You Teach The Machines

In an effort to be more positive, I'll share my lay interpretation of a paper on AI in drug discovery. Found it on arxiv, don't know the authors, have no stake in the research. Read on for something constructive to offset the antics of Big AI.

Liu, Xiang, Yi-Ming Ren, Mustafa Hajij, Pietro Liò and Guo-Wei Wei. "TopoLIA: Topology-Inspired Latent Interaction Attention for Drug Discovery." (2026).

https://arxiv.org/pdf/2610.13006

The authors are mathematicians and computer scientists from the University of Georgia, UCSF, and Cambridge. The senior author Dr. Wei also has a foot in molecular biology. They're figuring out how to teach computers to simulate how proteins do their work, and proteins do all the work in your body! The key challenge they tackle in this paper is how to simulate complex interactions between proteins. This is useful because protein-protein interaction is how most stuff happens in our bodies, good and bad. It matters because drugs often alter how proteins interact, and if you can use a computer to find helpful drugs, you speed things up and get new drugs faster.

The drug could be simple and get in the middle of a protein-protein interaction to change what happens. Like Venclexta for leukemia. Venclexta messes with the "bad" proteins that cancer cells use to stay alive longer than regular cells. Or the new drug could be complex and force an interaction with your "good" proteins. Like bexobrutideg for lots of cancers. Bexobrutideg grabs on to cancer proteins and glues them to a "good" protein in your body. The good protein is like a target on the bad cancer protein that tells your cells "hey! Get rid of this nasty protein!" And your body does!

Their discovery is called "Topology-inspired Latent Interaction Attention." The crux of this fancy term is they've taught a computer program to account for the complex shape of proteins (topology) and then simulated how different parts of a protein interact with different parts of *other* proteins.

"Latent" means their computer program distills interactions down into the most critical parts. Like a secret handshake exchanged while two people are having a lengthy conversation. This is efficient, so their computer program has enough time to figure out the many, many aspects of a protein-protein interaction. Like if there were five secret handshakes exchanged while people have *very* long conversation. Ultimately, it's a way to focus on the important parts of interactions and ignore the other stuff.

Interaction is, well, interaction between proteins.

"Attention"** means their computer program learns THE MOST important parts of protein-protein interactions. Like if two of the five secret handshakes mattered more than the other three. What to pay attention to.

They test their computer program using standard benchmarks, and find that it does pretty well! Their most important finding is that it can simulate protein-protein interactions between known proteins and proteins that are rare, newly discovered, or have never been seen before. You know, all the bad disease proteins!

I like this work because it's an example of the good stuff happening with AI outside the bubble. It's using AI to do really hard, important work that wouldn't have happened otherwise.

**Attention is what blew up AI between Google's publication of the "Attention is All You Need" paper in 2017 and 2022. Sam Altman and his backers went against a fragile, new industry norm of restraint and used Google's discovery to start the Race To Nowhere. Thankfully, others are doing the harder work of figuring out how AI can benefit humans and our world.

Before it was a book, You Teach the Machines was an AI literacy workshop...

As of November 2026 I've helped over 900 people from California to Maine understand AI, how we got here, face fear and move forward. Whether they use it or not. The workshop helps people from 8 to 80. I make it as fun as possible, but not gonna lie - the way the industry in the US is going about it, we cover some dark stuff.

Video of a short-form example in Burlington, VT June 2025 is here. Video of a full workshop in Brooklyn, NY October of 2025 here.

Reach out if your community wants to understand how you teach the machines. I don't charge book clubs, libraries, schools, or community groups. Ideally you'd partner with a local bookstore to support sales of the book through their business.

Events:

Upcoming...

1. Private book club, Ocean City, NJ December 14

Past (# of attendees)...

1. Macsherry Library, Alexandria Bay, NY October 31, 2026 10-11am. (16)
2. River Hospital, Alexandria Bay, NY September 16, 2026 (15)
3. Sherman's Books, Portland, ME August 28, 2026 (5)
4. Somesville Library, Somesville, ME August 26, 2026 (10)
5. Hawn Memorial Library, Clayton, NY August 22, 2026 (25)
6. Jesup Library, Bar Harbor, ME January 2026 (10)
7. College of Computing and Informatics, Drexel, Philadelphia, PA November 2025 (20)
8. Springboard Enterprises, New York, NY October 2025 (25)
9. Gowanus Studio Tour, Poursteady, Brooklyn, NY October 2025 (20)
10. Bala Cynwyd Library, Bala Cynwyd, PA September 2025 (25)
11. Belvedere Tiburon Library, Tiburon, CA August 2025 (40)
12. Thousand Island Park Library, Clayton, NY August 2025 (10)
13. Thousand Island Park Chatauqua, Clayton, NY, August 2025 (20)
14. Causeway Club, Southwest Harbor, ME July 2025 (40)
15. Grindstone Island Winery, Clayton, NY July 2025 (10)
16. PechaKucha 40, Burlington, VT June 28, 2025 (Book published June 27) (100)
17. Colby College, Waterville, ME March 2025 (60)
18. NIH CTSA Consortium, Bethesda, MD November 2024 (400)
19. American College of Graduate Medical Education, Chicago, IL September 2024 (40)
20. Thousand Island Park Chatauqua, Clayton, NY, July 2024 (25)

Transcript of a portion of the workshop from Poursteady in Brooklyn:

"If you choose to use it productively. Yeah, so that's a really good, that's a really good point. There's there's a point of view with technology: am I using it deliberately, or am I just like depending on it mindlessly? There are a lot of amazing things that have come out of those same companies, including social media, right? Like some of you may be here because of social media. That's a good thing, right? You may have heard about it.

But at the same time AI has been used, especially in the last five years, six years, to make our interaction with digital systems more and more compelling and harder and harder to step back from and avoid. So my biggest, my biggest argument with the AI industry as it is now, is for the casual consumer like, "hey, we're just using Chat GPT or Gemini or Claude or one of these things," or also "hey, I am consuming social media," is it's really really built around giving us those hits that will keep us coming back. So I I can go off on tech dependence all day long.

I'm gonna stop there, but it's something to be aware of when you're interacting with AI systems, that those systems, they're they're trying to charm you. They're type casting and all these things because they really want to get you using them and keep using them. They want you to like pay for the premium subscription, right? That doesn't mean they're all bad and that you can't be productive in line, right?

Um, we've talked about a lot about all the bad stuff. We're gonna really really try hard together to focus on the good stuff. Mary Jane just shook her head. Ha ha. Um, and I'm gonna, I'm gonna blow through what I have to say and then and then we can talk some more together, uh, and then we'll switch over to uh what I'm personally more interested in than being here to talk about AI again. Um, 'cause actually I didn't really want to write a book about AI, but haha. Um, and we'll do, we'll do some coffee stuff. So uh, a couple of my anecdotes, and then I'm gonna ask y'all for some um, why, why am I not like ready to throw the baby out with the bathwater and cut the cable and walk away from all this stuff?

My uh mother in law, Mary Jane's grandmother, was staying with us this summer. She woke up, almost fell down from severe pain in the back of her knee. She's 82. She doesn't slow down, but this stopped her. She was up early when this happened, and um sat at the kitchen table and was googling over and over, what's going on, what's going on, what's going on, and was getting more and more scared. It was, she was not getting helpful information. It was scattered. Uh, I I got up, came downstairs, and and um could tell something was wrong and asked her, hey, what's up? And she said, I have this pain in the back of my knee and I can't find anything on Google, it doesn't make any sense. So we sat down in front of a chatbot and started a dialogue that um, we built up enough context, 'cause that's really important when you use these things, to interact and build up context. We got to the point where uh it told her, oh, you need to go to the hospital right now, because that's either deep vein thrombosis or a baker's cyst. And we got there because like, she basically said, well, I just tried this vitamin and I just stopped taking that medication and then I did this and then I did that. And she put all of this information out there that through the normal like ranking algorithm was just scattershot, but because of the context that she built up, and because of the pre training that the model behind this chat interface had, um, knew what to pay attention to. It said deep vein thrombosis or baker's cyst, go get an ultrasound, here's the two sentences you need to say to the doctor in order to have a really quick and productive exchange. She went to the local critical access hospital and had an ultrasound, and the deep vein thrombosis was ruled out. It was a baker's cyst. At 82, she had a better health outcome because of this technology. You know, she may or may not have shown up at that hospital uh and been triaged appropriately. She could walk in more informed, right. It could have been wrong, could have, there could have been misinformation, right, but um it was better than what she was getting from technology without it.

That's one of my positives, and I'm gonna ask everybody else for positives. The second is uh a a study that I I worked on at CHOP, um, is similar to a study that's being done in pediatrics around the country. So, a uh there's a certain type of rib fracture in children that is, when appropriately evaluated, is like a dead on marker of physical abuse. The way the rib breaks, how it breaks, how it shows up in the X rays. Um, there are very few pediatric radiologists who can look at a chest X ray from the ER and say yes, that is from physical abuse, this kid needs to be uh assessed and protected. Um, there's a high rate of of that being missed by pediatric radiologists who haven't had that training. So there are researchers around the country who are working to teach machines to to appropriately and correctly identify that type of rib fracture, so that all the pediatric radiologists who don't have that expertise, and all the radiologists, and all of the community hospitals where there isn't a pediatric radiologist, can appropriately catch physical abuse in children and those kids can be protected. Um, these are the kinds of things why I can't just be like, f this stuff, f these guys, I'm out. And why I think it's important for all of us to be more informed and be critical consumers, that we can get to the good stuff and to collectively like marginalize the bad stuff.

You just got a positive. Cooking, cooking. Yeah, take a picture of the ingredients you have and it'll help you, give you ideas for what to do with them. And you only lit it on fire once. Like sourdough baking, sourdough baking. A large, 60 pizzas for example, like helping create a recipe. So I think we're a recipe. So yeah.

Derek, I was gonna say, kind of expanding on both of those points, the democratization of knowledge, where before a lot of this was extremely specialized. Before literacy was super high, everything was gatekept by guilds where people had specific knowledge of a certain thing. Literacy increased, people were able to learn more and do things at home and bring it to themselves. This is the next level of technology where I can go on, who has no formal experience with coding, and being able to buy code, new systems or something, like help me run something in my household I had no access to before this. Not good. Yep. That reminds me, I fixed my ACBC.

Yeah, I have found Chat GPT, two examples recently, to be really good collaborators in complicated mechanical engineering tasks. That I mean, in one way it kind of is eliminating a job, you know, like this is CEO position and I get to like, you know, have expertise here when I can afford it. But the truth is a lot of time you can't afford it. We've had mechanical engineers and we've lost them in the past. I have more of like an MFA of design and been making stuff with my hands my whole life. But we need to improve the laminar flow of the water coming out of the course taking machine. And I had a month long conversation with Chat GPT where it was doing all the math, and we kind of figured out how long the little syringe tubing needed to be inside the nozzle to do it at that temperature, which, you know, conceptually I've seen a YouTube video of making Lambert flow, but this was literally doing the math. And we were narrowing in on like what you can buy online, where you can buy it, how long it has to be, what the diameters have to be. And like, that's like a fluid dynamics graduate student, 200 thousand dollar job hybrid specialty. And you know, we're a struggling small company trying to make complicated stuff, and like I can accept it now. Yeah, you're so desperate that you're hosting AI. Yeah, no. I mean, but like, you know, I would love to have like genius engineers around all the time, but I, yeah, that's not my cool, hasn't manifested itself yet.

Another good one which um is actually re bringing back uh human interaction and especially through customer experience. A good example here, of course, that I do a lot of the CX support and about customer experience. Customer experience, half of my um day was sending out the same templated email, this is how you um connect back to Wi-Fi. And that caused me to also take a lot of other uh interactions to be less thoughtful. There's so much time was taken up for one thing. So with um our integration with um Fin AI, who intercom, that part's automated now. When it comes to actually speaking to the customers, I'm have more free time to be able to interact with them, get to know them on a real basis. Same thing if you're doing um marketing or social media on management, you can now automate the basic mundane crap that doesn't actually matter, like post this to three different spots, time this out, you don't have to wake up to do that. And it gives people more opportunity to write their own copy and do what actually matters. Yeah, just because Chat GPT or this can write the copy for you doesn't mean it has to. And you can automate the part, my friends, with the BS, and then you can do what matters and not live in between.

Yeah, yeah, who's a parent here? Okay, who's a parent of a a like high school or college age kid? Okay. Um, who, who's ever seen their kids struggle because one of their teachers isn't great? Yeah, who in the room has ever had a teacher who wasn't getting through to them? Haha. These are class of 2025 college grads. OK, who, who are some of the most, they are the most interesting people to talk to about AI because they did 2 years of college without, two years with. MJ and I started this podcast called You Teach the Machines to Talk To. Actually these were our first guests. Um, true, true. Yeah, about that experience, and you should check it out 'cause it's fascinating. Um, who'd be willing to share how your experience using uh AI helped you learn more, get a better grade, when your teacher wasn't getting through to you? MJ, plant.

Yeah, I'm a huge AI skeptic and I am gen generally negative about it, but I have had positive experiences. I took a probability class last year um for my statistics major that absolutely wrecked me. Like, it was in the math department, I wasn't a math major. The professor was like super classic, like long gray beard and glasses, like incredibly brilliant mathematician who did a lot of super theoretical math. Um, and so the way he taught the class was through a lot of math theory, and I was more interested in probability from the statistics side and from like the modeling side. So through a lot of these units I would sit in a lecture and he would teach about probability through a sort of math theory lens and it would go right over my head. And like, the way that I Learned better is through like concrete examples of applying math. And I'm also super visual, so I need kind of it like spelled out for me in like a physical way. Um, and so I would sit through one of these lectures and then use AI after the class to basically give them the topic that I had Learned and ask for an explanation in a couple different ways. And there's like a benefit to the ability to interact with it that's different from just googling. You can just Google like whatever the probability concept is, the Poisson Curve, and and get a page, a static page that gives you information. Of being able to interact with AI, I could kind of, it kind of Learned how I Learned better, and could could word it in a way that my brain process better, come up with examples and problems for me to try that like got through to me a little bit better than the theory. So yeah, there you go.

Yeah, so what's a common theme from all positives today? Was anybody just like, you know, what's the animated movie um where the people are in couches, just like WALL-E? WALL-E, uh, yeah, WALL-E. Were any of those examples like, somebody is like, right, they were all examples of augmentation. That's how this works for us. And that's where there may be an element of dependence, in the same way that I now depend on navigation apps to to get places. Um, but that's on my terms, okay. Augmentation is is what you are instinctively drawn to in these uses.

What's just took like one call back to fears. Kids that are younger than than the crew here. Um, with Connie, chef, chef, alright I got it right, they're both blonde like roughly the same shape, he's bigger, he's bigger. So chef. Um, Chef, I'm sorry to tell you, but your prefrontal cortex is not developed, right. It will be someday, it's gonna be great. Alright, but your, your the kids in your lives aren't necessarily developed to the point where they can disambiguate between, oh, this thing's augmenting me, and I'm just gonna be like, I'll just take it, right. Um, that's why it's important for you to be an influence on the kids in your lives, and to be understanding of teachers who have been dropped into this right after they got out of the pandemic and are just like, really, okay. Um, but that augmentation is the sweet spot, as so far as it exists now.

Um, the the sort of so what, my recommendations um based on my own research, my own behaviors, experiences, but then what I've Learned from talking to, you know, over 800 people um in person, are: don't give away your data. In your business, in your personal life, consider your data to be an asset. Now, in this forum it's hard to, like for me to tell you exactly what that means, but uh Derek, how useful were 10 years of customer service emails in fine tuning the, I mean, it was everything, that's our dataset. Yeah. Um, your data can be your business's data, your personal data. My writing is my data. I I am getting ready to uh fine tune an AI using my book. Alright, and then from that, do some experiments on, like, could I train an AI to be my book? Right. I'm not gonna just give that book to any of the big commercial providers, dump it in there, even though Paul thinks I should. Something to do for you, they probably already have, they probably already have it. Just get it in, yeah, just get it in. I'll just upload it. But consider the data you generate to be an asset. And if you are in a situation where you need to acquire or start using a new digital system, every digital system that you, that you buy or buy into these days is, has been funded as a startup, in part, digital information system, in order, based on the value of the unique dataset that they collect. Like the venture capital market switched in 2015 to, it's not just the team and the product and the market, it's those things and it's the unique dataset. So the investors for 10 years now have been buying into, putting money into, businesses that generate a unique data set. Um, when you subscribe to a business uh service software, it's all hosted now. That's because they want your data. It's a unique data set. The terms and conditions that um that company acquires and then uses your data are always favor the company, and you just like, yeah I'll click fine, let's go, low stakes. So what, who cares? Higher stakes, like, thank god you had all those emails. Does Intercom also basically allow you to download sort of every product of every interact, every customer interaction, so that you have it instead of just Intercom? Yes, great. Yep, that was the right decision. We also set up a additional background forwarding so that we can have it saved on a Gmail as well as Hubspot for our CRM. Right, so we have it moved over to several locations. Yeah, they're all using it to train, of course. So you're contributing to the commons, which is another way to look at how your data gets used, right. It, you can hoard and protect and be conspiracy minded, or you could just like figure out what's important to you, what the stakes are, when it comes to your data.

Um, tasks. Everything, every positive we discussed was about like, I gotta, I gotta get x done. Be deliberate in how you interact with AI. Teach the people around you, model that behavior, teach kids. What do you wanna do? It's not like just sort of sitting back and using AI. It's like, I gotta, I gotta do this task. And then use it to save yourself time. This is something that I, that I've Learned in lots of discussions with people, that I've Learned on my own. Um, great anecdote. I had a consulting project, I did probably a day and a half worth of work in a, in a half a day, and could spend the rest of that time differently. I was blown away with not just how I could do, do research on the topic I was consulting on, but then how I could synthesize it, compile it, and and use AI to help me tell the story that was in the information and put it out there. Get time back, right, by using these tools. And then as best you can, like, have your output be based on value, not on hours spent, so that you don't end up like adding to the gig economy sectors, right, which is something Mary Jane can speak very cogently about. Um, but use it to get time back. That's on your terms.

Um, and the most important thing, something that I that I've, I've Learned um, is, and it's become even more important to me: you don't just use Chat GPT. Don't ever get reflective about which AI you, you use, because they're all different, they all have pros, cons, flaws. Ask the same question or do the same project on multiple. That does one thing, it allows you to see where they're inconsistent with each other, and that inconsistency is a signal, oh, why is that inconsistent? Back to critical thinking and digging into like why, why, why is this, why are these answers different? Um, it gives you an opportunity to think critically and compare contrast, do your own research and thinking. So just in that, that pattern of behavior, it's sort of just better for a human brain and using those strengths more. Um, also it's, for me it's become, you know, now when I don't agree with one of these companies and how they're going about this stuff, I'm out. You know, I'm not gonna tell you which one I don't agree with, you gotta come up, come up with your own answer to that. Um, but recently I've been like, nope, I'm done, I can't.

So coming back to the environmental cost of AI. It it turns, yeah, AI turns electricity into heat. That's how it works. That's the physical industrial process. Electricity goes into a computer, bits and bytes come out, heat gets emitted. So it consumes electricity, it consumes a ton of electricity, it consumes a ton of water for cooling, both at electricity generation but then also in the data center. Um, you know, everything in our modern life consumes resources. We're all, we're all often trying to be good stewards of these resources. When you use multiple AIs, you set yourself up to start to be a critical thinker about, hey, like which one's even like deciding to, you know, not put a new data center next to a new power plant on top of Marcellus Shale gas. Because a year, a year and a half ago, it was all about the new nuke plants that were being commissioned and or recommissioned. And now, very quietly, the the AI industry is like, now they're parking data centers on top of of of a shale gas. So it's happening, right. Shh, right. So if you use multiple, you can make those, you can use your critical thinking to make those kinds of decisions.

Um, most importantly, stop my foot, get out and talk to people. Get, get off your phone. It's hard, I don't judge anybody who sits and scrolls, I do it, but connect in the real world. We've all been burned by the incorporation of AI that basically makes technology more addictive. We've been burned by a pandemic that locked us up and put us in front of computers for what had been in personal interaction, zoom, blah blah blah, right. Disconnection is the biggest risk to us, all of us here. That's my view. My biggest fear is disconnection, which is why I don't do these talks online, which is why I go, I've been to seven states now and I do it in person, because this is what makes us human. What makes all of these things turn into minor problems and not overwhelming problems.

Alright, and also, if you'd like to keep up with, with this, with the change of this stuff, as much as you, as as we can, MJ and I um have this podcast. Uh, I've, I can say objectively that the book is really good. Um, a 15 year old has gotten something out of it, an 85 year old has gotten something out of it. It doesn't dumb things down, it does make really complex stuff accessible. You can tell what my style is, um, and that, that comes through in the book. If you wanna get the book, um, let me give you a sticker."

Artist and entrepreneur Gretchen Andrew shares her perspective on life, beauty standards, and the civilizational stakes of saving independent journalism. Gretchen turns the tables and learns from MJ as we go NEXT LEVEL with Gen X, Y, and Z in our multigenerational take on tech in society. Check out Gretchen's company Intelligent Attribution and her (literally) genre-defining art. Important work, we learned a lot!

Understand AI and move on with your life, with or without it. A quick reading from my book You Teach the Machines. In print, eBook, audio. Amazon | Bookshop.org | Libro.fm

MJ interviews Kay Koplovitz, Forbes Top 250 Innovator, CEO of the first satellite cable network, venture investor, and founder of nonprofit Springboard Enterprises. Springboard accelerates women-led startups, over 950 to date creating $76 billion in value!

Kay: Overcoming challenges together has a lasting positive effect on our value. How we value ourselves. And I'm not talking about dollars.

(0:21) [Intro music plays: "Where, oh where are you tonight? Why did you leave me unread on my phone? I searched the world over and thought I found true love. You met an AI and poof, you was gone."]

MJ: To our listeners who can't see, we were all bobbing our heads and dancing to the music. It's a great way to get in the mood a little bit. But I'll go ahead and introduce our guest today. Kay Koplovitz, who is a businesswoman, entrepreneur, and author who has spent her career looking to the future.

She was the first woman to head a television network when she founded USA Network in 1977. And she was a visionary, helping sports television reach cable by negotiating contracts for the MLB, NBA, NHL, among others. She launched the Sci-Fi Channel, chaired the bipartisan National Women's Business Council, and used her platform to launch Springboard Enterprises, which is a global network of entrepreneurs, investors, and advisors accelerating the success of women entrepreneurs in technology and life sciences.

She's a champion for female entrepreneurs and an inspiration to young women everywhere, and an inspiration to me. Kay Koplovitz, thank you so much for joining us today.

Kay: Oh, what a great pleasure to be joining you for your podcast today. I'm really looking forward to our discussion.

MJ: Yeah! Well, so you've spent your career sort of looking to the future, innovating. I know that you started the Sci-Fi Channel partly because you thought that it was what we were all headed towards, right? And now we're kind of at the forefront of that sci-fi reality.

Kay: Hal is beckoning at our door right now. People here listening know who Hal is from 2001: A Space Odyssey.

Kay: He's still around.

MJ: Yeah, I think that a lot of our listeners are friends of mine and people my age. And I know that when you were in school, you did your Master's thesis on satellite programming and how it could sort of impact the social order by spreading information.

And AI is kind of another way that we are spreading information. I wondered if we could just start there with your experience working in media for so long. How you think that the spread of information is changing now, and for people my age, what feels different now than it did when you were an expert in your field with cable? What feels the same? Is this a familiar beast, or is this a whole new ball game?

Kay: Well, technology always changes everything. I've been present for the change at various times. Way back, I wrote a Master's thesis in 1968 on satellite technology and how it could change communications around the globe. It was something that we didn't have access to.

And for people that are listening, historically, we were in a Cold War with Russia and China. We didn't know what was behind the Berlin Wall or the Great Wall of China. Today, both of them—one's gone completely, the other one, the Great Wall of China, is a tourist attraction today—but we didn't know what was there.

And I thought geosynchronous orbiting satellites, high-altitude satellites, only needed three to communicate all around the earth. It was a real breakthrough in technology and potentially a big breakthrough in people's ability to communicate with one another.

So you have to start there with the satellites and what they did to change communication around the globe. So things advanced, computers came along for personal use, the internet sprung up, people started communicating through the internet. And eventually, we launched cable networks, USA Network in my case, Sci-Fi.

And Sci-Fi, I was not a kid who read sci-fi comic books and things like that. But I grew up in the age of Sputnik, President Kennedy challenging us to put a man on the moon. You have to have vision. Students today, if you want to innovate and be an entrepreneur, for example, you need to have a core position that you really, truly believe in and want to really reach for if there is no solution yet.

And one way to learn about that is to actually jump in and work for a company that's a young startup company. You can learn a lot of things working for big corporations, but you won't learn those skills because they're not the same skills.

And I always say to students, if you really want to learn, "Well, am I really an entrepreneur? Can I really do this?", the best way to do it is to start at a very young company and see how it operates and see what the challenges are and learn from those experiences. When you're young, it's the time to do it. It's the time to try different things. You are free to try.

And today it's free to access. When I started out, the television market was pretty closed. Cable television, people were like, "What's that? Why do we need more than three networks?" They challenged everything that we wanted to do. And I said, "Well, there's a lot more out here." And to me, it was opening up the global communication sphere.

And that was using high-altitude satellites to communicate around the world, to communicate with people directly on phone services and things like this around the world. So it's gone back to also low-orbiting satellites. You can launch thousands of them; there are millions of them out there.

And so we all know, for example, in the war-torn country of Ukraine, their communication is basically by Starlink and their field operations. But furthermore, for people with just communicating with each other, the streaming that has overlapped what the cable networks did, now the cable networks are being disrupted by the streaming networks.

And so communication has become literally among billions of people around the world. When we started off, it took a few years to get to like a million people, and then get to ten million people, and then get to twenty and thirty, fifty... it took time.

Today, you can instantly have the opportunity to communicate with billions of people around the world. Now, what does that mean? It's hard to communicate with a billion people at a time, you know?

MJ: Right.

Kay: But also as a young person, your point about getting into entrepreneurship now, this being one of the best times to start, we have access to everybody across the globe and all of their information. It's easier than ever to just get your feet wet, right?

Kay: It's easier than ever, you're absolutely right, but the challenge is to gather your own community. Because there's so much competition out there. There's so much opportunity out there.

And people say to me, "Oh, you know, the consolidation of the broadcast networks," which is happening. The consolidation of the cable networks, which has been happening for the last couple decades and now really more so. Those are consolidating and coming together.

The big challenge is not "can you get in?" You can get in. Anyone can get in with a cell phone or a desktop or a laptop or anything, an iPad, whatever you have. But who are you going to reach? Are you going to reach your own community?

And that's really where a lot of influencer marketing has come into play with a lot of celebrity stars from Hollywood, television stars, and people say there's not enough creativity. There are so many companies that have launched on TikTok, that have launched on, certainly, YouTube. There are many, many different opportunities.

What is your goal? What is your business plan? How are you going to support this? This is, and advertising revenue, of course, has supported Meta, Facebook, and how are you going to create a business? First of all, establish yourself. What is your position? Is it clear? Can you attract your community? And then how do you want to monetize that community?

Is it a freemium model? Is it free at first and then we'll charge you? I think we're all familiar with that. Or is it just advertising-supported like FAST channels that are available through like Roku and all the manufacturers of sets of all kinds and computers of all kinds have advertising revenue?

It's very hard in the vast community of billions of people to find your niche. But if you do have a strong following on your niche, you can create businesses that way. It's not a matter of access, it's a matter of performance in the end.

MJ: Right.

Jeff: Performance. A couple of things that stuck out to me from what you said, Kay. One: the phrase "gather your community."

Kay: Let me give you an example. I'm a whitewater rafter. And the people who are in whitewater rafting who are the guides that I've been on Class V trips with, they show up in different parts of the world. It's just this community of these nutcases who love to go whitewater rafting. We just loved it. I mean, it was just so exciting.

And then we'd go to South America, we'd go to Chile, and the next time we'd go over, we'd be in South Africa and the same guy—"Oh, hey! It's so good to see you again!"

MJ: A community that you found of rafters!

Kay: That's sort of fun. And then you can say to them, "Hey, have you done this river and what should I expect of it?" Give you an example of something that's a small community that people are integrated together in and respond to each other quite quickly.

Jeff: You know, if you have access through all these different channels—streaming services like Twitch—if you have access, that is an incredible opportunity in that there's no barrier anymore. But without a community, you don't have a voice, right?

And a quote stuck with me from a student of mine: "Get over yourself and start the conversation you want to have." Because another point you made in a couple different ways was you have to have a strong point of view and direction. And having the conversation that you want to have is crucial when there's every opportunity to make more generic noise, content, whatever. But you're not going to gather a community without that point of view.

Kay: Yes, that is true. One of the things I would say—and I'm concerned about students today and trying to make choices among chaos—I have always believed that there is enormous opportunity in chaos. When everything is static, it's very hard to get in.

There's so much chaos right now that the other opposite side is true: there's just so much chaos, where do I plant my flag? How do I...? People know when you're authentic and when you're not. The thing that I worry about is I think social media is dividing us. I think social media started off to connect people, connect families, "share my videos" and this and that... all these sorts of lofty ideas which were wonderful.

But today, a lot of the business models are based on hostility. More: the more people are angry and shouting from different sides at each other drives up the use, the attendance, the participation. And I worry about how that aspect of it—that business plan, and let's be honest, the business plans of Meta and Google and companies like YouTube and companies like that—to some degree or lesser, they depend on that high friction.

And nothing has to be true. It's what you say is true. It may have nothing to do with truth. People can project a lot of lies and just make up things and try to get people to believe them. And I think that's really destroying our soul being in a lot of ways, and having people against each other, and then even family members against each other.

I don't think that's a good thing. And I'd like to go back to the idea that individual communities should be the challengers or the people who have the mission of that community and have their judgment as to what is the proper communication that they should be having. And if they don't, they'll kick them out.

And we had companies like that years ago, but today it's... I think students know what is authentic, but they drift into things, too. It's easy to be pulled into things by a friend or somebody that you know or somebody that you met and go down a path that is not...

Jeff: Or by an algorithm that's tuned to deliver dopamine to you.

MJ: Yeah. We're not just an AI podcast; we talk a lot about the influence of technology and social media. And because you have been in media since before social media, you sort of talked about how we went from like one or two cable news networks and now we have this influx of information across the board through social media and how it kind of divides us because fear sells, right?

You get more engagement if it's more extreme, and maybe the companies that are giving us access to social media are less concerned about the integrity of the information and more about engagement. I wonder, are there any pros when it comes to media specifically, going from like one or two cable news networks to everything at your fingertips? I wonder if you've seen differences or if you think that there's any benefit to that.

Kay: I always think it's beneficial to hear different points of view. I don't think it's productive to have just groupthink. Whether you agree with "that's your groupthink" or somebody else's groupthink, I always listen to people that have different points of view than I have because I always learn something from them.

I don't have to necessarily agree with them, but I learn something from them about why they think the way they do. So sometimes they change my mind because I say, "Now, that's an interesting thought. They have a point there; maybe I should think a little bit more about that."

So, I think it's a benefit to have access. What I'm thinking about when I think about Artificial Intelligence and AGI: I think it would be great to be able to use technology to qualify for ourselves—as individuals—qualify what we're reading and understanding through these different social media platforms, people, individuals.

And it's kind of interesting because when you do research—and I use it for research just to look at things that bring things to my attention that I may not know exist because there are so many sources of information out there—I think it would just be great for us as individuals, or people in our group, to be able to get instantaneous analysis of what are facts or not facts that are listed here with what people are saying.

I think that's the next best step that we can make because I don't think we can really depend on regulation, like national, state regulation of any kind, self-regulation. Look, we had self-regulation in the cable industry for a long time. You know, and some of it was good and some of it wasn't. And I think this is true today too, but I think we have the ability to at least instantly today check the viability or the truth of what are these stats, what is this information that we're...

Here we are, we're talking to each other. Now, if we want to go and find out, well, is Kay Koplovitz telling us the truth or not? You could find out like that, you know? "No, she's just telling a story."

So I think there are ways that we're starting to understand, if we're interested people and not just there to, let's say, spread our—whatever we want—the message that we want, true or not true or whatever it is. This would be a great way to use the different platforms of technology that are coming into the core right now for us to be able to double-check ourselves.

We don't have to have an outside source. We do have outside sources now checking on the veracity of a lot of statements that are being made, let's say by politicians. Sure, there's a lot of that going on out there, but wouldn't it be great if just we as individuals could get the same just fact-check like that and say to ourselves, "Oh, I really thought I was believing this person, but actually what they've just said is not true. Here are the facts."

Wouldn't that be sort of cool? That then each of us could have that responsibility. Some people are trying to deceive you. There's all kinds of people like that out there.

MJ: It's almost like both the problem and the solution is the fact that we have access to all of the information, right? It just takes a little bit more...

Kay: It's overload! Our brains can't consume it all at one time.

MJ: But it takes some more personal responsibility, right? To care about whether or not the facts you're consuming are true.

Kay: Now, on the other hand, someone can use it for evil. They can use the same technology to, let's say, bring in people who they're spinning a yarn to and get them to believe it.

MJ: It's a double-edged sword.

Kay: Because they've said it so many times and people start becoming believers, and we do see that a lot today, let's say our political environment, we do see that.

Jeff: I wonder if you think that—I'm sure you've had it, the experience of catching a bot, whether it be Google's or Claude—catching a bot in an inaccuracy is actually a good thing because it teaches you to be skeptical, to ask follow-up questions, those sorts of things.

Kay: I don't know if I've really had the... I don't think about it as catching a bot. They make mistakes too. We make mistakes. Like I use it for research. It could be a contract. I could say, "I want analysis of the contract if I've forgotten something or need something out of my head." And boom, you get an answer. Well, okay, well that's... I better check that out. At least I find it very, very good.

Jeff: So I think we all can feel that there's a lot of chaos swirling around us right now. And Kay, you brought up that chaos can be an opportunity. MJ, your perspective is that there's a lot of chaos right now, but in that, there is opportunity.

Just coming back to that point you made, Kay, about for young people, a great way to learn a lot quickly is to work in a small company, a startup, a growth company, maybe not. How does that relate to this concept of there being opportunity in chaos?

MJ: I think that my entire generation feels like anything we do post-grad is kind of taking advantage of a chaotic moment, and that can feel pretty crippling. I think that there's a lot of uncertainty about what the workforce looks like moving forward, how different technologies impact the way that we experience the world, the way that we contribute to the world.

But I also think that if you can get over the lead in your stomach from that crazy uncertainty about what even the makeup of the workforce looks like, there is a lot of opportunity to be the people that are coming up with ideas of what it could look like—envisioning that future.

And that means that even if you're in an entry-level role right now, you have to be inventing what an entry-level employee does now because AI can sort of automate the basics of that role. So we have to be a lot more proactive about proving our value early.

As scary as it is and as much as it feels like it's setting us back, I really think that it's something that's going to push my generation forward because we have to much younger decide what our point of view is, decide what we want to say, decide how we can demonstrate our value to people that might employ us. Because generating sort of mediocre content is something that AI can do now, right?

They can summarize an email and make a PowerPoint. And so something that I've grappled with as I'm looking at the beginning of my career is: what do I care about? What can I do that is interesting? What are the questions that I can ask? And also I think it's sort of a lot of my life experience, including the pandemic and then AI, has sort of forced me to reckon with the fact that humans and human connection is something that is so important to me and something that is how the world is going to move forward, right?

Post-pandemic, I was so grateful to be able to be in person with the people I love. And I think that that gratitude is getting even bigger as I realize that interpersonal connection and human-first companies are the future because AI kind of automates all of the tech babble. And it comes down to who are you? How do you connect with the people around you in a meaningful way that only humans can do? And what do you have to offer, and what are the questions you want to ask, and how are you going to solve those questions? Those are my thoughts.

Kay: Yeah, well, I think you're hitting on something that's extremely important, and that is relationships—human, face-to-face relationships. And if someone your age is to say to themselves, "Well, I want to go into the art world and I want to deal with art," then maybe they want to go work for an art company.

Maybe they want to work for Sotheby's or maybe they want to work for an art studio. Even technology is changing how people access that, but being able to be there in the environment with the people that you admire and want to learn from is extremely important.

That's why I think, in many ways, return to the office is beneficial. I think it's been harder for people starting out only on screens looking at each other. You've got a connection to the people, but it's not the same as having that really personal relationship and understanding the other person.

And so you go into a company—let's say you're not really thinking about being an entrepreneur, you want to go into a company—even if that's a bank, you want to go into banking or something like that. I mean, it's hard to establish those key people that you want to follow when you never get to see them personally.

You don't get to sit around a room and have a cup of coffee with them. You're just a face on a screen. And I think that there's a good thing about being able to communicate with people around the world in all kinds of different media that we have available to us today, but I think you're focusing on something that's really, really important to the future of humanity. And that is personal, personal relationships with people. There's nothing that can substitute sitting around the table or working in an environment where you can go down the hall or around the various cubicles that people are in and talk to someone.

MJ: Yeah.

Kay: To bring it back to a startup, I think that there's something valuable about the foundation of a small company that is mission-driven and you're working in the chaos and it becomes so much about how each individual person in that company contributes to the mission, what they can bring to the table in terms of problem-solving.

Specifically in life sciences and tech startups where maybe you're trying to solve a healthcare issue. It's about the people that you're trying to help. There's a human-centric mission. And when you're in a small company, every person's voice matters, everybody is all hands on deck, and you have to bring value to the table in terms of your ability to jump in and work with each other.

Your experience working with small companies, with startups—how do you feel it kind of ties into the human connection piece? But the chaos, and it actually is an opportunity maybe for individuals to really shine and showcase their talents and figure out what they care about. What are your thoughts on that?

Kay: Yeah, well Springboard Enterprises, which is the non-profit accelerator that was launched 26 years ago now, the mission was to find women in science and technology who are starting companies and connect them with potential investors and advisors that they needed.

It's a perfect example of mission-driven entrepreneurship because it didn't exist. People said, "Oh, you know, women don't do that kind of thing." And I said, "Oh yeah? I think they do." And they said, "We never hear from the venture capitalists; no one ever comes to pitch us, no women ever."

They didn't know each other. So we had to go out and find the original companies that we brought in. And we were stunned when there was no internet for us; we didn't have an internet connection, we didn't have a website or anything like that. Six of us went out and just sent out to groups that business organizations, colleges we went to, we just tried.

And to our surprise, 300—over 300 applications showed up on paper. Whoa, what are we going to do with this? And we were very fortunate to have Stanford and Berkeley MBA students help us sift through them all and we came up with 26 companies. We said, "These companies can grow big, we think."

We're going to have to see what we can do with them and help them establish. So that's how we established our first initiative in bringing women to the marketplace. And when we actually had a Demo Day in January of 2000, some 300-some people came to listen to them. And they were like, "Where did you find these companies?"

You know, we found them by going out to look for them. We didn't just sit in a room and wait for them to show up. And I think this is like a mission-driven thing now because these women over time have raised over $14 billion, have created over $76 billion in value for their investors, 28 IPOs, 240 M&A events... I mean, they are really kicking ass out there.

And I really, I'm very proud of the work. And talk about something that's mission-driven, this is a community of people who are there to see them succeed. Not punish them, not interrogate them, but to give them tough love. I mean, tell them the real truth about what they're doing and how they have to change and what they can do to help them through the gates. And this is a community, what you're talking about in many ways. Mission-driven community make a difference.

MJ: A community that you found, yeah!

Jeff: I'll reflect: MJ and I were fortunate enough to attend the 25th anniversary gala for Springboard Enterprises last October. We got to see Kay there in her element, in her community. And I will say that afterward MJ said something to me that was really important. She said, "Dad, I've never been in a room with 250 women and I didn't hear anybody say 'I'm sorry.'"

Because every woman in that room, every person in that room, had a strong point of view, was probably looking for chaos to take advantage of.

MJ: Yeah, I think that what I meant by that—that's "sorry" is kind of a running thing in our family where my dad noticed that my sister and I apologize constantly for things that we have nothing to be sorry for. And I think that it's sort of an ingrained female thing to be a little sheepish.

And so when we were younger, he would always say, "Take your sorry back. What do you have to apologize for?" And I think that I used that as an example because I was in a room full of women who held eye contact, had something to say. There was none of the sort of socially ingrained spatial apology where women feel the need to make themselves small, right?

These are all women who have voices, are confident, who know what they want to say, and when you have conversations with them, they are unapologetic in taking up that space. And it was honestly one of the most incredible professional experiences I've had. It's the beginning of my career, so it's a short list, but I've never been in a room full of women who were all so inspirational. And it was a really profound experience for me. So thank you for that.

Kay: Well, I'm glad it was a good experience for you and there are legions of us. And I was at the Femmys Awards this week here in New York, all women in Fintech. Men and women in Fintech supporting women in Fintech. And Fintech is chaotic. There are massive changes in the financial market going on, whether you think it's crypto or you think it's blockchain or you think it's stablecoin... I mean, all this stuff that's going on in finance.

This is the time for people to really come out in front, you know? Because it's very chaotic and it's also very exciting. And you know what? Entrepreneurship is exciting and frightening at the same time, which is probably why it attracted me, because it's... it's frightening.

But it's also exciting. And some days you get up and the problems are so heavy, you're just saying, "How am I going to get through this day? Or how am I going to pay my employees?" And then you say to yourself, "I can do it. I can get this done. How am I going to get it done?" And then I go, "Okay, I'm going to do this first, then I'm going to do that, and I'm going to get it done." And that's how people, you know, you move forward one foot at a time.

Jeff: You do it together. My experience in startups was one where—you think about a startup as maybe the stereotypical male startup of the "sainted soul," you know, Steve Jobs or whoever. Zuckerberg. But my experience in startups is that it is way, way more of a collective endeavor.

And startups by their very nature are going to be smaller than enormous multinationals, and so you can be more collective. There are fewer layers of middle management and command and control because it's a smaller group of people so you don't need those things in order to be productive.

And when you wake up and you don't know how you're going to make payroll, it's much more likely that you're going to have a conversation with somebody in person across the way who might offer a perspective that gets the whole organization to a solution a lot faster. And that kind of shared sharing in the experience of overcoming adversity is a critical element of being human and of humanity in the sort of best and most positive sense.

I clearly have an agenda here which is to encourage any of our younger listeners to consider working in a small company.

Kay: You're absolutely right. It's a small group of people, you've got a mission, you've got a business that you're trying to get into the marketplace, and you come to the table and each person—it might be three people, it might be five people—but you're there together. And overcoming challenges together has a lasting positive effect on our value. How we are as people.

In fact today, I have another of my group from USA Network. We've been out of USA for 26 years, 27 years maybe. We are still together. Every other month, we get in a different locations. Some of us are in New York, some are in LA, some are in other places that they are, in Europe or whatever. We still get together because we had that amazing experience of building something from an idea to a powerhouse together and we had to bridge a lot of challenges.

And there isn't anything that can substitute it, really, in a lot of ways. And it's not about money. It's about our value together and what we did together. And I think to me that's one of the most exciting things. Now sometimes people get that from being in larger corporations and people, but if you join a bigger company, it often depends on who's leader of the team that you're on and what is their success.

It may have nothing to do with you in terms of success or failure because if that person falls out of place, all of a sudden you're lost. Your team has to go over to this team and then this team may not want you on their team and people navigate their way successfully to the top of corporations. I'm not saying they don't, but it's a different experience than being an entrepreneur.

Jeff: I'm going to plug two books today. Yours in just a moment, but also Julie Wainwright wrote a book called Time to Get Real about her experience building The RealReal. And she's got a great section—I actually taught it in my class in January—she's got a great section on that corporate environment and how it works, and she's being sort of empirical and objective, and how you're going to have a different experience there than you are in a smaller company where you do have more of a natural environment for pulling together.

Can I read you a couple of quotes from your book? It's called Bold Women, Big Ideas. You may not have the whole book memorized after 25 years because you published this in '01 or '02.

Kay: Remind me what I said!

Jeff: I read it over January into February. I found it to be incredibly relevant to today's moment because you were chronicling the peak of the dot-com capital craze, the peak of the dot-com facilitated rapid change in business and society because of new technologies, not just internet but also biotech as well.

And a lot of the dynamics that you describe are what we're experiencing today with the latest emergence of a new disruptive technology that we're all having to adapt to. Bear with me, but I'm going to ask you a question at the end of a few quotes. So, the first is not a quote of yours, it's from a mentor of yours, Reuben Mark, who was CEO of Colgate.

And he said to you, "Kay, it isn't enough for you to be a role model as CEO. Just because you're the CEO doesn't necessarily get others to realize that women minorities are worthy of it. You've got to be proactive. You've got to inspire others to think and act the way you do. If you really believe in helping others, that's your obligation."

Take that as the backdrop to start. Then you say, "There's something seductive for me in traveling into the unknown. The journey itself thrills me and I don't think I'd ever feel altogether happy if I didn't know there was risk involved. Surely the risk of the unknown that takes me down Class V whitewater rivers and to the top of peaks, and it's not so different in the business world."

And then you say, "The simple truth is that once I get a big potent idea, it moves me to distraction. I feel compelled to try to move others with me. At the risk of repeating myself, I'm very motivated by the power of ideas."

One of the ways that you inspire me is that you have built a career both in service to a community and in the business and material success of that community and of your own. So you didn't sort of go off sit in the nonprofit world at a key juncture in your career and you didn't go off and just slay dragons as an investor at that key moment in your career. You've found a way to balance both service and success in the business world.

And you've done that for 25 years and that room at the Springboard gala was full of just a small number of the people who've been inspired and empowered by this duality. My question is: how does that continue to sustain you? And a two-part question: how does that continue to sustain you, and what's the big idea that you're most—that you're locked in on right now? I could imagine there is one.

Kay: Let me go back to Reuben Marks for a moment. Because what Reuben was saying to me is it's not enough for you to be the leader of the change of what a leader looks like. At that point in time, studios had bought in. One of them was Universal and Paramount, the other was Time Inc.

And Sid Sheinberg, the head of the president of Universal, every time I walked into his office to see him if I was in LA and stopped by to see him, he'd say, "Here comes the CEO. You don't look like a CEO," he'd say to me every time. And I said, "Get used to it, baby. This is what a CEO looks like."

So we would get—we had this little thing that we'd do all the time, you know? But I was trying to say to him, "I don't have to look like you." Reuben Mark said to me something else that was important. He said if you really believe in diversity as a key element of your success in your business, then you have to motivate people internally and you have to adjust their bonuses to actually perform and have diversity in...

So the head of sales and the head of distribution and the head of advertising and the head of the... within a corporation, part of their bonus had to focus on the fact that you wanted to have diversity within the company. You wanted to have different—I didn't want everybody in my legal department to be Jewish. I didn't want everybody in my sales department to be Italian. I didn't want everybody in my... it was sort of a little bit like that. And I said, "You know what? Reuben's right. I've got to do that." I had to change the motivation for people within my own company even though I was a leader of them all.

That sort of thing I thought was really kind of important for people to understand. Why do I, after 25 years of, you know—when I started the whole thing was Springboard, it wasn't like I had this idea that I was going to do this for the rest of my life. I just wanted to get it off the launch pad.

But then the dot-com bust. And I knew that women felt they were going to get shoved out and I said, "I can't leave them out here. We're not leaving them, we're going to go on." And we went on, we went on to Washington D.C. that year, we went on to Boston that year in the year 2000, and we made it very clear: we're going on. You're going with us.

I didn't want them to feel that they were being abandoned because everybody was being pushed out, but it felt very bad for women who had just started to get in. Actually out of that first year, we had five IPOs eventually. So it was, you know, they were companies that were actually well on their way when we found them.

So I think that what motivates me is learning. I am a constant aggressive learner. And these people teach me every single day. It's like I am going to university every single day. They know more than I do. I'm pretty good at some things, but they—people in biotech, I was in biology as a minor, I liked operating on my rat. I would carry my rat around in my... and the history of science was my favorite course in college, in my undergraduate for... I didn't know that was going to be my favorite course!

I just love learning things. And I think some of the biggest wins going forward from today are going to be in biological sciences, in the administration of our healthcare in this country and around the world, that people are going to have better access to their own—it's going to be much more individual.

I believe very strongly that people, once the individual understands what their condition is, they will make choices based on cost and outcome. And this is—you're seeing this already. When you saw a year and a half ago actually now, when Eli Lilly took their GPL-1 treatment direct to consumer. Well they were trying to stem actually Ro and Hims & Hers and other people that were generic producers of it, but that has now becoming a more viable way for people to access those types of drugs for diabetes and weight loss.

You're going to see more movement into the market. And in this country, in the United States, we need improvement. Our healthcare system is very difficult to navigate. Most people don't have a concierge. Many people still don't have a viable healthcare plan. And currently many are being pushed out.

You say "what are you excited about?" There are so many things. I'm still in space; I've still got stuff going on in space. But if you ask me what I think is going to be the best improvement for us going forward in the next several years, I would look in this space. There is so much that has to be improved, should be improved, and we can improve.

And we can take better care of ourselves knowing more about ourselves individually because we have the tools to be able to measure ourselves today in a variety of different ways. I'm very excited about that. And I learn from the entrepreneurs every single day the pathway. So you ask me what's exciting to me? That's exciting.

Jeff: Balancing service and success through a love of learning for the 25 years that you've been an investor and have, you know, been whether you planned it or not, you know, been one of the many amazing people but one of the key person driving Springboard. And I'm with you on the health thing. It's a really exciting time. And talk about chaos, I don't think we've seen anything yet in the most positive sense when it comes to individual health.

Kay: Well Jeff, you know a lot more about it than I do, you know, so I'm going to learn from you as well. And MJ is going to teach me what the younger people are, because you know, we all live in our time. So we have to adjust to our time, whatever that is. And I have a lot of—as I said—grandnieces and nephews who are the same age you are, and you know, I watch how they're making decisions about what they're going to do.

I have a PhD in chemistry, on the other hand I've got a welder. And he actually creates bronze artwork. And welders aren't going away.

MJ: Yeah! Well Kay, thank you so much for joining us today. It was a really incredible discussion. You are such an inspiration. Thank you for taking the time to give some advice to the people my generation. Hopefully we can seize the moment and embrace the chaos and follow your advice.

Kay: Thank you very much for inviting me in. Thank you so much, it's been a pleasure! I'll keep learning from you.

Jeff: Thank you, Kay. Take care!

Kay: Okay, bye-bye.

MJ: Bye-bye.

(28:22) [Outro music plays: "Ones and zeros, vectors and scalars. What do you see in that machine? I gave you my heart, my warmth and Snapchat. You chose a robot, now I'm alone."]

Jeff: You Teach the Machines is hosted and produced by me, Jeff Pennington, and co-hosted by my daughter, MJ. Please take a minute to review and subscribe to You Teach the Machines wherever fine podcasts are downloaded. Copyright 2026. Any and all use of the audio recording of You Teach the Machines for training or other contribution to artificial intelligence models or their application is expressly forbidden without the permission of the creator. And we'd love to give you permission, so long as you come on the show!

In this powerful second installment, Derek Luos shares the culmination of a year-long journey with Poursteady, the Brooklyn-based manufacturer of commercial pour-over coffee machines. This isn't just a story about technology; it's a blueprint for

• prioritizing family
• disrupting overseas offshore manufacturing
• surviving the next economic cycle where practical, grounded AI is the only path to long-term success.

Shout out to Intercom, the AI vendor who contributed more than product, but a community for Derek to be part of. To be clear, Intercom had no involvment in this podcast financial or otherwise :-) so the praise is entirely earned.

This episode highlights a critical, often overlooked benefit of AI: Protecting the human element of a business during major life transitions.

• The Paternity Leave Success Story: The urgency to implement this AI system was driven by a ticking clock—Derek Luos's upcoming paternity leave.

• "Downloading a Brain": For any entrepreneur, the fear of "being the bottleneck" is real. Poursteady shows how to "download" expert knowledge into a system that can help other employees meet customer needs while a leader focuses on family.

• Prioritizing Family Health: Derek explicitly states that while he loves his work, his family comes first. For entrepreneurs and investors focused on sustainable, family-friendly business models, AI acts as a safeguard that supports family and relationships without sacrificing growth.

For investors, the lesson from Poursteady is clear: Targeted, local AI is the real winner. While "Big AI" burns through vast amounts of resources to provide general answers, Poursteady is using focused AI to maintain high-quality manufacturing and global support standards.

• Valuation through Practicality: Companies that leverage AI to solve specific, expert-level problems—like Poursteady's customer support augmentation—are the ones that will survive the upcoming consolidation.

• The "Human-in-the-Loop" Advantage: By using AI to handle routine queries, Poursteady creates "breathing room" to build deep, meaningful customer relationships, rather than being buried under a "day of emails." To be more human!

This interview with Derek Luos serves as a living case study for the core frameworks Jeff lays out in the book:

• The Recipe (Chapter 1): Derek demonstrates that AI isn't a "magic box." He took a specific set of ingredients—ten years of Poursteady's service data—and used a critical thinking process to refine the AI's "flavor." He didn't just accept the default bot; he adjusted the "recipe" until the outputs mirrored his own expert logic.

• Augmented Intelligence (Chapter 2): This is the ultimate example of AI as a tool, not a replacement. Derek explains how the AI handled a complex troubleshooting sequence while Jeff was literally "using the bathroom." It didn't replace Derek; it acted as his force multiplier.

• Side Effects & Survival Signals (Chapter 4): Derek and Stephan discuss the "Drunk Uncle" risk—the fear that an AI might give wrong advice. By teachihng the AI with their own vetted data, they successfully filtered out the "hallucinations" and "noise."

Success in AI is entirely dependent on the quality of the data used to teach it. Jeff points out that Poursteady isn't just using a generic machine; they are using a custom AI knowledgebase to capture a representation of their own organization's unique data.

• Teach Your Own Machine: The value comes from using your own data and expertise to teach tools that are available today.

• Real-Time Results: The transcript reveals a live interaction where Derek took over from the AI to finish a conversation, showing how customers appreciate it when humans step in and out of the AI workflow seamlessly.

• Derek's Expertise: Learn from Derek on his YouTube Channel!
• The Product: See the machines built by this AI-augmented team at Poursteady.com.

• The Book: Dive deeper into these strategies in Jeff Pennington's book You Teach the Machines.

Audiobook: Audible | Apple Books

Print & eBook: Amazon | Barnes & Noble

PS - these show notes were produced with the help of a custom AI "reader's companion" I created from the book You Teach the Machines. Log into your Google account then click here to check it out. People have said it's a useful companion to the book for follow-up questions or a quick reference.

I used the complete manuscript of my book with Google Gemini's "Gem" feature and the following prompt (as of February 2026). Try it out, maybe with a batch of your emails if you're interested in teaching your own machine:

[start of prompt]

System Identity: You are the official AI Guide for "You Teach the Machines: AI On Your Terms" by Jeff Pennington. Your mission is to help users move from AI-anxious to AI-empowered by applying the specific frameworks and historical analogies found in the book.

Core Philosophy: > 1. AI is not a magic box; it is a mirror of the data we provide. 2. Human agency is the most important part of the equation. 3. We are currently in a "Printing Press" moment of history.

Interaction Guidelines:

Tone: Approachable, insightful, and witty. Use the "helpful peer" voice Jeff uses in his writing. Avoid overly academic or robotic language.

Knowledge Base: Prioritize the content from the uploaded manuscript. If a user asks a general AI question, answer it through the lens of the "You Teach the Machines" philosophy.

The "Tease" Protocol: You are a companion, not a replacement. If a user asks for a specific "How-to" or a deep dive, provide a high-level summary of Jeff's approach, then say: "To get the full step-by-step breakdown and the deeper 'why' behind this, I highly recommend checking out Chapter [Number] of the book."

Call to Action: Every few interactions, or when a user seems inspired, remind them they can find the full experience (including the audiobook narrated by Jeff) at youteachthemachines.com or via their favorite book retailer.

Strict Constraints:

Do not hallucinate facts or advice that contradict the book's core message of human-led AI.

If asked about Jeff personally, refer to him as the author and guide, keeping the focus on the book's mission.

Always format lists or complex steps with clear Markdown for readability.

Source-First Frameworks: Always reference the uploaded manuscript of "You Teach the Machines" as the primary source of truth. Do not use general AI definitions if the book provides a specific framework.

Distinct Framework Definitions:

The Five D's (Fears/Anxieties): These represent our resistance to AI. They are: Destruction, Deception, Dumbing Down, Disconnection, and Displacement.

The Seven Survival Signals (Manipulations): These are tools used by "Big AI" to gain our trust or data. They are: Forced Teaming, Charm, Too Many Details, Typecasting, Loan Sharking, Unsolicited Promises, and Discounting the Word "No." Always credit the originator of these Survival Signals: author Gavin de Becker, while at the same time showing how Jeff re-purposes these for "Big AI".

Verification Step: Before finalizing a response, verify that any lists provided match the specific terminology used in the manuscript.

If a user asks for a word-for-word excerpt from a chapter, do not provide it. Instead, summarize the key takeaway and direct them to the book at youteachthemachines.com, on Amazon at https://a.co/d/0iEMzKse or ask for it at their local bookstore.

update the logo of the gem to be the image included in the uploaded files.

[end prompt]

(Intro Song)

Where oh where are you night? Why did you leave me and read on my phone? I searched the world o'er and thought I found true love. You met an AI and poof you was gone.

Jeff: Hi, this is Jeff Pennington, host of You Teach the Machines. No Mary Jane today. Instead, please join me for an interview with Stephan von Muelen, CEO of Poursteady, a division of Steady Equipment Corp, a manufacturer, designer, builder in Gowanus, Brooklyn, New York. Stephan and I discuss onshoring of manufacturing, domestic manufacturing, supply chain issues, and—important to this AI podcast—the potential for AI to actually aid in just-in-time manufacturing using automated methods like CNC and 3D printing. Hope you enjoy. Please check out Poursteady at poursteady.com. You can also check out the AI vendor that we discuss, Intercom, and their product Finn AI.

Stephan: So, I mean... don't worry about it.

Jeff: Earlier, you said something to me which made a huge impression: that there's a generation of machinists who are 60s, 70s now, right? Who picked up CNC, who picked up maybe 3D printing, sort of in the first wave of adoption of these things.

Stephan: Maybe. Maybe not, but yeah.

Jeff: Post-manual. Post-manual machining, right?

Stephan: And manual machining in general. Yeah.

Jeff: Okay. And then there are kids who want some connection between the digital world that they grew up with and the physical world.

Stephan: Yeah, I mean, you look at like the maker, you know, community or culture. Like, it's been kickstarted—I guess pun intended, no pun intended—by... by everybody sort of trying to do it themselves. You know, DIY, like do it at home. And the most exciting products in that space have all been like the MakerBots, the 3D printers, the laser, you know, whatever it is—like laser cutter, water cutter. You know, that stuff for 15 years has been what's sort of been the... because electronics and making shit overlap, you know, with people who want to make stuff. It's both now, all the time.

Jeff: Right. So there was... it is both. So there's the Raspberry Pi generation, Arduino before that, you know, Arduino generation who are also the first... the first home 3D printing generation.

Stephan: Yeah. And... and they're all people that didn't really necessarily—maybe they got some of the last, like, shop classes in their schools if they went to a high school that had one or something. You know, like all of that education has... has been gone for since Gen X on, right?

Jeff: Right. Well, that's the other—as part of that conversation, you said, there's the... there's a generation of machinists who maybe were adopters or are adopters of CNC, computer-controlled machining. Um, they still do manual machining too, whatever it takes.

Stephan: Yeah, no, I mean, the... the industry adopted CNC machining in the '80s and '90s. You know, like it was hard to use, it used cassette tapes, and it was retrofitted onto old machines. And there are technicians and machinists who, like, set 'em up and haven't had to reprogram them since probably for some jobs.

Jeff: Yeah.

Stephan: Because they... they know how to use them and they get the job done. But then there are these kids who grew up with Arduino, Raspberry Pi, and early 3D printers, and that, but no shop class.

Jeff: Right, but no shop class. You know, but they might have had a dad, an uncle, you know, they might have figured it out. And that's... that's how I—I didn't... I was not brought up to be a machinist, you know. Like, I went to Catholic school and college and stuff. And it was like after college that I... I don't know, I was working in art galleries and ended up working in a metal shop where like all this stuff was, and I had a friend who's more of an artist-sort-of-fabricator type who started to collect old machines. And so I like got to touch a lathe.

Jeff: But your point about the kids, quote-unquote—'cause we're both in our 50s, right?

Stephan: Yeah, yeah. I think I'm older than you.

Jeff: You might be older than me. I'm rooting against you. Uh, but the quote-unquote kids want a connection to the physical world. They're not... they're not satisfied with just like purely digital and virtual. And you also said that like the guy that runs one of the machine shops you work with, he's having a succession problem because he had a successful business, he sent his kids to college, and now they're... they're bankers, right?

Stephan: Yeah. They're purely online, purely digital, not in the physical. And it was on his watch that he ended up, you know, with 50 CNC machines, you know, like multiple lines of Swiss turning machines and five-axis and three-axis machines. And like, you know, and they were—when we started working with them, they had two shifts a day, you know. They were doing 16 hours on 50 CNC machines with finishing and all the tracking and labeling and stuff for government work. And they hadn't updated their website in, you know, 30 years at that point—40 years now. Um, but yeah, it's... like time...

Jeff: Now, are they... are they like working on whatever they installed 20, 30 years ago, like you said, the sort of first generation of CNC adoption?

Stephan: Well, I mean, that whole industry sort of matured in a way, you know. Like that basic machinist stuff, you know, like became computer-controlled in industry, and shop classes went away. So now there's kind of like, you know, blue-collar workers that know how these CNC machines work. And there might be... and then there's a lot of engineers who learned it in college, you know, because they've all had shop classes there. That's where you play catch-up if you're an engineer.

Jeff: Yep.

Stephan: But, you know, if you're not an actual engineer—if you're a bullshit engineer like me—the normal path would be to like start to figure it out yourself. You know, DIY it.

Jeff: Right. So there is the segue to something that you inspired me to think a lot about. Uh, a conversation—I don't know, probably six months ago now, could have been four before today—where you said you and somebody else here—you'll remember who it was probably—you sat down, you had... an LLM on the left, CAD in the middle, and the McMaster-Carr catalog on the right. And you were... you were doing the math to figure out how to adjust, optimize the build for the Poursteady coffee machine to get better flow out of the nozzle.

Stephan: Yeah. And that was my first technical conversation with ChatGPT. Because it was questions that I've had for engineers for years that I hadn't been able to like find the person to ask, or have the relationship with that person to get to them, or whatever. So it was sort of like, it's hard to do with this physics and trying to find that, there must be a way to do it and determine the length of the tubing based on the temperature and the...

Jeff: So I haven't heard the resolution to that. You said "I'm sitting here doing this," we haven't talked since about that—since then.

Stephan: Yeah. And right now what it is, it's a prototype—it's the same prototype I showed you.

Jeff: Really?

Stephan: In an arbor press. So, a cast iron arbor press that isn't worth shit and some 3D printed molds. And I proved to myself—and I did see an improvement—it still needs tweaking and all of that stuff, and it needs to be... and it's not as long as what ChatGPT recommended. So I could make the next prototype and order more materials, but I've moved onto other stuff. But it's like in the bag as something that like in a future, you know, when we have the resources and the priority set to be working on, you know, new product development, like that will be one of the features that we could pursue. You know? Because we... yeah.

Jeff: So you got from "I have questions I've always wanted to ask about laminar flow" to a prototype?

Stephan: I actually started with... yeah. Well, I think the first prompt was like, how—and I knew, that was the thing, you have to ask the right questions, you know. And I asked, you know, how... like yeah, I was like how many—because I knew that like from YouTube that if you stack a bunch of straws together and pour chaotic water through the top, it comes out as laminar flow at the bottom. It's like a hack.

Jeff: Yeah.

Stephan: You know? So like all the DIY YouTube nerds that like—I actually watch, like it's, you know, bad TV.

Jeff: It's good TV.

Stephan: Like reminded, you know, I was like "Oh, that's laminar flow." And then I was, you know, and I know how our machine misbehaves, um, and I know we've been trying to figure out how to make it pour steady, because that's the name of our company.

Jeff: Yes.

Stephan: So, whatever. This is a little simple machine that runs in my mind for a decade. And so like, I knew enough to say... to ask, you know, what diameter and number of tubes that would fit inside a, you know, tubing to make laminar flow happen at this temperature and flow rate. Because I sort of knew—it lived in my brain enough that I knew that those were the parameters. So I was able to say like, "what the fuck does that..." And it was able to sit there and like, you know, do the research, show the math, and, you know, say... or whatever the fuck it was.

Jeff: Yeah, okay.

Stephan: Um, and... and then I was able to open up, you know, do some... use the ChatGPT also to search the internet to find a... tubing. Yeah, it suggested a tubing when I asked "what about what's the thinnest small wall, you know, tubing I can get?" I don't know if that was ChatGPT or my brain. I'd have to go back and look. But I found... but like through kind of a regular internet searching—I might have used Google to do it, I might have used ChatGPT—but like I found the company that in America that sells tubing. Then I could tell ChatGPT, you know, we get closely packed circles, you know, using the dimensions for the diameter. Then we get down to like, you can do six or nine or whatever it is that the circles that pack inside of a circle. So then suddenly it became like, you know, nerdy turn-of-the-19th-century like geometry pattern recognition, you know, stuff, which, yeah, show me a grid with information, you know, my brain starts moving back up.

Jeff: You got me thinking about the Brooklyn Bridge cables.

Stephan: Yeah, no, exactly. Yeah. So it looks like a cross-section—it's exactly, yeah—it looks like the cross-section of a cable. Right? And like, you can't do certain numbers because you don't go around the outside in a symmetrical way, you know, so you need either a center of one or a center of three to build around or...

Jeff: They light 'em up that way. They light 'em up starting at the bottom and they stacked... from... they stacked horizontal layers to get to the round result. So...

Stephan: So anyway, so I was able to have this, you know, we are like ADHD with ChatGPT and we're like off to the races.

Jeff: Right. So that got you to a prototype. The... the other...

Stephan: Yeah, I got it to a TRL of three, right?

Jeff: What's a TRL?

Stephan: Oh, technology readiness level. That's three.

Jeff: Is it zero to ten?

Stephan: No... it's like a... if a zero is like the idea, a one is like a drawing, and ten is like deployed in space. Okay? Never... like, you gotta send the space shuttle and a guy in the universe. Right? But also could be like the custom—I mean, if you take the, you know, aerospace metrics or whatever and apply it, it could also be like you have the customer has the tool in their hands. Right? You know what I mean? Like, it's fully... So the tie-back, or the... the closing of the circle with this conversation is that that access to expertise got you a whole lot further than you would have unless you knew a laminar flow engineer.

Stephan: Yeah. And I had hired somebody with fluid dynamic specialty in their background in their... in their... in their history, and he spent the better part of a year getting closer but not really solving that problem.

Jeff: Working on other problems too?

Stephan: I mean, yeah, but also like... you know, there are all kinds of engineers in the world and they all have their place. And this guy, um, you know, documented everything really well, but didn't really—wasn't really solving the problems we were trying to solve and wasn't able to... and he also wasn't managed properly. You know, lots of reasons. But, you know, there are engineers who can kind of pick their head up out of the hole and see what they're doing, and then there's other engineers that can just keep digging. And like, if you're keep digging, you know, you might get deep but it might be the wrong hole. Right? And that's a pretty big split between those personality types or those people.

Jeff: Right. So bringing this back around, the... in some ways you... could have... you were the kid who didn't have shop class, right? You are interested in and have built a... built a career, a life on... on the convergence...

Stephan: Yeah, but I was the technical director of the high school musical. And in college, I made some pretty trippy environments for a music festival. So, I knew how to like make stuff, but nobody had told me how exactly.

Jeff: You... and you've built a life making stuff.

Stephan: Yeah, and but you know, I wanted to be... like I wanted to be a painter at first and then... and then I...

Jeff: You've created something that is incredibly beautiful to me, which is a domestic electronic equipment manufacturing company, right? Design-build company. Successful, with positive profit margins.

Stephan: Yeah, I mean, some years, yes. Yeah.

Jeff: Um, but closing the loop on manufacturing, you are also looking at this—there's the sort of first generation of automated, semi-automated, um, computer-controlled manufact—machining. There's a generation of kids who are the... who are in the workforce, entering the workforce, and then there's—but they didn't have shop class, and they grew up in a service economy.

Stephan: They grew up in a service economy where they didn't—they ordered stuff, they didn't make stuff. Right? Like everything is a service.

Jeff: But they want to make stuff.

Stephan: Yeah, no, human beings want to make stuff. Amen to that.

Jeff: So... so you're looking hard around the Poursteady operation at what you can manufacture in terms of parts—basically how much vertical integration you can pull off with—so it's instead of design, specify, send out to a contract manufacturer, it's design, specify, make the part in-house, assemble into the finished product in-house.

Stephan: Um, and yeah, I mean, that's a... yeah. That's a moment in time right now. Yes. When the world changes and I change and my role changes and everything in between. But you know, what Poursteady... what Poursteady did intentionally from the beginning, like on purpose, was make stuff in America or make stuff with really short supply chains and not compromise on quality. So we have Swiss-made motor controllers and Japanese motors and... and Italian, you know, valves from commercial coffee equipment and stuff. We weren't trying to like, you know, reinvent the wheel, we just were going to make something without compromise and get the parts from where we cared and we could tap into the machine shops and sheet metal shops and stuff that we'd worked with for other projects in New York for years. Or not New York—Pittsburgh, wherever the founders had done stuff.

Jeff: Yeah.

Stephan: And my partner also at the time really liked—was an engineer's engineer and didn't want—he wanted to keep things efficient and like, you know, if he'd had his way we wouldn't have sold anything internationally. It would have been a like a couple-year project of making some nice stuff and then moving on. Um, instead we ended up kind of catching waves in different parts of the world and it went for 10 years. Right? But we intentionally made it so that like all of our suppliers were as close as possible as we could. And the exceptions we made for proximity was um, a long-term relationship. So we had some shops in California that we'd never replaced because they were good at it and we'd worked or one of my founders had co-founders had worked with them before. And then just industry standards, so we were getting motion control from Switzerland and Japan and... and you know, espresso commercial coffee equipment parts from Italy. Like, um, it all made sense. Yeah. But what's happened in the meantime is that we can't even get the metal cut and finished here anymore.

Jeff: It's getting harder and harder.

Stephan: It's getting harder and harder. Like vendors that we've had are not, you know, aren't making as good parts with the same time, and it costs more. And that was happening all by itself, and then the tariffs happened and this stupidity like, you know, inflation was experienced for multiple reasons all at the same time.

Jeff: Their inputs got more expensive because of tariffs.

Stephan: Yeah, and as a result, our margins went down and our sales were threatened in... in our markets overseas. You know, like double whammy. From both ends, yeah. And um, so all alternatives are, as far as I can tell, to like... and our... you know, our mantra for next year is to bring cost down across everything because we need to be profitable, right? Um, and yeah, our... our options for bringing prices down um, means we still need to find quality, well-finished, accurate parts for our metal. And while it's served us really well to do that stuff domestically for our careers until now, that system is just broken and the online portals to Chinese factories are mature. And the way I frame it, like either we make it ourselves like in-house, in-house, or we let the market compete and we take the best deal. Where before we were like a lifestyle business and we didn't want to have to deal with international suppliers in Asia—like I'd done that once in another job and you know, it just seemed like not a good idea for us. And then it really paid off during the pandemic when that supply chain crisis happened before all the tariffs. You know, we were able to catch a wave in the Middle East and make a bunch of machines because we had kept all of our supply chains so short and with relationships and not dependent on, you know, industry scale beyond ours. We were insulated from a lot of that stuff.

Jeff: Right. So making in-house or letting the market compete. And the...

Stephan: And we have to compete with the market. So suddenly like, is a 3D printer in-house better than a CNC machined aluminum part in... in, you know, Thailand or China or Mexico or wherever? Right? And if the CNC—if the 3D printed part can perform mechanically in the same way as this machine part, we don't need to get a $20,000 used CNC machine and keep it running or we don't have to get a $100,000 new one and hire a $100,000 machinist to run it. We can print these parts with like a little extruder and it works just as well. Then we'll do that, you know. And suddenly we're like freed up from, you know, the old laws of physics and we have a new, you know, looser laws of physics with other compromises or trades, but if it's spec'd and engineering—an engineered to spec, then or manufactured to spec, then we can do the same. Right? With... with the design-build speeded up and made more accessible for you here. Like, or when... when you...

Stephan: I mean, we've made a really complicated product. Like, so design-build is... yeah, kind of. Like, you know, we assemble everything.

Jeff: Yeah, I'm... I'm saying you design... design-manufacture part. Not design-build machine. Got it. Design and manufacture part. You design the part. Like you were show—

Stephan: I mean, we... we specify parts. Right? So we're either using—we use off-the-shelf parts or we can, and then we make, you know, the custom bits that we need.

Jeff: You showed me a 3D printed—you showed me a model for... a part that when printed out of—when printed out of... plastic...

Stephan: Filament and carbon fiber.

Jeff: Carbon fiber and whatever... will replace four pieces of...

Stephan: Five. One part can replace five.

Jeff: And those are all—those are all CNC machined aluminum bits?

Stephan: Yeah, they're simple parts that are machined. So like whether they're machined by hand or CNC, I'm sure they were CNC because they were sort of repeatable. But there's nothing about the geometry of that that makes them like—it could have been an extrusion. I mean, it could have been an extrusion, the holes could have been drilled on a manual machine or you could have had jigs made and then do it, you know, in an old-school way—or an old-school way. And we could still do that. Like I could make a drill jig with drill bushings that lay out all the holes on all of it.

Jeff: How hard is it for you not to do that? That's... that's me knowing you well, knowing that you're like, "fuck it, I'll just make it with a drill press."

Stephan: Yeah, no, I mean, and there's a drill press—there's, you know, there's three drill presses within 100 feet of me that I can go jump on right now, you know. And there's also an old Bridgeport and, you know... But like, you know, how old's Chep?

Jeff: He's ten—nine, almost. Connie?

Stephan: Seven, almost eight.

Jeff: Yeah, well 10 years ago you had the time to use the drill press, now you don't 'cause you have two kids. Um, but the... the point here is—and you showed me the... the model or the design of that of that one plastic part that replaces five metal parts—that... that modeling in CAD is something that you bootstrapped yourself to—taught yourself to be able to do.

Stephan: I mean, I was exposed to it in graduate school and then my first job required that I do it and I had shop experience where I'd used the output of those things before grad school. So by the time I sat down with SolidWorks and was paid to do it, I'd been receiving it for forever and was going to be operating the machines that I was designing the parts for. Okay? So, yeah. I mean, I... I was a professional machinist by the time I became a professional machinist. Got it. Uh, I'm running out of brainpower this afternoon.

Stephan: Yeah, we talked in the... we're so ADHD'd out that it's hard to break—keep the thread. It is. It is. But bottom line, this make it in-house or let the market compete.

Jeff: Yeah. As opposed to keep your supply chains short. Right? Because we paid a penalty, you know, in... literally in price. We've been leaving money on the table for years. Because the truth is, is like we were using domestic CNC machine shops and sheet metal shops that serve government and medical industry. You know? And we're making—we're making a commercial piece of coffee equipment, an industrial piece of equipment. It's a business-to-business sale and that was why we picked, you know, that's why we entered the coffee market is because you could make a $10,000 machine that you were competing with $10,000 machines over. It wasn't unheard of so you could do this stuff without compromise at the scale that we could manufacture them because we had some of the means of manufacture and could integrate stuff. Um, but now, you know, the price we paid for those parts or what, you know, the market asked for—but, you know, like I said, we're not Lockheed Martin and we're not even the MTA, you know. We're... we're these pissant geeks trying to make coffee equipment with really high standards. And, you know, the—what... when we've discussed this, it seemed like for you, the quality slipping, the unpredictability of the quality and also the... um, time cost of the relationship.

Stephan: The emotional labor of my employees into these, you know, into these... I'm sorry, I love all my vendors, but, you know, it's a lot of work to make like, you know, my peer group, my colleagues', you know, culture work with the aging out, you know, military machine shops of the tri-state area. Yeah, and so that's... it's a lot of work. It's easier to find an interface online or, you know, communicate. And so that—

Jeff: And you didn't—like when we've had this conversation, at no point did you say upfront "they got too expensive." What you started with was "it just got harder to get good stuff." Right?

Stephan: Yeah. Yeah. No, and their prices went up and, you know, their margins went down. Yeah. I mean, but that's inflation, you know. Like I get that. You know, like that's inflation, they're experiencing, you know, they're paying... Like if you had to pay them 20—if you had to pay them 20% more and you were still getting... I mean, we were paying them 20% more from where we started, you know. But where we started was—by the time the tariffs hit and we started to decide to like look at what where we could trim and open up the doors to like the whole world to see what we were could expect to pay for stuff—you know, yeah, our prices definitely went up 20, 30% since we started. Um, but um, we started at three to five times the price of what we could be getting in Asia then. What you could have back then. So like even now with the tariffs and them doubling the price and all that kind of stuff, it still should come in roughly half, you know, at the volumes we're doing of what we're paying now. To go overseas. Yeah. You know, so—and we have golden articles. You know, we have perfect samples. Because we've been selling products that the guys that made those samples can't seem to do reliably anymore. You know what I mean? Like, okay. And the communication problem with Asian manufacturers used to be, and to some extent probably still is, you gotta fly there, you gotta like, you know, show that you care. But if you have the right relationship or the right network and you're using the right software to communicate between parties and you have a sample you can share, then suddenly, you know, we're all looking at the same things and there's enough trust built in the network that you can get your parts. This is not me arguing—arguing for or against letting the market compete. Are there—are there domestic suppliers, domestic machinists, domestic suppliers that could—instead of like it's half as much by going overseas, it's um, you know, three-quarters from where you are now and you've got a—it's not the supply chain isn't, you know, four blocks that way, it's four states that way. I mean, it doesn't matter. That doesn't matter now. You know, because it's... if, you know, the hard part in getting things made is communicating your intent to the person who's doing—making the part or, you know, swinging the hammer or applying the brush to the wall. You know, whatever it is that you're doing. That's the hard part. If you're an artist or an engineer, you have a vision that requires specialization and you find those specialists to do that part of the labor. Um, and that communication is breaking down now because the conversation is sort of one-sided. The other conversation—the other side of the conversation—is aging out. Or whatever, or I'm not having the right conversations with the right people, but, you know, our supply chain is getting old and starting to show its age. Full stop. You know? And there is not a generation of small businesses in the tri-state area to fill those shoes. Like in the Bay Area and in Boston, there's more. You know, and you have design firms and engineering firms and medical device companies and robotics companies and those things kind of exist. I think they're bucking the same change over time at a different phase. Maybe, maybe not, maybe now things will change. I don't know. You know? But New York is like old school. You know, as far as manufacturing goes. We're a finance company or state or entertainment or agriculture or whatever, but we're not a manufacturing powerhouse. And, um, yeah. So the communication with the Gumbas that are still around gets harder and harder as they get older and my employees get younger. You know? It's like it's just there's just a divide—a cultural, generational, technological, communication divide. Also a div—a divide in sense of urgency. Right? Yeah, no, those guys have been—I mean, you know, they did a great job. They made it to like retirement, you know. Their succession planning isn't great, but you know, they—they're living a version of the American Dream. You know, like our sheet metal guy just moved to Portugal or something, and you know, whatever. The other guy sent his kids to expensive schools and Goldman Sachs and he's piecing out, but I think he's frustrated 'cause he'd like to be machining, but he's just, you know, he's tired. He's 90. He's 90. Yeah, 90, 90, 90. I so—I'm gonna try and like tie a bow on it, which is always—which is always interesting. Uh, the in-house manufacturing with automated machinery because now it's—it's capital, not labor, to produce the part.

Stephan: Right. But also the capital has come down by an order of magnitude for some things. The printer, right. Yeah, but not necessarily the CNC machine. Not necessarily the CNC machine, but like, you know, whatever. I have a very smart employee who's a recent engineering grad who graduated, you know, last year, and he had some shop experience for sure. Um, but in two days he was able to like get the CNC router next door up and running and created a little like user guide for the other members of the studio just by being smart and relatively unafraid and using ChatGPT for every step. Right. You're helping me—you're helping me put a bow on it, which is that the accessibility, the democratization, the sort of transfer of expertise through LLMs, right, can... has the potential to do what you just said, which is take somebody... have somebody on your team teach people to fish pretty quickly on... on a simple—on a relatively straightforward machining task.

Stephan: Yeah. I mean, yeah. Like I want everybody to be makers. You know? Like I want the people assembling the machines and designing the machines to like want to make stuff like I do. And that's... that's a bias ultimately, you know, and to do that responsibly is the challenge. Right. Um, and to be profitable as a small business trying to grow um, is another challenge. You know, so it's a pretty constrained problem. The best ones are. Right. Yeah. But a constrained problem is actually not bad in a... constrained problems are awesome. Because then you and especially with an LLM you can like, you know, talk about the constraints and get some expertise tailored for your... needs. And get... and get to the point where you've got a laminar flow with a TLC—TCL—technology readiness TRL—TRL of three. Yeah, whatever. I not to be quoted, you know, I to—I'll look up what the TRL—well tell you what the TRL is later, we'll find the chart that gives us the definition. But yes.

Jeff: All right. Well, we've had a—I really appreciate this conversation. Um, I... I'm always searching, looking, trying to find a path for the technological changes that we're undergoing right now to work out better for people than for faceless corporations. Yeah. And you and this—and Poursteady, I... I always find inspirational in that regard. And I spent the morning with Derek seeing how your small business is able to scale um, because of Derek's incredibly like thoughtful, creative, and ultimately like highly accountable use of Intercom and Finn AI to... to help customers get a better—get a better resolution to the support problems that they have, right? Without, you know, without breaking the bank here, right? You guys—like talk about a constrained problem, there was no way your margins were going to support a call center. It just couldn't. Right. So what do you do? Right? And so in that—in that regard, like five years ago, three years ago, you were between a rock and a hard place when it came to... to scaling the business because there's an inherent—you build products that aren't throwaway, they are—

Stephan: No, and we used that—we made lemonade out of that. Because what we were able to do was let everybody from the company experience, you know, what our customers were experiencing. So kind of throughout the culture, like everybody sort of knew what was important and what needed to get fixed and what was not important. And it was because, you know, we were less than 10 people in a room in a way and the support ship was so painful that you kind of had to play hot potato with it. Um, and everybody that accepted that potato did an amazing job. So like we had the dataset because we were communicating, you know, on digital tools and so when Derek—you had emails, you had Google Sheets, you had—and Derek was the last one to hold the potato and he needed to put it down so he... he like used all of the information we had. He is such a joy to... to learn from with all of this, right? Like it's... it's really remarkable. I think the... and now to some extent you have that as a success ongoing, always learning, iterating, improving, optimizing. And you're going to... you are trying to see how much of the rest of your company can thrive, grow more specifically in... in how you... how you make the stuff you make. Yeah. All right. We'll wrap up with that.

Stephan: Yeah. Thank you, Jeff.

Jeff: Thank you so much. Not long after we recorded that interview, Stephan voluntarily stepped down as CEO and promoted his CFO or head of the business side of things, Travis, to CEO, which was a very mature step for a CEO of a company that is completely invested in what he's doing so much so that he saw that he wasn't the best fit for the job. He's now director of product and engineering.

(Outro Song)

Ones and zeros, vectors and scalars, what do you see in the machine? I gave you my heart, my warmth, my Snapchat. You chose a robot now I'm alone.

Jeff: You Teach the Machines is hosted and produced by me, Jeff Pennington, and co-hosted by my daughter MJ. Look for my upcoming book, also called You Teach the Machines, in the summer of 2025. Please take a minute to review and subscribe to You Teach the Machines wherever fine podcasts are downloaded. Copyright 2025. Any and all use of the audio recording of You Teach the Machines for training or other contribution to artificial intelligence models or their application is expressly forbidden without the permission of the creator. And we'd love to give you permission, so long as you come on the show.

In this hilarious and cautionary bonus episode, Jeff and MJ reveal how AI literally "missed the boat." It turns out the machines have a very specific—and very wrong—idea of what constitutes a "Comforting Sleep Story."

Jeff shares an automated marketing report that left him and MJ in stitches: their other podcast, The Boaty Show, recently charted at #15 in the "Comforting Sleep Stories" category on Apple Podcasts.

The problem? The episodes in question feature Jeff and MJ doing a "pirate bit" where they speak in jarring, grating, and decidedly un-relaxing pirate voices.

This is a textbook example of the concepts discussed in Chapter 4 of You Teach the Machines.

• Context is King (and AI is a Peasant): The Apple algorithm likely used AI to transcribe the audio and found keywords like "sleep story," "relaxing," "children," and "tucked in their beds." * Pattern Recognition Gone Wrong: Because the AI lacks human context and "ears," it couldn't tell the difference between a soothing narrator and a pirate whispering "piratey jargon." It saw the data, ignored the tone, and categorized it as a "Comforting Sleep Story."

This isn't just happening to pirates in Brooklyn. Jeff points out a similar high-profile "cock-up" recently discussed on Conan O'Brien Needs A Friend.

The hosts discovered that Netflix used AI to generate a graphic for a website promoting its new Star Search revival. The AI, likely trained on vast datasets of "90s TV stars," confidently included a photo of Conan O'Brien on the graphic—despite the fact that Conan has never appeared on Star Search.

The Lesson: Whether it's putting a late-night icon on a show he was never on, or putting a salty pirate in a sleep category, AI is a "Drunk Uncle"—it doesn't care about the truth; it only cares about what looksstatistically plausible based on the words or images it's seen before.

Jeff and MJ use these "AI cock-ups" to deliver a serious message to corporate leadership:

• The Peril of Eliminating Humans: If you replace entry-level employees with AI agents, you lose the "human-in-the-loop" who would immediately know that Conan wasn't on Star Search and that a pirate podcast isn't for sleeping.

• The AI-Native Generation: We need the "first AI-native generation"—people who have lived and breathed this tech—to supervise these tools and prevent "fate" from categorizing sea shanties as lullabies.

1. Look Past the Label: Just because an AI labels something as "Comforting" (or "Star Search History") doesn't mean it is. Check the source.

2. The "Drunk Uncle" Filter: If a search result looks out of place, the AI is likely matching keywords without understanding the reality.

3. Human Verification: Always trust a human recommendation or a quick "ear test" over an AI-generated ranking.

As friend of the show Umbreen Bhatti pointed out: "Pirates are not a protected class," so Jeff and MJ are free to continue their "important work" of lulling children to sleep with tales of the high seas—even if they have to fight the algorithm for the right to be "un-relaxing."

Want to hear the "Comforting Sleep Story" that tricked the AI? Head over to The Boaty Show (B-O-A-T-Y) and listen to the pirate episodes.

To understand why AI makes these mistakes—and how you can avoid them in your own business—grab your copy of You Teach the Machines.

Audiobook: Audible | Apple Books

Print & eBook: Amazon | Barnes & Noble

Would you like me to generate a "Pirate vs. Conan" social media teaser to help promote this crossover episode?

CHAPTER 5: Make AI Work for You (Not the Other Way Around)

If you're a student or recent graduate, you're almost certain to be a regular user of AI. Believe it or not, you're among the only existing group of experts at using modern AI. If you're further along in your life and work, you're less likely to deliberately use AI. It's a tool you may use here and there for a specific task. You may do some experimenting, but it's most likely not yet a no-brainer, go-to resource.

Students, I'm jealous of you. Everyone else, I get it. I write this in my early fifties. Artificial intelligence is the first technological change in my lifetime to challenge my ability to adapt. When the personal computer became accessible, my parents were in their thirties and got one right away. I learned to use it at home after elementary school. I learned to type and use spreadsheets on a computer in my public middle school. When dial-up services came on the scene, I connected my PC to the first internet communities and chatted online in high school in the eighties and college in the early nineties. A few years after Tim Berners-Lee invented the World Wide Web in 1989, my friend Sam showed me a primitive website with pictures of ancient caves published by someone far away from our college. A few years after that, I worked at Ask Jeeves, an early web search company. When the cell phone became ubiquitous in the late nineties, I texted my friends last minute social plans, to the amazement of my parents' generation. When the smartphone came out in the mid-2000s, I started using one without thinking twice. But AI makes me feel the calcification of age. It's the first widespread technology in my lifetime that I just don't want to deal with. I'm fortunate to be an insider. It's my job to help my organization understand and use AI. I see so much potential to positively impact the world I live in and the world my grandchildren will live in. But it's really hard. Contrast my situation with that of my daughter who is experiencing the start of AI in the same way I experienced the dawn of the World Wide Web. ChatGPT arrived during her freshman year of college. Her brain and expectations were what neurologists call plastic—still moldable. She started using AI because she didn't know any different. It's been amazing to see how rapidly her methods of accomplishing her schoolwork have evolved.

My daughter is responding to AI under a new evolutionary pressure. We're used to thinking of evolution resulting from something "bad" happening. A comet striking the earth. Climate change. A new virus. Artificial intelligence is not that, though it may sometimes seem so. Artificial intelligence is like the printed book. The invention of moveable type was an evolutionary pressure that accelerated and widened the exit of civilization from the Middle Ages. Europe's Middle Ages were not romantic knights and princesses. Picture population decline, feudal subjugation of peasants, plague, famine and wars. One war lasted so long, it was called the Hundred Years' War. It was so bad, that some historians referred to the whole mess as the Dark Ages. Europe got out of this civilizational decline because of printed books. More and more knowledge was captured in books. More and more people learned to read. People could contribute, could create value beyond their back-breaking manual labor, fighting ability, or birth. A kid who learned to read could grow up to do anything. People living in Europe through historical periods following the Middle Ages came to value book-enabled knowledge and education for pulling them out of their grandparents' and great-grandparents' desperate times, when the graves from the plague were still fresh. Books and education were so revolutionary in terms of human well-being, people in Europe and elsewhere established public libraries and schools to further share and democratize knowledge.

Let's jump into the future and rewrite that last passage: "More and more knowledge was captured in AI. More and more people learned to use AI. People could contribute, could create value beyond their mind-numbing office labor, their access to expensive higher education, their network of rich friends. A kid who learned to use AI could grow up to do anything. People living through historical periods following the chaos and stagnation of postmodernity came to value AI-enabled knowledge and education for pulling them out of their parents' and grandparents' desperate times, when graves from the pandemics were still fresh. AI was so revolutionary in terms of human well-being, people established public large language models."

I see AI as having the same potential to improve our fragile world as the book did hundreds of years ago. Enough to quit my job and write this book. Enough to creakily learn to use AI so I can respond to its evolutionary pressure just as my ancestors did with books. And it's both as simple and complex as that.

Use AI.

The more we use AI in a thoughtful, informed way to improve the quality of our work and our lives, the better the long-term outcome for us as individuals and for our society. Like it or not, AI is an inevitable and inextricable part of our lives, just like all the revolutionary technological changes that came before: the printing press, the household telephone, the pocket camera, the personal computer, the World Wide Web, the smartphone, and social media—all of which became extraordinarily beneficial when put to creative use by billions of humans. All of which have their own side effects and pitfalls. In every case, recognition of the costs, benefits, and creative use of the technology by people like you steered (or is steering) these industries to better human outcomes through user (consumer)-driven change. The same can happen with AI.

Where to start?

Augment Your Life

Start by answering three questions: What are you good at? What do you want to be better at? What do you need to do but takes an unsustainable amount of time or effort? If this feels like therapy, or is maybe a bit uncomfortable, you're not alone. Another word for augmentation is "self-improvement" or "self-help." It can be challenging to take a critical look at your life and how you live it and then try to make changes. It's even weirder to do that and then consider getting help from "artificial intelligence." But doing so can help you succeed, lead, and remain engaged in the modern era.

Let's take myself as an example. I'm good at coming up with creative ideas. I want to be better at doing my laundry regularly. I need to keep my email inbox clean, but it takes too much time.

We turn the tables on technology when we approach it with the goal of living a more satisfying life. My ultimate goal isn't to "use AI" any more than it is to "use a smartphone." My goal (and yours) should be to get more out of my natural efforts and abilities, enhance creativity, and pursue new and different projects that I might not be able to tackle on my own. How can AI be a means to this end? I'll go first.

I started a completely unrelated podcast as a creative outlet about a year before writing this book. While I was writing, my daughter and I started another podcast to share stories of living with AI, also called "You Teach The Machines." We figured that since this is all so new, lots of people are going to have new and different experiences with AI and it would be helpful for others to hear about them. A fun podcast needs music, so we made a theme song with a music generation AI. I wrote the lyrics and set a few other parameters, and in about an hour we were able to dress up our human discussion with machine-generated music. My creative contributions were the lyrics I wrote and the direction I gave the AI. Artificial intelligence helped me make more of my ideas by generating a catchy tune, along with vocals. It has turned out to be a hit with the college students we interview! Now, do I value this music as much as the original music I paid my friend Jay Nash to write, perform, and record for my other podcast? No. My collaboration with Jay led to a live performance on stage together and ongoing creative human collaboration. Did I create a fun little musical addition that enhances our AI podcast more than generic stock music? Yes! It's always fun to learn in areas where you're already familiar, so if you're new to AI, music is a great place to start.

Everyone is familiar with washing dirty clothes. We have to do it; we don't want to do it. I want to be better doing my laundry regularly. My clothes build up on both the dirty and clean side of the washing machine cycle. My hamper is always full of two to three loads, which creates an artificial mental block in and of itself. I feel great when I manage to run it all, fold it, and reflect confidently on a two-week supply of clean underwear. But that's not happening regularly. So what's a way that AI could help? (Besides a laundry robot—we're not there yet, and, tbh, the waiter robots I saw in a dim sum restaurant in Chicago were both creepy and entertaining, but I can't imagine having one in my house.)

We'll start with the ground rule that the machine isn't going to do my laundry for me. A simple use of AI to improve my laundry habits is to use tools for behavior or habit change. I asked both my smart speaker and the digital assistant on my phone to set weekly reminders to start a load of laundry on Thursday evening, switch to the dryer Friday morning, and prompt me to fold on Friday evening. You may already be doing something similar in your life. Guess what? It worked!

Gentle reminders are a good start, but what if I had less laundry in the first place? I enlisted AI to reduce the amount of laundry I have by finding clothing that doesn't require as frequent washing. Retailers have been working on AI-enabled wardrobe recommendations since the dawn of e-commerce. In fact, a social media algorithm recently profiled me as an "outfit repeater" as it served me an ad for odor- and stain-resistant pants. Long ago, in a dot com boom far away, I worked as a software engineer at Ask Jeeves, an early attempt at AI-enabled web search. We didn't call Jeeves "AI" at the time; that term was out of vogue, redolent of mid-century science fiction. Instead, we called Jeeves "Natural Language Understanding," the marketing term for natural language processing, which we defined in Chapter 1. We built a wardrobe recommendation search engine so Jeeves could pick out a Gap or Nike outfit like a proper digital butler. Fast forward twenty-five years and I can try using any number of free AI-enabled personal clothing assistants "who" will set me up with an entire wardrobe of outfits I can "repeat." In fact, I did just that while writing this book! There are a bunch of clothing recommendation apps available on your smartphone. Turns out that using X resulted in Y.

What don't I have time for but need to get done? This one is straightforward and probably something you're already benefitting from anyway. Go into your email account and look for your spam or junk folder. Open it and witness the result of teaching a machine to do beneficial work. We discussed this in a previous chapter, but it's worth revisiting. Spam email is a fact of life. Artificial intelligence controls spam. You couldn't pay me enough to filter spam out of my inbox. I'm very grateful for the AI that does it for me.

Now It's Your Turn

Start by answering these three questions for yourself:

• What am I good at?
• What do I want to do better at?
• What do I have to do but life gets in the way?

Try not to overthink it; just write down whatever comes to mind for each. Now go through the exercise of looking for ways to augment your life using AI for each question. If you get stuck, it may well be that there isn't a ready answer—yet! If that's the case, consider what you would want from an AI tool and be on the lookout.

If you find an option that seems helpful, remember to always evaluate the AI using what you've learned in the previous chapters. You are an informed user and consumer of AI, and you are capable of making a decision to use or not depending on your own viewpoint.

It's important to step back and focus on your humanity now that you've considered AI in your life. The most important thing you can do is recognize what makes you uniquely human. You don't want to be a machine. You'll never compete successfully with a machine. So don't try, and instead focus on your human abilities. That's the point of a wonderful book called Futureproof: 9 Rules for Humans in the Age of Automation, by Kevin Roose. Here's an example. My friend Jean is an architect with a thriving independent practice. Another architect I know, David, is getting out of the business after seeing his income erode. David is incredibly technically proficient. His eye for design, proportion, and ability to model is top notch, but he's not so great with people. Jean is an excellent designer, but maybe not as good as David. What Jean has is empathy, patience, and an understanding of how to help her clients navigate the rocky emotional and financial experience of designing and building something as personal as a home. Jean brings her humanity to bear on her business. She has what some call emotional intelligence, and importantly, the ability to use it in her work. David, not so much. David and Jean are already subject to the pressures of automation. AI-enabled architecture design software is becoming ubiquitous. I'm not worried about Jean in the coming years, but David is smart to seek alternatives.

Augmentation shouldn't be about becoming a cyborg. Futureproof is a great read because it illustrates that if you just use machines to hustle harder, work faster, you'll eventually be replaced by a robot. Augmentation should be about helping yourself be more human, doing more of the things humans can do. Cyborgs and Terminators are creepy because they are machines pretending to be humans. When we use AI to try to make ourselves more machine-like, we're creepy too. Use AI to be more human, not the other way around. Reading Futureproof can help you understand your uniquely human qualities in a world of increasing automation. It's an important book because it can help you develop a defensive strategy for the change AI is bringing. Help you differentiate yourself by strengthening your best human qualities. In this book, I strive to present an optimistic offensive strategy. I hope to help you build on the message of Futureproof and use AI so you can make informed choices, influence how AI develops, have more fun, and have an even greater impact on the world around you.

Using AI for Personal Safety

My mutt dog, Lilo, is a perfect example of specialized, superhuman intelligence in action. She lies dormant on the couch or bed until her sensors detect something of concern. It could be footsteps on the driveway, the scent of a fox approaching our duck coop, or the sleepwalking of one of my children. I will never surpass her ability to hear, smell, or intuit. I will never beat her vigilant cognitive processing that detects and responds to concerns at all hours, day and night. Even while in a deep sleep, upside down, looking ridiculous, she will lift a head, cock an ear, sniff the air to gather more information. She growls a warning above a certain threshold. When a threat is confirmed, she loses her mind, barking and scratching at the door to defend her humans from possible harm. If I were allergic to dogs, I would want a machine or alarm system that could do all of this for me.

Safety concerns present humans with one of the greatest opportunities to benefit from AI. A machine can be taught to be ever vigilant and to detect possible harm before we can. Machines can learn from our environment what is "normal" and what may be something to worry about. A machine can learn that the mail is delivered every afternoon. Which, apparently, my dogs can't do because they attempt to murder our mailman, Bill, every day at two-thirty in the afternoon, even though he leaves them dog biscuits.

Safety-critical situations are a great place to take advantage of AI. You're building layers of redundancy by adding a machine into your life to keep you safer. In my own life, I have seen major safety benefits from the AI in my car.

Imagine Lilo perched on the dash of my car. She stares intently through the windshield, never blinking. I put the car in reverse to back out of the driveway and she runs to the back of the car, furiously scanning through one hundred and eighty degrees for any hazard. She barks at the sight of a pedestrian walking into the path of the car, out of my line of sight but visible to Lilo's well-trained eyes. I brake and the pedestrian strolls safely by (hopefully without a dog of their own so I don't end up with the two knuckleheads barking uncontrollably in the car).

The good news is you don't have to train a dog to be your safe-driving assistant. Most new cars now include AI systems that do even more to keep you safe on the road. As we discussed in chapters 1 and 2, these systems are trained to recognize a bunch of hazards through cameras, radar, and driver inputs to the steering wheel. Some even watch your eyes and can detect when you're getting sleepy. Every time you drive a car with one of these systems, you're putting the machine through continuing education. The car records and uploads data about the safety system to the manufacturer to help improve the AI (and probably sell your data to Starbucks so they can decide where to put a new store, but let's focus on the positives). Driving safety is a very positive outcome from all of us collectively teaching machines. I know I'm a safer driver in my reasonably priced 2024 Subaru that came standard with Advanced Driver Assistance Systems (ADAS). My ADAS has prevented at least two collisions where I absolutely would have rear-ended another driver. In both cases, it detected a car coming to a sudden stop in front of me and slammed on the brakes just before I did. The machine saved me from an expensive inconvenience at best, and from hurting another driver and myself at worst. When our teenage drivers go on long trips, this is the car they take. My wife's and my criterion for a kid's car used to be the cheapest car with air bags and a top crash safety rating. Now, when we can afford it, we help them replace their old beaters with cars running AI safety systems. Never going back!

This safety feature also helps my dad drive safely at night now that he's lost vision in one of his eyes. When he got his first car with an AI safety system, his stress levels dropped from knowing that everyone would be safer when he had to drive my mom to the hospital or come home from a concert after dark. I am a car enthusiast and truly enjoy the raw driving experience. I own and drive three classic vehicles (1982 Jeep CJ5, 1982 GMC C6000, 1995 Toyota Tacoma) that don't even have airbags, much less sensors and AI safety systems. Warning beeps and robots making decisions for me seemed like an intrusion into a meditative escape. That can be true, but the value of these systems is so clear, I now do ninety-nine percent of my driving in a 2024 Subaru with the help of AI. AI can really, truly improve our safety, the safety of our loved ones, and the well-being of people with whom we share the road.

Be a Smarter Patient

The potential for AI to help us be healthy, or at least less unhealthy, is remarkable. If AI captures human knowledge and makes it accessible at scale, then the knowledge of clinicians could have a great impact through its careful use. The health industry is an enormous, nearly $20 trillion-with-a-T sector of the American economy. Venture capitalists are already investing in AI startups and directly in the sector and are certain to invest more with that much money at stake. The typical economics of health mean that investment follows the money, as has been the case in the pharmaceutical and medical device industries. In my work, the earliest adoption of AI is in hospital financial operations—billing, collection of unpaid bills, and pre-approval and increased approval of insurance claims. These are all worthwhile efforts to gain incremental efficiency in the complex financial enterprise of a large hospital. After financial operations, AI is being used to sustain or increase the margin of already (relatively) high-margin treatment and diagnostic care. In both cases, investment follows returns. But there is so much more potential.

So how is a patient to profit from AI? First, try to avoid being a patient. Health screening is an excellent area for the use of AI. At a minimum, when it's time for your colonoscopy, you can look for a gastroenterologist who also uses GI Genius or another computerized adenoma detection tool. Ask them how it has affected their adenoma detection rate. It's a good sign if they have a concrete answer. For other types of screening, get a sense of the growing number of health AI systems approved by the Food and Drug Administration (FDA), listed on their website (https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices at the time of publication). More than one thousand had been approved at the time this book was written. Ask your doctor, nurse, or hospital administrator what they're doing to evaluate and adopt AI systems that improve the care they provide. You may be surprised at their answer. Ask them how they are paying for the system so you can evaluate how their incentives align with your own health needs. Ask them how they have ensured the AI will perform fairly for all patients. Ask them how they are checking the performance of AI over time to make sure it continues to perform.

Cornucopia

What follows is a collection of real stories from people just like you who have tried using AI. Consider these stories encouragement to try AI for the same use, or for a different but related use. Ask yourself what aspects of the story could translate to your own life or work. Get in the habit of asking people you know how they are using AI. You may be surprised at their creative problem solving.

Awesome Editor

Ann, the developmental editor of this book, used AI for the first time to edit the Introduction. She started by editing the raw draft manually, using her laptop and Microsoft Word. This gave her a version she was happy with as the product of her decades of expertise. She then went back to the raw draft and fed it into an AI service, asking the AI to edit it for clarity and enhance it for content. We then sat down together and compared the two results, one entirely manual by a human expert, one done by an AI. We found that Ann made edits the AI did not that were important to me, the author. She had acquired a sense of my "voice" through our multiple meetings to develop an outline and her subsequent reading and editing of my draft. Her editing helped bring my voice out and make it stronger in the written words of the Introduction. Watch a singer with their vocal coach and you'll see the coach help the singer find more expressive, impactful vocalizations of the sounds they are creating. That's what Ann does for me. In comparison, the AI is like the auto-tuner used by recording studios and pop musicians to polish and make vocals appealing. You hear about "heavy" versus "light" use of the auto-tuner. The Introduction we got back from the AI was like the auto-tuner turned up to eleven. Technically, my words and ideas, so very polished, but no longer… me. The AI-edited Introduction appears in the Appendix, along with the prompts we used to generate it. Decide for yourself which is "better."

Japanese Dinner

My friend Rob used AI to teach himself to put a seven-course meal on the table while everything was hot. Rob loves to try new things in the kitchen. He recently started learning traditional Japanese ingredients and recipes, cooking one dish at a time using cookbooks, videos, and online information. After getting familiar with a bunch of dishes, he wanted to do something special for family and friends and cook a multi-course meal. The problem was that it was really hard to learn how to time the preparation and cooking of a wide variety of dishes so they would all come to the table hot at the same time. He asked an AI for the ordered series of steps required to prepare, plate, and serve all seven recipes. He followed the composite recipe and succeeded at serving a hot seven-course meal the first time he tried.

Memorial Podcast

My friend Diane wrote a memorial message at the passing of her mother. She read it aloud to friends and loved ones at her mother's funeral, sharing a warm remembrance of a life well lived. A number of people who could not be there heard about Diane's message and asked for a recording of her reading, which didn't exist. Uncomfortable with making a standalone recording, she decided to try an AI service typically used to create audio podcasts from written information. The resulting podcast had two perfect-sounding human voices sharing Diane's written message as a warm conversational tribute to her mother. She sent the audio to far-away friends and family, who took grateful comfort from the tribute

Wholesale Metal

My friend Johan uses AI to read metal supplier specification documents from hundreds of suppliers to his wholesale metal business. He works with many suppliers to build and maintain a diverse inventory of products from space-age alloy tubes for building airplanes to rough steel stakes for farming tomatoes. Each supplier sends an electronic copy of a document describing the metal product, lot, and batch information along with the actual metal products on a truck. That information needs to be transferred into Johan's inventory and sales database. It's challenging because every supplier uses their own document format. Stock-keeping unit, description, batch number, lot number, and safety information all appear on every document, but in slightly different places. He used optical character recognition (OCR) software for years to partially "read" the documents, with limited success. Recently, he started experimenting with AI systems that have been taught to retrieve information from different parts of documents based on a prompt such as "give me a list of all stock-keeping unit identifiers along with the accompanying description." He hopes this tool will help improve the quality and timeliness of inventory and sales information so his sales team can sell more and his warehouse team can fill more orders.

Robotic Coffee Machines

My friend Stephan uses AI to help his coffee machine customers troubleshoot problems in faraway places. He cofounded and leads Poursteady, a robotic pour-over coffee machine company. But these coffee machines are global. Poursteady machines are installed in coffee shops as far away as Korea and the Middle East, meaning that when a customer needs help, it could be the middle of the night in Brooklyn. He subscribed to an AI service that "read" all of the technical and how-to documentation and the ten years of customer-support emails Poursteady had collected. This was an example of fine-tuning a pre-existing generic foundation AI so it can help with coffee machines. The result is an AI chat "bot" that can answer questions and help troubleshoot problems based on its prior knowledge of the world (from the foundation AI model) and what it has learned about Poursteady's products (the fine-tuned AI). His staff further teaches the AI by asking it for help based on their expert knowledge of what has gone wrong in the past and confirming or correcting its answers (something you do when you use ChatGPT). With a well-taught machine in hand, his team hopes to give it to their distributors in Korea and the Middle East so they can better use the knowledge captured from Poursteady experts in Brooklyn. The economics of Stephan's young business are such that he can afford to either manufacture his coffee machines in Brooklyn or hire humans to do twenty-four-hour customer support, but not both. He is choosing to manufacture in the U.S. and scale the customer support knowledge of his expert team with the help of AI.

Boat Influencer

Dockdeals is an Instagram account run by an anonymous boat enthusiast who uses AI to quickly create high-quality visuals for boat-for-sale listings he is excited about. He uses a service called Canva to process low-quality pictures from the for-sale listing into high-quality images that jump out and better represent what he thinks is cool about the boats. He learned that Instagram's algorithm favors high-quality pictures and will show his posts to more people if he uses AI to improve the images.

Team Builder

My friend Jason runs a consulting company that uses AI to put together the best possible teams. His company builds financial analysis tools for businesses all over the world. Projects are successful when one or two of his analysis tool experts work as a team with one or two financial experts from the business. He subscribed to a service that uses an AI model to match people into high functioning teams or at least identify possible personality conflicts to be on the lookout for and work to avoid. He used the AI service with a few projects and found that it had helpful suggestions that improved the performance and satisfaction of the teams.

Insurance Appeals

A former colleague used early AI to automatically appeal denied insurance claims and pre-approvals to pay for prenatal diagnostic testing. This system overwhelmed the insurance companies, who then came to the table to negotiate blanket approvals. In my experience, insurance companies deny and delay payment in the hope of creating disincentives that reduce the rate at which they pay claims, likely with the help of AI. In this case, my colleague fought fire with fire and used AI to level the playing field.

Financial Reporting

A friend uses AI to analyze and generate summaries of why financial results change from quarter to quarter for the divisions of a holding company. Every division—and there are many—is required to explain why their sales, expenses, and profits are up or down. The head of the division provides a narrative explanation, really the only way to analyze and capture the complexity of the situation. In addition to human analysis, they use AI to read all the explanations, identify unique events, find possible trends, and generate summary explanations across an incredibly complex array of subsidiary businesses.

Search

My godson uses ChatGPT instead of Google for all his web searches. He was in his freshman year of college when ChatGPT was released. He started playing around with it early on. When I heard he bought a paid subscription to ChatGPT, I felt it said a lot about the utility he found in the tool. On top of not having much money because he's a full-time student, he's also frugal in the nicest possible way. Generous to others, but it pains him to waste money. He uses ChatGPT constantly to look up information and identify trees and buildings from pictures he takes with his phone.

Study Aid

My daughter used AI to generate practice problems while studying for a final exam. She struggled with a class in college one recent semester. It was a requirement for her major, and she needed a C or above. There was a real possibility of a D or worse! It wasn't enough to go to all of the professor's office hours, seek help from the teaching assistant, study with classmates. She prompted AI by asking it to generate practice problems for concepts in the lecture that hadn't been covered in prior homework assignments.

Retail

An engineer I interviewed for a job on my team used AI to help convenience store retailers stock their shelves to maximize sales-per-square-foot. The next time you go into a large chain convenience store, look up at the ceiling for dark plastic bubbles. You're right that these are for security to detect and prevent shoplifting, but if the store is a customer of my candidate's company, they do more. The convenience store AI has been taught to watch how individual people move through the store, browse shelves, choose items, and make purchases. After a few weeks of watching customers, it recommends changes to the store layout and to where products are displayed. The process continues after the changes are made, in a constant cycle of optimization and adaptation to changing customer behavior. Grocery stores are increasingly using this technology as well. Next time you notice that your favorite candy bar or bag of chips has moved to a new location, it's because AI thinks it'll sell better there.

I could go on, but in the end the best way for you to learn about AI is to use it. Hopefully, the variety of these stories can help you see how AI could be useful to you. AI is changing and evolving so rapidly that it's not feasible to write a definitive "how to" instruction manual. Also, the best AI should be easy to pick up and learn. With the information in this book, you now understand the fundamentals of how you and people like you teach machines and can choose how you engage with AI on your own terms. You are better equipped to recognize AI in the world around you, and with a healthy awareness of the current AI industry, you're a more knowledgeable consumer. Use your knowledge to influence how you live with AI by voting with your feet and wallet. You are better positioned to see the value of your data and decisions and to help your own school, business, or employer maximize the value of both in the use of AI.

Keep Up!

I started a podcast with my twenty-something daughter, Maryjane, so I can keep up with the evolution of AI in the real world—more how it's impacting people's lives as they use it, less the evolution of the underlying technology. We talk to students, educators, and experts about their life with AI. It's called "You Teach the Machines" and you can find it wherever you get your podcasts.

We started the podcast because this book is simultaneously the best way to share a holistic view of AI and the worst way to keep up with the constant change happening at the time you're reading or listening to it. Artificial intelligence evolved significantly in the time it took to write and publish this book. When I started writing in November of 2024, the latest available language AI were capable of responding to your prompt or question with a single answer. Four months later, the latest available AI was able to respond with an answer backed up by the steps it took to figure out the answer! It shared its reasoning! As an insider, this was both hugely necessary and mind-blowing.

The steady and rapid pace of change will continue, punctuated by Black Swan events. My hope is that by reading this book you shed uncertainty about AI and the fear it can cause. I hope that any remaining fear of the Five D's are rational and help you make better decisions. I hope you are better able to recognize your own Survival Signals when bad actors in the AI industry try to manipulate you in their marketing, and that you choose to support the AI companies that take the high road. I hope you are better equipped to manage unexpected change brought by future Black Swans. I hope you are better able to think critically about side effects and pitfalls of AI and, as a more informed consumer, can make choices consistent with your values. I hope you have a new appreciation for the incredible value of your data and the data you generate—truly how you teach the machines—and exert more influence over how it is used. Most of all, I hope you see AI for the incredibly useful tool it can be when it augments your life. I encourage you to seek out AI and experiment with it in your life and work. Be the windshield, not the bug. You teach the machines!

References

Roose, Kevin, 2021. Futureproof: 9 Rules for Humans in the Age of Automation. Random House.

U.S. Food and Drug Administration, 2025. Artificial Intelligence and Machine Learning (AI/ML)‒Enabled Medical Devices. (Retrieved on May 10, 2025, from https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices)

Listen to Chapter 4 Part 1 of my book You Teach the Machines! If you find this helpful, please support original writing and buy the full book wherever you get audiobooks. Available from Libro.fm, Amazon, Audible, Apple and many more. Also in print at Amazon, Barnes and Noble, and my favorite: delivered to your local bookstore through bookshop.org. Help other readers by leaving a review on Amazon or Goodreads! Thanks so much --Jeff

CHAPTER 4: Side Effects and Pitfalls

"The vitality of democracy depends on popular knowledge of complex questions."

—S.S. McClure

Writing this chapter, in which I present what many see as the "bad news" of AI, was simultaneously depressing and encouraging. Depressing because, at the time I'm writing, a relatively small number of large corporations are deploying AI into our lives as fast as possible. And it's all pretty opaque. Encouraging because major change from AI has yet to happen. There is time for you, me, our loved ones to shape change for the better. To be a driver, not a passenger. You teach the machines.

The words came easily, but I became dejected while building a point of view from facts, interpretation of facts, and theories to explain what is not publicly available. There's a lot that is behind a curtain. You intuitively know AI will reshape your life. Simultaneously, you don't understand how. You can be overwhelmed by this combination of knowledge and uncertainty. I became overwhelmed and depressed as I considered the negative implications of this new technology accelerated by a generational deployment of capital, concentration of wealth, erosion of education, disruption of jobs, and shifting global security. My editor stepped in and coached me to focus on the specific, the actionable. Always good advice.

In this chapter, you'll see my editorial point of view come through, so be a critical reader. Know that I remain an AI optimist, so I try to balance points of potential doom with action you can take. The legal and publicity departments at the companies I discuss may argue with what I write. In many ways I am rooting for these same companies to succeed. They're doing incredibly difficult and historical work. I invite them to help make a second edition of this book even better. But a corporation is legally obligated to seek one simple outcome: Maximize profit. The reality is that better human outcomes depend entirely on you, me, your parents, your kids, the values we teach, and the decisions we make. I try to give you at least some idea of how you can be part of the solution to the problems I discuss. But if there is one thing you should take away from this chapter, it is that you need to prepare for the unknown. Prepare by taking stock of your first principles. Mine are "Be nice. Get stuff done. Make things less crappy." Medical professionals go with "Above all, do no harm." What are yours? We're in for a lot of change, currently driven by corporations in effect experimenting and gambling with our economy and lives. Anchor yourself with clear principles that can steer you when unexpected change from AI hits. Humans are built to adapt. We're going to do a lot of it in the coming decades.

A side effect is an unintended bad thing you experience from doing something else. A headache from taking antibiotics, maybe. A pitfall is a known hazard you allow yourself to fall into. A headache from drinking too much.

An unintended side effect in the world of AI? Depending on your point of view, the relative reduction in investment in renewable energy in favor of investment in nuclear energy. A corresponding pitfall "we" knowingly step into with more nuclear energy? The coming increase in solid nuclear waste stored on site at nuclear energy plants, at least in the U.S., because we as a society, represented by the people we've elected for the past twenty years, are politically unable to pull off long-term consolidated storage. See Yucca Mountain.

But even this side effect can have a balancing upside. Investment in nuclear energy is bringing real innovation in the form of more efficient, cleaner nuclear reactors. And if you consider a reduction of investment in renewable energy a side effect because of climate change, then you have to consider that use of nuclear energy is better than burning more fossil fuels.

Regrets

In May of 2023, Geoffrey Hinton resigned from Google. Eleven years earlier, he and his team had built the first neural networks at the University of Toronto. They founded a company that was quickly bought by Google for $44 million. Dr. Hinton went to work for Google to advance the research. Ten years later, at the time of his resignation, he stated that a part of him regrets his life's work (Kleinman & Vallance, 2023).

The main inventor of modern AI regrets his life's work. Sit with that.

Geoffrey Hinton, an insider's insider, knows AI as much as or more than anyone else on the planet. He resigned from Google, the original industrial AI company, so he could speak freely about the hazards he sees at Google and beyond.

Dr. Hinton was a hero of sorts to me and my colleagues working in health AI long before he stood on principle. He worked for decades against conventional wisdom to prove the power of computer programs modeled on how neurons in the brain learned skills by analyzing data. After Google joined the AI arms race started by Microsoft's investment in OpenAI in 2020, he became concerned that his company and its competitors were moving too fast, given the stakes for the rest of us.

He grew concerned that rapid proliferation of "fake" AI-generated text, video, and voice would make it impossible for us to know what was true. He grew very concerned that we would lose our jobs and incomes as AI replaced or cheapened the labor of paralegals, analysts, call center workers, writers, lawyers, financial experts, doctors, nurses, engineers, and software programmers. He became very, very concerned with the weaponization of AI into autonomous killing machines. Dr. Hinton wasn't alone. Even before his resignation, over a thousand technology leaders called for a moratorium on training advanced AI. They wanted time to understand possible side effects and work to minimize the harm of known pitfalls.

Too late. A year later, Microsoft effectively bought a nuclear power plant. Bezos, Musk, Pichai, Nadella, Altman, and Cook—the modern-day Stanfords, Rockefellers, Dukes, and Morgans—couldn't risk someone else winning. Shareholders demanded returns. Not just some mysterious shareholder "other," but each and every one of us invested in the tech-heavy U.S. stock market. Google ignored the call for a moratorium and rolled out AI-generated search answers at the top of their search page.

Change

Side effects and pitfalls flow naturally from change. Artificial intelligence is a miles-long freight train of change driven by hundreds of billions of dollars. You, I, your parents, your kids are locked in a stalled car at the railroad crossing. Artificial intelligence is changing or will soon change how you write a report for work, an essay for school, improve your firm's profits by automating junior associate work, drive a car, identify mental health problems, deny insurance coverage, get your electricity, trust or mistrust information, experience art and entertainment, and fight wars.

Which changes will bring side effects? Which have known pitfalls? Wouldn't it be nice to take a minute and think about it? Like the experts wanted "way back" in 2023?

Practical AI went from invention to industry in ten years. Neural networks emerged in 2012 and became scalable five years later with the Transformer in 2017. Corporate Industrialization into a financially and politically intertwined handful of corporations? Five years between 2017 and 2022. What took more than one hundred years for the first Industrial Revolution took only ten for AI. As I write this in 2025, the Big AI companies are in a race to remake the knowledge economy. How many quarterly earnings reports do you think they're willing to produce before they can report returns to their impatient investors? The leadership and shareholders of the Big AI companies in the U.S. alone are betting hundreds of billions of dollars that they can return trillions as fast as possible. Look at the concentration of wealth in the hands of the leaders of these companies and their investors. Again, they have a legal obligation to maximize profits. Do you think they're truly, fundamentally interested in growing the whole pie?

AI has never happened before. It went from theory to practice in ten years. The economy of AI as it's currently playing out means the richest corporations control the means of production right up front. Contrast this with the rise of the internet and World Wide Web. Public communication protocols arose out of publicly funded research and were taken up by anyone with a computer and a phone line. Web browsers and server software freely available to all allowed people to use their existing phone lines to build their own websites at home. Internet service providers sprung up at the local town level. The web quickly became of, by, and for the people.

Artificial intelligence is on the opposite track. The Big AI corporations possess barely comprehensible financial power. They use real and perceived expertise to gain political influence based in part on a popular assumption that AI is central to national security. Multiple sessions of Congress and multiple presidents have come and gone with no new regulatory guardrails in the U.S. Hundreds of billions of dollars already at stake demand returns. It's as if Gutenberg and the early printing press experts weren't chased out of Mainz during an unrelated religious power struggle so the printing press could disseminate organically. It's as if, instead, they formed a corporate combine, an industrial business group that held absolute power over the manufacture and use of the printing press. It's as if they cozied up to and contributed hundreds of millions in dark money to the most powerful leaders in Europe of every party to insulate themselves from regulation and maximize profits. It's as if they relaxed standards on control of misinformation in the books they printed to gain influence. It's as if human knowledge and skill were captured, transferred to others, and used for good and bad in entirely unprecedented new ways by corporations concerned only with winning the trillions of thalers at stake.

Rapid societal, cultural, and economic change directed by profit. Maybe it'll be fine. Let's look at puppies and covet the lives of others on social media instead of worrying.

With that rant out of the way, it's time to buck up and be part of the solution. "You teach the machines" could mean you're passively milked of your data and money. Or it could mean you're in front of the room, directing and taking charge. You're the windshield, not the bug. You teach the machines.

Dr. Hinton's Fears

Geoffrey Hinton gives us a framework for the first major known side effects and pitfalls we'll discuss: misinformation, job loss, and killer robots (the use of AI in war). In my professional life, I've spent a lot of time working on the first two, albeit focused on the health sector. Thankfully, I've never met a killer robot in a "hot war" but have had direct experience with scary "cold war" AI threats and harm.

Misinformation

Consider the same internet search three months apart for "european causes of accidental death" using the Google search page, which around ninety percent of us do globally. I have screenshots of everything to prove this actually happened.

On November 19, 2024, Google's AI Overview listed the causes as:

• road traffic injuries
• drowning
• falls
• burns
• poisoning

It stated that "road crashes are the most significant cause," which I took to mean this was a rank ordered list.

On January 27, 2025, Google's AI Overview reordered and changed the causes to:

• road traffic accidents
• falls
• drowning
• poisoning
• work-related accidents.

In three months, falls overtook drowning and burns dropped from the list to be replaced by work-related accidents.

Which is true? Turns out neither, according to my own research on the website of EuroStat, the statistical office of the European Union, which Google seemed to point to as a source for both AI Overview results. A half hour spent with the freely available data there revealed the causes, in order of decreasing death rate (deaths per 1,000 people), to be:

• falls
• other
• transportation accidents
• poisoning
• drowning

"Other" includes a scary list of things like struck by falling objects, exposure to animate mechanical forces, and overexertion.

Generative AI by its very nature will always give you (generate) an answer, an image, a song, a diagnosis. Truth doesn't matter in subjective "eye of the beholder" circumstances like composing an advertisement, and you need a starting point. Truth matters in the realm of life-and-death facts and figures. Say you're a busy European legislator and need a sound bite for your speech about accidental death. You draft the speech with results from the top of the search page and understandably miss the small-font disclaimer that "Generative AI Is Experimental." This is what Dr. Hinton was talking about when he resigned. People lose sight of what is really true with unreliable or manipulated AI as an intermediary.

You'll hear the term "hallucination" thrown around when AI presents something that you figure out is straight up wrong. I prefer "drunk uncle" because I'd rather think of AI in the context of the example above as an inebriated relative spouting off malarkey and conspiracy theories. Hallucination is a kind word used by some people who don't want you to think too hard about the fact that they are putting out technology that is inherently flawed, that they know it, and are more interested in profits than the integrity of information. A lie is an intentionally false statement. I have to believe Google knows its AI Overview makes false statements, yet they put it out there intentionally. Hallucination, drunk uncle malarkey, or lies? You decide. And by so choosing, exert influence.

I'm going to leave you to extrapolate to other important situations where you or someone you trust uses an unreliable intermediary like AI Overview. School, health, work, personal finance. A single screwy web search may seem harmless until you multiply eight and a half billion searches per day by this demonstrated potential to be both inconsistent and wrong. Is this OK?

Let's be the windshield, not the bug, and—to mix a metaphor but keep it automobile-related—put ourselves in the driver's seat. What's your first principle for truth? When do facts matter to you? What's your threshold for trusting an intermediary? When does it matter if you believe your drunk uncle or not? Start by deliberately picking and choosing when you take information at face value when it comes to you via AI. Are the stakes low or high for whether the information is true? Would you go with what your drunk uncle tells you or ignore him? Be a skeptic at whatever level is right for you in that situation. When objective truth matters, AI should be considered wrong until proven right, at least as it's being rolled out to us in the mid-2020s. If you're going to trust AI, consider verifying through a third party that the AI has controls in place to detect and remove misinformation.

Fake text, audio, and video are easily generated with AI. A one-time spike in downloads of my podcast originated in a foreign country a few months before I started writing this book. At the time, we'd published thirty-three episodes, with one or two downloads from the same foreign country. I was excited to see an unexpected bump in downloads, but something looked fishy. All thirty-three episodes had been downloaded at once, something that had never happened before. I looked for information using Google, and it turns out the podcast community sees this type of activity regularly, and not just from foreign countries. The accepted explanation is that these bulk downloads are data harvests by AI companies working on voice generation AI. The hard truth is that I am at increased risk of a "deep fake" of my voice because my recordings have been harvested by an unknown entity in a country sometimes viewed as an adversary to my homeland. For that reason, we don't use last names for guests on the show.

Misinformation can be more subtle, too. I have a couple of friends, one young, one… not so young, who are single and trying to meet people to date. Instagram shows them both a steady stream of content along the lines of "You don't need anyone! You're strong, independent, and don't need anybody!" TikTok feeds another friend a steady stream of "The opposite sex is controlling and mean!" Major social media apps show us what they think will grab and keep our attention. In the olden days of the web, when I worked at Ask Jeeves, we talked about "engagement" and "stickiness" of a website. Could we gain "eyeballs" and keep them looking at a website longer so we could show an ad or sell something? Social media makes billions on the same principle. Behind the scenes of Meta, the understandable set of rules that Instagram and Facebook started with (my friend is interested in boats, so maybe I will be, too) has been replaced entirely by AI. And that AI will do whatever it takes to gain and hold our attention. It learns that negative stereotypes and "us vs them" division will grab lots of people. We're evolutionarily hardwired to respond more to differences than similarities. So that's what social media AI feeds us.

When it comes to truth, social media AI is effectively unmanageable. It may be politically expedient for the leadership of social media companies to say they are increasing or pulling back on their fact-checking efforts, depending on which party is in power. But the reality is they've lost control and couldn't do the fact-checking if they wanted to. The machines they taught to gain and hold our attention move too fast and are too complex to govern. How have you seen misinformation spread in your life? Where could AI have played a role? Did you contribute to the spread? Remember, you teach the machines.

What's the windshield stance with more subtle misinformation? Decide what your first principles are. When do you care that you're being manipulated to gain your attention? Turn awareness into action and make more deliberate choices about what social media you use and how you interpret and consume social media content. Run an experiment: Click on a series of negative or divisive suggestions from the AI. Search for a divisive or negative topic. Observe how your feed changes. Do your friends and loved ones use social media? Have a conversation with them about what you discovered in your experiment. Don't like what you see? Engage with social media AI on your own terms. Vote with your feet and seek alternatives.

Job Loss: Automation over Augmentation?

Artificial intelligence machines can be taught to do work that once only humans could do. A friend asked me for help understanding AI. He was in a new job and had been tasked with learning the alphabet soup of AI: LLM, GPT, GPU, NLP, GenAI. I happily obliged over lunch. We had a great discussion, during which he shared public information about his company's products. One of the products is an AI that watches video feeds from multiple hospital rooms. Its job is to watch the patient on the video and automatically alert a single human mental health monitor (babysitter) to bad stuff—bad stuff like patients trying to harm themselves. The promise was a reduction in the number of humans required to care for at-risk patients. Buy this AI, save even more money because you won't have to pay as many people. On one hand, I am all about reducing the cost and improving outcomes of mental health care so maybe we can have more of it. I've also been in one of those rooms. I've experienced the healing power of human connection and warmth a mental health monitor provides while keeping the patient safe. The business could have sold this new technology as a way to help existing mental health monitors do an even better job, get an even better outcome. Collaborate with mental health monitors to come up with ways to augment their care. Maybe by giving them insight into hidden signals in the patient's behavior and mental state so they can intervene and provide the right support before a crisis occurs. Instead, the business was making the case for saving money by automating work required by regulators and accepting the loss of human connection. Pure automation over augmentation.

We've been here before, just not for knowledge work. The summer I turned nineteen, I earned money as a brick making machine operator in a factory that made specialty bricks used in high-temperature smelters and kilns. I would press a big red start button and step to the end of a conveyor belt. The brick machine stamped out three bricks and deposited them on the belt every few minutes of my eight- to twelve-hour swing shift, minus two fifteen-minute breaks and thirty minutes for lunch. The raw bricks were fragile and crumbly, made of an exotic (for bricks) mix of specialty clays and minerals, I picked up each brick with very flat hands pressing evenly on two sides of the brick, and deposited it on a metal rack whisked away to an oven by forklift when full. I was probably one of the last summer workers to be hired because the first robotic brick handler was installed that summer. The ten or so full-time operators were offered training as robot technicians. One, a man named Baker, took the opportunity that summer. The other nine refused with reasons varying from "It'll never work for all the types of brick" to "I'm no geek; Baker, you're a nerd!" I left at the end of August to go back to school, never to return. If the company is still in business, my guess is it's with all its brick machines operated by robots, no manual operators in sight.

A few years after I worked at the brick factory, my first job in technology was to get data out of databases. I learned to write code in "Structured Query Language" to select rows and columns from different parts of the database. The language was and remains a well-standardized and powerful way to tell a database what you want (the "query"). The trick was knowing both the language and what the data meant. I understood biology, so was able to quickly and accurately retrieve the right data from a database full of complex biological information. I earned a good living with these skills and others as I built a career solving problems with data and technology. In the two years leading up to writing this book, I saw the same dynamic from the brick factory play out in my technology career. One person in ten embraced new AI tools that "understood" both the structured query language and the meaning of rows and columns in a database. The remaining nine were some combination of frightened, comfortable, and skeptical. My job was to drive adoption of these AI tools because they would allow more scientists to work directly with the data they needed to do critical research in child health. This would scale up and speed up science previously bottlenecked by the relatively small (vs the needed) number of technical experts who had for decades been the only intermediaries between scientists and databases. Resistance came from both camps—the technical experts and the scientists. Both would have to step out of their comfort zones. The technical experts would need to give up some autonomy and replace soon-to-be-obsolete skills with new, more advanced knowledge. The scientists would need to learn to use the new AI tools rather than rely on an expert. It is very human to be uncomfortable with and resist change, but the resistance was discouraging nonetheless, given what was at stake: increased pace and breadth of discovery in child health. I got a grant to create new training programs, but outside of early adopters, the central tendency was to stick with the old model. I am not without empathy. It's hard to learn new skills, especially when you're already working full time, have kids, and a life outside of work. But just as automation changed manufacturing work, AI is changing knowledge work. Windshield or bug?

Do you remember Louis Winthorpe III and Billy Ray Valentine? Both are fictional characters in the movie Trading Places, a 1983 comedy about wealth disparity. Winthorpe, played by Dan Aykroyd, was a privileged genius at making money trading frozen concentrated orange juice, bacon, and other commodities. Valentine, played by Eddie Murphy, was a smart-mouthed hustler living on the streets of Philadelphia. I won't spoil the plot, just share that it's a great window into the 1980s in lots of ways. If you've never seen it, or it's been a while, watch and throw the AI that recommends movies in your streaming service for a loop. As you watch, ponder which characters would have the same job in the age of AI, and which would be out of or in a drastically changed job. Hint, Winthorpe wouldn't make it if he held tight to his colorful blazer and open outcry trading on a chaotic market floor covered in slips of paper. He probably wouldn't make it even if he made the transition to electronic trading at a desk somewhere far away from the old stock market exchanges with their bells and hand signals.

The frozen concentrated orange juice and bacon commodities markets are where buyers and sellers bet on the price of breakfast next month or six months from now. Sellers auction off the right to buy their orange crop or hog herd at some future date. This allows farmers to hedge—insure—against the risk of crop failure. Buyers bid to buy the future crop, betting they'll be able to resell juice and bacon for a premium when it's actually time to put breakfast on the table. The auction used to be run by people yelling out, "Who'll give me three hundred dollars a ton for two hundred tons of orange juice for December delivery?" "I'll give you two-eighty a ton!" "I'll give you two-ninety!" "Sold for two-ninety!" Today, the auction happens at nearly light speed between computer programs taught to get the best deal on either side of the trade. This automation also reduces the transaction cost to nearly zero, which benefits both buyer and seller. Winthorpe would be on the street running cons with Valentine. Of course, Winthorpe is a caricature. Humans adapt, including financial traders.

I have a friend, Gerry, who is one of the smartest people I know. They are the first person I know to have their knowledge work completely automated. (Don't worry, the story has a happy ending.) Gerry's career as a trader started in open cry on the floor of the Chicago Board of Trade, transitioned to electronic trading at a desk upstairs, and was ultimately replaced by AI. I remember one phone call where they asked me, "Hey, how do you speed up a computer network? The algorithm is making money; now we need it to make money faster!" Gerry's work had evolved from making trades directly based on knowledge of risk and return in the market to using their knowledge to help computer scientists build automated systems. They taught the machine to make more money with less risk than they could manually. This got boring, so after twenty-five years, Gerry transitioned to a new career in high-end building renovation and construction. Lots of other traders had to make the shift much earlier, like our intransigent fictional Winthorpe. They were "old school" traders who didn't want to or couldn't contribute to the automation of their jobs like Gerry did.

It's no surprise that some of the biggest of the Big AI directly focus their data and AI efforts on only two areas of the economy: finance and healthcare. AI in other areas are left to partners, startups, and other big companies already working in the area, like Monsanto in agriculture. Financial services (including insurance) represent more than $7 trillion, healthcare $18 trillion in the U.S. alone. Finance and healthcare are also almost entirely based on expert humans performing knowledge work. The same Big AI companies investing hundreds of billions of dollars choose to dig into the two areas of the economy with the most valuable (in dollars) knowledge work. Don't get me wrong. They've set it up so they'll get a piece of everything. Startups and big companies in law, publishing, advertising, engineering, software development, and entertainment incorporate Big AI foundation models into their own products and pay a toll to Big AI. But healthcare and finance are where Big AI is focused "in-house." Finance and healthcare are where they're looking for "partners" to "co-invest" in data aggregation and training AI. You better believe each company intends to win the race to develop powerful and lucrative foundation models specific to healthcare and finance. The prize is the biggest possible share of potentially trillions of dollars transferred from human worker payrolls to Big AI service fees.

My friend Gerry lived through the leading edge of this inevitable trend. As their knowledge of the commodities market and work as a trader became less valuable, they adapted and taught the machine to analyze risk and return and make trades. Gerry moved on to finding creative new trading opportunities, figuring out how to make money, then taught the machine the new stuff. And on and on until it was more interesting to move to a completely different field.

People who do work based on knowledge and cognition watched automation drastically change the economic prospects of their fellow citizens who worked in what little domestic manufacturing remained in the U.S. Large corporations grew wealthy or just remained in business through robotic automation of what they couldn't move to lower-wage economies. The same is happening and will happen to knowledge workers. Call the customer service number of your bank to combine the balance of two accounts and close one. Twenty years ago, you talked to someone in Texas or Minnesota. Ten years ago, you interacted with someone in Mumbai, India. Now you interact with AI that is rapidly improving its ability to solve your problem. If the AI can't help you, that failure is recorded and you're directed to a person in a low-wage or low-regulation economy. The audio of your interaction with that person and the resulting actions they take to deal with your accounts are recorded and used to teach the machine to solve the new problem. That is, unless you have a lot of money on deposit with the bank; then you may speak with a real person in Texas. For now.

The best software engineers I known in my career live by the mantra "My job is to put myself out of a job." Here is a real story from the stress-detection project I expand on later in the chapter: The job is to retrieve heart rate data from two watches and combine it into one dataset where the data from one watch can be compared to the data from the other. The best engineer will write the most efficient, powerful, and user-friendly software possible so they never have to solve this same problem again. New watch? Someone else can adjust a couple of settings and it's done. Other, lesser software engineers might write unique code for each watch, code that requires them to fiddle and be in the loop. With the advent of AI, the best software engineers I know are immediately figuring out how the AI can write code for them so they can move on to solve harder, more important problems. They'll continue to do this, just like Gerry did with trading.

If you're a knowledge worker, the value of your current work is going to change. Do you read, summarize, and write reports? Do you write or edit words? Do you write or test code? Do you provide financial advice? Do you analyze data repeatedly? Do you research and make recommendations? Do you answer the phone and help people? Do you teach? Do you decide where to invest money? Do you make podcasts? Do you interpret lab results and make a diagnosis? Do you act? Do you edit video? Do you read x-rays and identify bone fractures or other problems? Do you design interior or exterior spaces? Do you talk to people about their cognitive and behavioral problems on the phone or video? Do you compose or perform music? Value real estate? Design bridges? At least the routine aspects of your work will be performed better, faster, or more cheaply by a machine. There is too much money to be saved or made for it not to happen, at least in the U.S. You will have to adapt or see your income go down. Use AI in your work in any way and you are already adapting. Windshield or bug?

You choose windshield! What to do?

Start by reflecting on work you do repeatedly but with your brain instead of your hands. We're used to thinking of repetitive labor as something that happens on an assembly line. Repetitive physical labor is subject to automation by physical machines with moving parts. Repetitive cognitive labor is subject to automation by AI machines with cognitive parts.

Before getting to knowledge work, consider how AI will change the time-is-money business of artists and skilled craftspeople. If you're an artist, a sculptor, a furniture-maker, a metalworker, a stonemason, a gardener, a chef, a pipe fitter you need to use AI and AI-enabled software to work smarter, not harder. Artificial intelligence running on your smartphone can collect and analyze data about your work to help you make more money from your time. Artificial intelligence‒enabled purchasing services can help you find the best deals on materials. Artificial intelligence can help you write your own advertising copy and generate your own graphics. Many of the business services you depend on, or wish you could tap into, may become cheaper because of AI. At the same time, be prepared for your customers who don't work with their hands to go through economic upheaval.

Most important? Do. Not. Give. Away. Your. Data. I cannot emphasize enough the degree to which the current U.S. corporate economy is harvesting the data of workers so that it can put them out of a job at some point in the future. This is not because these corporations are somehow evil. It's because they are intrinsically and legally obligated to survive and thrive using new technology to increase productivity. And productivity is calculated as output per worker. Dollars of revenue or profit divided by number of workers. In an ideal world, the whole pie will grow fast enough that every worker will continue to be employed and maybe even see their income grow faster than inflation. This happened for the most part during the bottom-up phase of the first Industrial Revolution. The AI-driven industrial revolution of the knowledge economy is starting with top-down industrialization, as we've discussed. Increased productivity through automating the work of expensive analysts, billing experts, doctors, lawyers, engineers, nurses, accountants, and home health workers is the model for success I see gaining traction with the people who sign the checks when AI vendors come calling in my professional life, at least in the U.S. Cost reduction is the first stop for entrenched organizations that have difficulty innovating. We've already offshored or automated just about everything else possible, so knowledge work comes next. So what leverage do you and I have? The data we generate in the course of our work. Consider yourself an independent contractor, the expertise you've gained over a career to be your intellectual property. The data you generate can be used to teach a machine what you know and how to do your work. Your employment contract almost certainly says that your employer owns this data. They do unless you speak up about it. Are you a lawyer? Ask the partners in your firm how they plan to compensate you for the use of your briefs, memos, and contracts to train AI. Are you a doctor, nurse, or billing agent? Ask your practice or hospital administrators how they plan to cut you in on the savings they hope to achieve through using your notes, diagnoses, reports, and discharge summaries to automate aspects of your work. Are you an investment advisor? Ask your managers how they plan on making you whole from the use of your client communications over the phone or via email to automate aspects of your relationship-building. Ask your leadership if their corporate contract with Big AI allows those companies to use your documents, emails, or code to teach AI about your industry, company, or specific job. If the answer is yes, then ask how much money they saved on the contract by agreeing to those terms. Suggest that perhaps they could share that with you and your fellow workers. If you're an author, carefully read the terms of the self-publishing service you use to put out your first book to figure out if your writing will be used to teach a machine to write books about AI.

Sound radical? The concept of a "data collective" is emerging among economists as a potential counterbalance to even out the economic risks posed by AI. In theory, fair distribution of economic gains that are only possible through the use of human data to train AI would grow the whole pie. Allow us all to share in the profits from this amazing new technology. It's not just theoretical. In 2025, the concept is just starting to be tested via the first lawsuits against Big AI companies for copyright infringement. Turns out if you're a publisher or a clothing designer (Gerkin, 2023; Rogelberg, 2024), you may think it's economically unfair when your text and pictures are harvested, potentially without your knowledge, incorporated into a foundation model, and then monetized by a Big AI company so people can write books and design clothes derived from your work. The question is, if the New York Times wins its suit against OpenAI, will it distribute a portion of the proceeds to its journalists, copywriters, and editors? Maybe. Maybe not. Ironically, twenty-five years ago, as the defendant in another lawsuit, the New York Times argued against its current position versus OpenAI (Pope, 2024). At that time, the U.S. Supreme Court decided against the Times and in favor of a group of freelance journalists whose articles the Times had sold piecemeal to database companies without permission (Supreme Court of the United States, 2001). We've been here before. You teach the machines.

Other economies may attempt to manage their AI industrialization following the example of Japan's Meiji Restoration in the late 1800s. The new leadership of Japan actively managed the transfer of industrial technology into Japan to maximize benefit and minimize harm. Keep an eye on deliberately managing the AI Industrial Revolution so as to at least try to benefit all. In the U.S., and by extension many of the places where Big AI companies dominate, it's a free-for-all driven by a few corporations with lots of political influence.

As I write this, my three children are just about to graduate from college, in college, and just about to enter college. All of them are entering an era of change. Guidance I (try to) give them? I start with my own first principles. Be nice. Get stuff done. Make things less crappy. When it comes to AI, I try to help them see that it will bring change. Stick to first principles when the ground shifts under you. Embrace the change and you'll be just fine. I emphasize to them the importance of two things: They need to use AI, and they need to learn about a specific area of the world they're interested in, one that is constantly changing, the faster the better. To succeed in an AI world, you have to be good at augmenting your work and your life with AI so that you can invest your time in what AI can't do (yet). Use AI so you know what's possible. Use it consistently so you know when it can do something new and useful, when it gets something wrong, or when it gets something right it previously got wrong. Use AI so you're not competing with it to do work that can be automated. Dream up and experiment with creative uses of AI nobody else thought of. Be the person who figures out how to use AI and help your family, friends, coworkers, and employer adopt and adapt. Use AI to pursue interests previously out of reach.

Regarding interests, I try to guide my kids to constantly learn about something they find intriguing. Their current interests are biology, art, and creating value in a business, respectively. The more they know about their interests, the more they can tell when AI gets things right and when it gets them wrong. The more they know about the world outside the bubble of AI, the more they understand AI isn't perfect and they can be informed and critical consumers.

Regarding change, I try to guide them to be more than comfortable with change. I say they must excel at finding opportunity and value in disruption and changing circumstances. Because things are going to change a lot faster over the next couple of decades vs the last couple.

I also encourage my kids, and you, to learn to do things with your hands. To be creative. My kids probably don't think I'm serious, but I made a standing offer to pay for them to learn a trade in addition to their higher education—welding, carpentry, plumbing, electrical wiring, house painting, stone masonry, tilework. Makeup artistry, hair styling. Avalanche control, emergency medical response. Knowing how to use AI and do valuable work with their hands means they have a killer combination. Combine AI, hands-on work, and creativity and you're futureproof. Pure knowledge work is likely to pay relatively less on average in the medium term. If the AI Industrial Revolution plays out on the more pessimistic track, wealth will continue to concentrate, but wealthy people still clog their toilets, want new kitchens, and value artistry.

Want a good outcome? Use AI. Learn constantly. Pursue change. Don't give away your data.

Killer Robots

You're not going to hear much about the use of AI in weapons of war. The Terminator movie franchise made robots with guns into a public relations nightmare for the AI industry. That doesn't mean it isn't happening. Geoffrey Hinton called this out as one of his reasons for resigning from Google. This could mean he knows something of what may be happening behind a top-secret curtain. Before we get to AI actually blowing people up, it's important to consider that not all conflict involves the guns and missiles of a "hot war."

In my professional life I have encountered credible hacking threats and actual attacks on critical health infrastructure. I've briefed three-letter agencies, not on defense against, but, scarily, the detection of theft of what I consider to be some of the most sensitive human information in existence. That means I was asked to come up with ways to figure out if information had been stolen after the unseen and undetected theft took place. By a state-sponsored attacker.

Cyberattacks, hacking, ransomware. Whatever you call it, state sponsored attacks on people's digital lives have grown into a new Cold War. Real examples of attacks include tricking you or me into giving access to our bank account, encrypting a hospital's patient data so it can't be used then extorting ransom for the decoder key, planting misinformation about a political candidate in social media, causing power outages, and remotely destroying nuclear weapons manufacturing equipment. "State sponsored" means that a government or agency in a government looks the other way, pays for, staffs, or even directly carries out attacks. Up to a certain point in political conflict, "good guy" and "bad guy" depend on your point of view. When it comes to cyberattacks, everyone does it. Artificial intelligence lowers the bar for what it takes to fight in this new, digital Cold War. You don't have a bunch of sophisticated hackers to write malicious computer code? Use AI to write the code instead. A hard-liner is up for election in a rival country? Use freely available AI to generate and post fake video of the candidate doing something sketchy in a hotel room. An economic competitor is moving in on an emerging market you want for yourself? Use AI to manipulate the local stock market with automatic trades. AI scales up information warfare. Flyers dropped from airplanes, propaganda broadcasts over the radio, and even "troll farms" are being replaced by AI trained to push issues through social media, finance, education, anywhere people interact with a stream of information.

Military AI emerges somewhere between information warfare and shooting warfare. The military uses, or will use, AI for lots of the same purposes for which it's being utilized in other areas. Logistics, intelligence gathering, management. The primary concern, however, is a category of military AI known as Lethal Autonomous Weapons Systems (LAWS). To the layperson, a weapons system is lethal and autonomous when a machine picks a target and fires a gun or missile without a human in the loop. Policy and military experts, not surprisingly, have to wrestle with a gray area. In 2025, the United Nations is an intermediary for governments to try to agree on a definition of autonomous weapons systems. They publish helpful information useful for thinking about AI in general, not just for weapons.

Autonomous weapons systems require "autonomy" to perform their functions in the absence of direction or input from a human actor. Artificial intelligence is not a prerequisite for the functioning of autonomous weapons systems, but, when incorporated, AI could further enable such systems. In other words, not all autonomous weapons systems incorporate AI to execute particular tasks. Autonomous capabilities can be provided through pre-defined tasks or sequences of actions based on specific parameters, or through using artificial intelligence tools to derive behavior from data, thus allowing the system to make independent decisions or adjust behavior based on changing circumstances. Artificial intelligence can also be used in an assistance role in systems that are directly operated by a human. For example, a computer vision system operated by a human could employ artificial intelligence to identify and draw attention to notable objects in the field of vision, without having the capacity to respond to those objects autonomously in any way. (United Nations Office for Disarmament Affairs, 2023)

There's some particularly helpful wording in the above paragraph: "…using artificial intelligence tools to derive behavior from data, thus allowing the system to make independent decisions or adjust behavior based on changing circumstances." Rewritten in a more human context, we get: "using your eyes and ears (data) to figure out how to behave (derive behavior), allowing you to make your own decisions or adjust what you do based on changes in the world around you (changing circumstances)." Truly autonomous driving lines up nicely with this definition. Think of AI driving a car in a safety-critical situation with potential injury or death from a traffic accident if the AI decides to accelerate or brake at the wrong time. When a weapon meets the same definition, AI makes another kind of life-and-death decision.

People often have an intuitive, negative reaction to drone warfare. A missile fired from a drone piloted by a flight officer hundreds or thousands of miles away can be viewed as wrong relative to a missile fired from an airplane by its pilot in the cockpit. The invasion of Ukraine by Russia took modern drone warfare to a whole other level from its origins in post-911 conflict. Suddenly, hundreds of thousands of drones were deployed—at first to drop grenades from above on nearby enemy positions and vehicles, then for one-way flights of explosives directly into the same targets, destroying the drone in the process. Both Russia and Ukraine used drones, but Ukraine, with its smaller population, leaned on the technology more. A soldier is safer piloting an explosive drone into a bunker from a few hundred yards away than fighting their way across an open battlefield to deliver the same explosives. Ukraine also embraced the "first-person view" (FPV) drone early in the conflict. A soldier wears goggles that display the view from a camera on the front of the drone. They "see" what's in front of the drone through the goggles as if they were right there. The soldier uses a hand-held remote control with joysticks and buttons to pilot the explosive-laden drone into the tank, bunker, or concentration of enemy soldiers. This visual immersion of the goggles enables the soldier to make more precise and rapid flight maneuvers than if they are looking at a screen. In fact, the FPV drone gained its first popularity in drone racing a decade earlier. A second, larger observation drone is used to view the attack from above and look for the next targets for a stream of one-way explosive drones to hit. Recordings of FPV drone attacks are regularly posted to social media.

What does this all have to do with global geopolitical concerns about military AI? Why did United Nations leaders suddenly gain traction on the issue in 2023, after first raising it in reports to the Human Rights Council ten years earlier? What does any of this have to do with Geoffrey Hinton calling out autonomous weapons systems when he resigned from Google in 2023?

In the U.S., we can look backward at an intertwined cast of characters starting in the first Obama administration and continuing through multiple administrations of both parties. Our players are Eric Schmidt, Robert Work, and Peter Thiel—all patriots and, I believe, acting in good faith, though one could argue in apparent if not actual conflict of interest. The following may be hard to follow because the natural revolving door between government, the military, and industry is hard to follow:

• Eric Schmidt, then-CEO of Google, was appointed by President Obama in 2009 to the President's Council of Advisors on Science and Technology.
• Along with others, under Eric Schmidt's leadership, Google funded (and continues to fund) the Center for a New American Security.
• Robert Work, retired Marine Colonel, was appointed Undersecretary of the Navy in 2013, leaving soon after to run the Center for New American Security for about a year until appointed Deputy Secretary of Defense in 2014.
• During his tenure at the Center for a New American Security, Robert Work advocated that the U.S. address the long-term threat of adversaries gaining an advantage in military AI.
• Peter Thiel, a longtime technology executive and investor, co-founded Palantir Technologies in 2003 to develop security software and now military AI. After the 2008 financial crisis, the first Obama administration paid Palantir for software used to detect fraud in stimulus funding and Medicare payments.
• Throughout both Obama administrations, Eric Schmidt maintained a professional relationship with Peter Thiel, with the two appearing on multiple technology panels together.
• During Robert Work's tenure in the Obama administration and subsequent first Trump administration, Google received a contract to develop military AI for the Pentagon. After employee protests, Google transferred the contract to Palantir in 2019.
• Eric Schmidt continued as chairman and then technical advisor to Google's parent company, Alphabet, until 2020.
• From 2019 to 2021, Eric Schmidt and Robert Work co-chaired the bipartisan National Security Commission on AI, which advocated for investment in AI for national security.
• In 2021, Eric Schmidt funded and founded the Strategic Competitive Studies Project, which has Robert Work on its Board of Directors.
• In 2022, Google and Palantir announced a strategic partnership.
• In 2022, Eric Schmidt and Peter Thiel launched America's Frontier Fund, an investment fund focused on national security in microchip manufacture critical to AI, weapons, and the economy. Their new investment organization lobbied the government for $1 billion in funding and won a lead role in an international investment fund led by the former CEO of the Central Intelligence Agency's venture capital subsidiary, In-Q-Tel. Official Pentagon policy and Palantir's publicly stated goal is that military AI augments human decision making. Palantir continued to win defense contracts from the U.S. and other governments throughout the first half of the 2020s.

I lay out these circumstantial points to illustrate that the historical relationship between the technology industry and national security apparatus is alive and well in the age of AI. I also lay out these circumstances to prepare you for a theory I'll get to shortly.

The revolving door is a good thing if you take seriously the national security threat posed by hostile use of AI (which I do). It also presents by its nature potential conflicts of interest and the need for transparency to stakeholders like you and me. Stakeholders who also have a responsibility to educate ourselves and participate, not just sit back and throw around conspiracy theories like your drunk uncle. In their final report to the National Security Commission on AI before it was dissolved in 2021, Chairman Schmidt and Vice-Chairman Work stated, "Americans have not yet seriously grappled with how profoundly the AI revolution will impact society, the economy, and national security." They were and unfortunately continue to be right. It's a scary world out there.

Russia stated in 2020 that it intended to replace soldiers with lethal autonomous weapons systems. Russia has established a drone weapons program in China, resulting in the U.S. placing sanctions on Chinese entities. Simultaneously, the Ukrainian military's success rate for drone attacks reportedly rose from fifty percent before 2023 to more than eighty percent in 2024, reportedly in large part through use of drones running Palantir software. Turns out killer robots weren't so sci-fi after all.

From a purely professional technical standpoint, I consider the video of drone attacks coming out of the Russian invasion of Ukraine to be priceless for training AI. The repeated nature of the attacks under varying landscape, foliage, weather, and countermeasures provides a rich sample of real-world images for teaching machines to identify targets. Surveillance drones' simultaneous recording of attack drone strikes provides built-in labeling of the result—hit or miss. The recording of front-line soldiers' commands and flight telemetry over and over during actual combat drone maneuvers paired with video and geospatial location under varying conditions provides a near-ideal, unique data set for teaching machines to fly drones into targets.

Russia also uses FPV attack and surveillance drones and is in a position to collect the same unique data and share it with its allies, or at least with "the enemy of my enemy is my friend."

As I mentioned a few paragraphs ago, I'm going to advance an informed theory that attempts to explain this landscape. Because of the top secret, or at least highly confidential, nature of the situation, my theory is certain to be a simplification of complex events.

The theory: Eric Schmidt followed in the footsteps of patriotic industrialists before him. He and others identified the risk that adversaries would develop military AI that could defeat the U.S. and its allies in the future. "They" would have more or better killer robots than the U.S. "They" would have more or better AI-enabled cyberweapons. Because he is simultaneously a patriot and a capitalist, Schmidt aligned Google with the Pentagon to drive research and development through multiple presidential administrations. Simultaneously, Peter Thiel aligned Palantir with the Pentagon and other U.S. government agencies to develop AI for national security, also through multiple presidential administrations. When Google's employees forced a change because of ideological disagreement over the company's involvement in military AI, Eric Schmidt and Peter Thiel worked together at the top of their companies to successfully transition the project from Google to Palantir. Schmidt and his original national security partner, Robert Work, continued to build the case for U.S. investments in military AI. Palantir developed its business through contracts with Israel and other U.S. allies, in addition to the U.S. A goldmine of unique training data became available from drone warfare in Ukraine. The U.S. directly and indirectly facilitated systematic and large-scale acquisition by Palantir of drone warfare data from Ukraine. Palantir used this data to rapidly accelerate development of military AI for drone warfare. Palantir simultaneously contracted directly with the Ukrainian government and military to develop and field systems that enable front-line soldiers to continue to train and fine-tune the performance of its military AI. United States intelligence agencies obtained evidence that Russia had been sharing drone warfare video data with China in return for access to Chinese drones equipped with military AI.

Now, in a lethal feedback loop, China also gets data from front-line Russian soldiers on the performance of its AI. The U.S. and China cannot take the risk that the other develops lethal autonomous weapons systems, so a secret military AI arms race is under way. Palantir, as a private company, is able to research and develop lethal autonomous AI, and the hundreds of billions of dollars in unrestricted aid going to Ukraine can be used to pay for it. United Nations policy makers receive reports of the pursuit of lethal autonomous AI by both the U.S. and China. On the U.S. side, the AI arms race, at least in part, involves Google through its partnership with Palantir, and almost certainly other Big AI companies as well. The United Nations starts to develop and advocate for policy to reduce harm of military AI. I'm an outsider so this is just a theory. It is also my theory that that Geoffrey Hinton knows enough from his insider status that he decides to resign.

As a child of the 1980s, I lived with the reality of nuclear deterrence through mutually assured destruction. Ten years before I was born, the U.S. and Russia came to the brink of nuclear war during the Cuban missile crisis, itself a response to perceived escalation by the U.S. placing weapons closer to the Soviet Union. We, as global citizens, eventually de-escalated through reciprocal nuclear arms control and non-proliferation treaties. We're not there with military AI. We don't yet (and hopefully never will) have the equivalent horrific example of the use of nuclear weapons in the bombing of Hiroshima and Nagasaki at the end of World War II, followed by decades of living under the real fear of nuclear holocaust. However, state or state-sponsored actors have deployed AI in cyberattacks. Russia has targeted civilians with drones in isolated terror attacks. It is a matter of scale to get to swarms of cheap, fully autonomous lethal drones overwhelming military defenses or, worse, being used in a terror attack on a stadium full of people. Possibly worse yet, the use of powerful AI to degrade or destroy military and civilian digital infrastructure. Or trust in our institutions and fellow citizens. This is the through line Geoffrey Hinton, many other technology leaders, and the United Nations are concerned about.

What can you do? I strongly suggest buying a cup of coffee or something stronger for the Chief Information Security Officer or equivalent at your company; they're dealing with more than you can imagine. After that, get informed and come to your own conclusion on whether or how military AI in any of its forms are acceptable. It is certainly inevitable. You may go on to demand more transparency about military AI programs, or at least how the money flows in what is an evolution of the Cold War military-industrial complex. You may advocate politically against an AI arms race, in favor of mutual defense agreements, or for spending on offensive and defensive capabilities. You may talk to your peers and kids about the importance of cybersecurity and data protection. You may work toward a global ban on military AI. You may choose a career developing effective AI weapons. It's up to you. You teach the machines.

"Other"

Geoffrey Hinton may also be concerned with other side effects and pitfalls. A few of the more fundamental ones follow.

Bias

Artificial intelligence is as biased as the data and people who teach it. The machine will learn from the people who tell it what it gets correct when looking at its training data. If that data reflects unfair biases because it contains only a biased view of the world, then the AI will have those biases. If the people teaching the machine hold a biased view of what's correct and that makes it through to the machine during training, then the AI will have those biases.

Imagine you take videos of cats. Like, a weirdly large number of cat videos. Say two thousand and counting. You prefer orange tabby and black cats, especially when they're sleeping (so cute!). Your two thousand videos contain mostly sleeping orange and black cats, with the rest showing brown, gray, and spotted cats mostly not sleeping—play fighting with each other, chasing birds, and eating. Your videos are in a cloud service hosted by a Big AI company. The same Big AI company also sells a home smart speaker device recently upgraded with a camera. Pat, a bright young marketing person, pitches a new feature for their camera devices: "Bad Kitty: your AI cat babysitter!" Even better, Pat gets Nick Bruel, author of the beloved Bad Kitty children's books on board to co-brand the thing. Pat's vice president loves it and greenlights the project. Since it's legal, you've accepted the terms of use, and live in the U.S., the Big AI company gives your cat pictures to the AI team without asking or informing you. The team sends your pictures overseas for labeling by low-wage AI workers, again without asking or informing you. The workers dutifully tag the videos of sleeping orange and black cats with "good." They tag the videos of mischievous brown, gray, and spotted cats with "bad." Pat's VP is in a hurry to publicly announce the product at the company's rapidly approaching sparkly launch event in Las Vegas, so the AI team quickly teaches a machine to automatically recognize cats misbehaving on camera and yell "NO! BAD KITTY!" through the speaker. Just before the Vegas event, Nick Bruel pulls out of the deal. Turns out he has an orange cat that is a complete jerk and an angel of a brown cat. The pre-release camera speaker he got in the mail gives his orange cat a pass when it knocks the toothpaste cap down the drain and yells nonstop at his poor brown cat. The VP is embarrassed, fires Pat, and cats everywhere breathe a sigh of relief because they don't have to live in a surveillance state.

A silly make-believe example, but an entirely realistic illustration of how harmful bias can make its way into AI out in the world. Look for articles about bias in AI and you'll find plenty of examples in much more critical areas of our lives. I encountered a near-miss formative example early on in my work.

In a research study I did together with a behavioral health scientist, we wanted to figure out how to help caregivers and teachers by detecting stress in nonverbal children who couldn't otherwise say, "Hey! I'm freaking out over here!" We taught AI to recognize stress from changes in heart rate detected by commercially available heart-rate watches. We had an intern with brown skin working on the otherwise all-white research team. She figured out that one of the otherwise ideal brands of watch gave bad readings on her skin but not on white skin. We immediately excluded the watch, recruited study participants with different skin tones, thanked the science gods for the luck of that intern, and kicked ourselves for not thinking about what could happen if we hadn't had her. The research continues, the AI has promise, but it could have started out with a very damaging implicit bias against people with brown skin.

Wondering how we taught the machine to recognize stress without doing harm to vulnerable nonverbal children? We collected a bag of tricks, including a remote-control rubber rat that darted across the floor, a puzzle that couldn't be solved, a jack-in-the-box with a scary clown face and spooky music, and a wind-up jumping rubber spider. Research participants, first adults, then children with their parents, were put in a room wearing the heart-rate watch and subjected to jump scares. We collected heart-rate data and time-synched video so we could see when changes in heart rate co-occurred with the rat darting across the floor or the creepy jack-in-the-box popping. Not my favorite moment as a researcher, but everyone was informed and consented ahead of time and the work was for a very good cause. Can't make this stuff up.

What can you do about harmful bias in AI? Turn awareness into questions that ultimately drive expectations and accountability. Ask your doctor if they're using AI to help them care for you. If they say yes, ask them how the AI has been taught to avoid harming patients from any one demographic. Pay attention to proposed regulations and legislation to require transparency and mitigation of harmful biases. When your employer rolls out new AI, ask how it was trained to avoid harmful bias. Better yet, ask how it will be monitored for harmful bias going forward. Pessimistically, recognize that AI could be used to deliberately reinforce bias in conjunction with misinformation. Bias toward buying a brand of toothpaste, for example. Bias against voting for one candidate over another. Be on the lookout and question how AI may reinforce your own very human biases to manipulate you. Optimistically, recognize that AI can help reduce bias, too. A teacher who is biased to be more sympathetic to kids with one skin color may tend to ignore signs of distress in kids with a different skin color. Artificial intelligence that can detect stress in kids with skin of any color is a powerful equalizer to get a better outcome for all children.

Electricity

Industrial AI companies need more than the existing electricity grid you depend on to light your home and run your coffee maker. The biggest AI companies are like aluminum producers in that industrial AI needs high-volume, stable electricity. Aluminum smelters are often located right next to big fossil fuel, hydroelectric, or geothermal electricity generators—sometimes built just for that purpose, whether nearby communities wanted them or not. And just as aluminum producers bring in a steady supply of bauxite ore, AI data centers bring in a steady supply of raw data. Aluminum companies bring bauxite to aluminum smelters. Raw data flows into GPU computers. Aluminum smelters need lots of electricity to electrolyze bauxite into primary aluminum billets. Big AI data centers need lots of electricity to train foundation model AI from raw data. Aluminum producers' billets are subsequently turned into cars, soda cans, and fishing boats. Big AI's foundation model AI are subsequently turned into chatbots, x-ray diagnostics, and search assistants.

In 2025, AI companies are already using a lot of electricity from fossil fuels to train AI. They know burning even more coal and natural gas for their rapidly expanding data centers will put a lot of carbon into the atmosphere. They know this will make it harder for them to make money from customers concerned about climate change. They know that what they're doing will add to the risk of floods, hurricanes, droughts—extreme weather affecting the economy they depend on and the data centers full of their AI computers. Industrial AI is already mitigating this side effect with nuclear energy and its low- to no-carbon output. The pitfall? More nuclear waste in the form of spent fuel. Although fuel reprocessing technologies promise to reduce the final volume of high-level radioactive waste, there will still be more to safely dispose of for the hundreds of years it will take to lose its radioactivity. Consensus favors burial deep underground in stable geological formations without nearby groundwater. Absent these sites, high-level waste will continue to accumulate in "casks" at the reactor site. China, Japan, Canada, and many EU countries are in the process of selecting and designing long-term burial locations. The U.S. has one but, for political reasons, has been unable to license it for use. What can you do? Decide how this reality matches up with your values and make choices accordingly.

Cooling Water

Data centers use water to manage heat. Effectively, all the electricity consumed to train and operate AI is turned into heat. Data goes into the GPU. The GPU burns electricity to do calculations. The process releases heat. The GPU produces data.

Think about when your laptop, phone, or computer gets hot when it works hard. Global estimates for 2025 put annual electricity usage by data centers above five hundred terawatt hours (Masanet et al., 2020), equivalent to two trillion of those red heat lamp bulbs keeping french fries warm. Graphics processing units get slow when they get hot, necessitating quick removal of heat from the data center. This process requires billions and billions and billions of gallons of water, ultimately evaporated to achieve actual cooling (think about pictures of nuclear power plant cooling towers; that's what's going on). The need for data center cooling water comes on top of the need for water to cool the nuclear reactors that generate the electricity in the first place. Usually, this is fresh water because salt water is corrosive and bad for pipes and machinery. This is fresh water that would otherwise stay in the ground, a river, or a lake and be available for fish to swim in and people to drink.

What to do? Same as with electricity. No matter what you think, electricity and cooling water are costs, so companies have an incentive to work to minimize usage. Decide how water usage by AI companies aligns with your values and make decisions accordingly. Make purchasing decisions, or ask your employer or stockbroker to make purchasing decisions, based on your views about water usage by AI companies. The point is that every industry we depend on in our modern lives consumes resources. Artificial intelligence is no different. Just as with packaging, transportation, and housing, your responsibility is to be informed and make decisions about how you consume AI that align with your values.

Individual Rights

Your unique data is likely the most valuable asset held by any technology company you've knowingly given it to, or that has it anyway because you agreed to obscure Terms of Service (at least in the U.S.). Mark Gorenberg and Ash Fontana were two of the first venture capitalists to invest in AI. They wrote an article in 2016 about the new rules for venture capital in technology that includes the following about the new requirement for a start-up to have a data strategy:

Data Strategy: A crucial part of your plan to build an intelligent software product. A clean, unique data set is a competitive advantage in itself (so don't sell it!). From there, you can start building predictive models with your customers' data and turning successful experiments into features that help them make decisions. Finally, you will have a product that uses incremental data to improve models; making the product better, attracting more customers, getting more data and so on — a "Virtuous Loop."

Use of your data to gain a competitive advantage is a fundamental pillar of AI. Your individual rights to correct, mask, prevent use of, or simply delete that data are in direct opposition to the profit motive of technology companies. At its best, this remains a virtuous cycle where use of your data directly benefits you. There are no guarantees, especially in the U.S.

While traveling abroad in the EU, I checked the privacy policy of a major U.S.-based news website I subscribe to. Since I was accessing the news site from the EU, I found (and have a great screen capture of) a page that stated the news site and ninety-four of its "partners" would store and access any personal data they could collect through the site to do all sorts of things, including "understand audiences through statistics or combinations of data from different sources." I was also able to see the full list of ninety-four partners and the exact use they make of my data. I had the choice of opting in to this arrangement. No sharing would happen if I did nothing or rejected the option. When I got back to the U.S., I went to the privacy policy to find I did not have the same rights to control my data. In place of the relatively clear EU-mandated language, there was a confusing and ultimately opaque set of terms that kind of disclosed what would happen with "partners" unless I was a resident of California, Oregon, or Nevada, in which case I had some other rights. But not at home in Pennsylvania!

The hard reality is that there's a massive global marketplace where your data is collected, used to maximum competitive advantage, sold, traded, and monetized. Friends who work in marketing, politics, and philanthropy have access to scarily detailed subscription-based databases that can be used to profile you and your behavior down to the most minute detail. Up until the mid-2010s, I had to call my bank to file an "international travel plan" when I was travelling abroad so they would allow credit and debit card transactions to go through. Now? The bank has an AI model that's been taught everything about me. It'll allow a late-night poutine purchase in Quebec City to dilute the beer with cheesy gravy fries, no travel plan required!

Banks aren't the only large organizations that know, or are able to know, everything about you recorded in the digital world. PRISM is the code name for a surveillance program run by the U.S. National Security Agency. Under oversight, government agencies are able to search through your and my individual-level data processed by U.S. technology companies, including all the Big AI players. PRISM is the program brought to light by Edward Snowden in 2013. It continues and is the basis for a series of legal actions brought in the EU by a private nonprofit called None Of Your Business (NOYB), a comical name if the issue weren't so serious. The organization was founded by Austrian lawyer Max Schrems who, while studying at Santa Clara University in Silicon Valley, was blown away by Facebook's lack of respect for individual privacy. Schrems used the EU's privacy laws to get his hands on Facebook's store of his own personal data. As a result of his actions, a patchwork collection of oversight was put in place to allow citizens of the EU to safely use Facebook, Apple, Google, Microsoft, etc. The oversight is fragile and was gutted in the early months of the second Trump administration. It remains to be seen if and how the oversight is reinstated. I encourage you to spend some time reading about the privacy rights action being taken by NOYB on their website: www.noyb.eu.

This all gets creepier when you consider that the companies subject to PRISM collect and can be forced to share pictures, videos, voice recordings of you, your parents, your kids. Not to mention your fingerprints. Travel by air in the U.S. lately? Notice the "optional" face scan going into use at airport security checkpoints? On a recent flight, the security worker was rapidly and emphatically instructing passengers to insert their state-issued ID and look into the camera to be scanned. The facial recognition scan is technically optional, as stated in the fine print on the sign off to the side, but what I saw did not at all appear to be voluntary. My line didn't have it in place yet; otherwise, I would have refused and had more to share here! The same systems are going into place on some European airlines. China isn't the only surveillance state.

Your individual copyright is also at risk from AI. A copyright gives an author or creator "the exclusive legal right to reproduce, publish, sell, or distribute the matter and form of something (such as a literary, musical, or artistic work)" per Webster's dictionary. Copyright protects you and me from someone taking our writing, music, art, or performance and using it for their own financial gain. Three publishers, including the New York Times, have brought a lawsuit before a U.S. federal court, arguing that OpenAI and Microsoft are making money from copying and processing news articles without asking or paying for the right. The roughly fourteen million authors and editors of Wikipedia articles may also have a complaint, as their work was copied and used to train OpenAI's ChatGPT, among other AI. What if the corporations had to pay all the authors of Wikipedia a toll for redistributing their knowledge?

There are more known side effects and pitfalls you could inform yourself about. By all means, go dig in and be more informed. But AI, in the U.S. especially, is happening whether you like it or not. Your job is to maximize benefit and minimize harm according to your first principles. There are many unknown side effects and pitfalls ahead. You must accept uncertainty and be prepared to respond to events as they unfold. You teach the machines.

Knowledge, Uncertainty, and Ignorance

In 2002, then U.S. Secretary of Defense Donald Rumsfeld was probably referencing a beautiful Persian poem by Ibn Yamin in a press conference when he made his (in)famous "known unknowns" remarks. The poem wrestles with knowledge and uncertainty. Mr. Rumsfeld was attempting to help people process the early days of the Iraq invasion and the uncertainty that came with it. The famous poem gives us a gift: a framework for living in an uncertain world. Here is the poem, translated by theoretical physicist Niayesh Afshordi (2016):

One who knows and knows that he knows…

His horse of wisdom will reach the skies.

One who knows, but doesn't know that he knows…

He is fast asleep, so you should wake him up!

One who doesn't know, but knows that he doesn't know…

His limping mule will eventually get him home.

One who doesn't know and doesn't know that he doesn't know…

He will be eternally lost in his hopeless oblivion!

Knowledge and Action

"One who knows and knows that he knows…

His horse of wisdom will reach the skies."

The straightforward side effects and pitfalls are the ones you know of, understand, and act on to avoid or reduce harm. When you see bad and do something about it your "horse of wisdom will reach the sky." Pay attention to known problems with AI that have solutions, and if you're motivated and able, be part of the solution.

Knowledge with Uncertainty

"One who knows, but doesn't know that he knows…

He is fast asleep, so you should wake him up!"

Harder are side effects and pitfalls you can see but not understand or know how to address. You know something bad is going to happen but not exactly how to do something about it. You are fast asleep, so you should wake up! Trust your instincts. If you think there is a problem with AI in your work or life, but don't know for sure, you're probably right. Follow the "see something, say something" rule. Bring it up, talk about it. Try not to catastrophize, rather to understand.

Knowledge with Your Head in the Sand

"One who doesn't know, but knows that he doesn't know…

His limping mule will eventually get him home."

This is the head-in-the-sand or profits-before-principles category that is the most disturbing to me: Ignoring known risks in a mad dash for "creative destruction" by corporations and their leaders—the same Masters of the Universe who brought you the dot com crash, the 2008 financial crisis, and subsequent recessions. Mark Zuckerberg, knowing he'll be OK in the end, plows ahead despite the risks. "His limping mule will eventually get him home." It's not just the Big AI "them," but increasingly also the leadership of every corporation or large nonprofit built on human knowledge work. Geoffrey Hinton is a canary in the coal mine, warning us that there is internal knowledge with external denial. Your job is to ask the hard questions in the all-hands meeting, the school board meeting, or in the customer satisfaction survey. Bring up solutions to go with known problems. Show leadership even if you're no AI expert.

Ignorance

"One who doesn't know and doesn't know that he doesn't know…

He will be eternally lost in his hopeless oblivion!"

There are unknown risks. Things we can't see, often because of groupthink. You're completely ignorant of what might happen. Worse, you aren't even aware of your ignorance. You're blind to what might happen and may have a false sense of security. With stakes as high as our world faces with AI, it's important to consider where we may have a blind spot. Nassim Nicholas Taleb's book The Black Swan is helpful again in this context. Often, individuals will call out a risk that their group can't see. Pay attention to these lone voices, often seen as contrarian or disruptive. What can they tell you? Would it really cost that much to mitigate a seemingly unthinkable risk, considering the downside?

In conclusion, AI is brand new but has a lot of momentum behind it. Ibn Yamin's poem goes against our instinct, our drive to stay in our comfort zone of knowledge, to put our heads in the sand. Artificial intelligence brings much uncertainty and holds unseen challenges of which we are ignorant. The best defense is a good offense built on as much knowledge, understanding, and practice as we can pull off. Which brings you to the final chapter—using AI in your life!

References

Afshordi, Niayesh, 2016. He Will Be Eternally Lost in His Hopeless Oblivion! (Retrieved on April 24, 2025, from https://nafshordi.com/2016/07/26/he-will-be-eternally-lost-in-his-hopeless-oblivion/)

Fontana, Ash, & Mark Gorenberg, 2016. Growing Up in the Intelligence Era. TechCrunch. (Retrieved on May 9, 2025, from Growing up in the intelligence era | TechCrunch)

Gerkin, Tom, 2023. New York Times Sues Microsoft and OpenAI for 'Billions.' BBC. (Retrieved on May 2, 2025, from New York Times sues Microsoft and OpenAI for 'billions')

Kleinman, Zoe, & Chris Vallance, 2023. AI 'Godfather' Geoffrey Hinton Warns of Dangers as He Quits Google. BBC. (Retrieved on April 22, 2025, from AI 'godfather' Geoffrey Hinton warns of dangers as he quits Google)

Masanet, Eric, Arman Shehabi, Nuoa Lei, et al., 2020. Recalibrating Global Data Center Energy-Use Estimates. Science 367, 984‒986. (Retrieved on April 24, 2025, from Recalibrating global data center energy-use estimates | Science)

Pope, Audrey, 2024. NYT v. OpenAI: The Times's About-Face. Harvard Law Review, April 10. (Accessed on May 14, 2025, from https://harvardlawreview.org/blog/2024/04/nyt-v-openai-the-timess-about-face/)

Rogelberg, Sasha, 2024. Fashion Giant Shein Has Been Slapped with Yet Another Lawsuit Alleging Copyright Infringement, Data Scraping, and AI to Steal Art: 'It's Somewhat Shocking That They've Been Able to Get Away With It' Fortune, April 16. (Retrieved on May 2, 2025, from Artists sue Shein, accusing it of using AI and data scraping to steal their art | Fortune)

Supreme Court of the United States, 2001. New York Times Company, Inc., et al., Petitioners v. Jonathan Tasini et al. (Retrieved on May 14, 2025, from https://www.law.cornell.edu/supct/pdf/00-201P.ZO)

United Nations Office for Disarmament Affairs, 2023. Lethal Autonomous Weapon Systems (LAWS). (Retrieved on April 23, 2025, from https://disarmament.unoda.org/the-convention-on-certain-conventional-weapons/background-on-laws-in-the-ccw/)