Humans + AI

– Tim Stock

Tim Stock is an expert in analyzing how cultural trends and artificial intelligence intersect. He is co-founder of scenarioDNA and the co-inventor of a patented Culture Mapping methodology that analyzes patterns in culture using computational linguistics. He teaches at the Parsons School of Design in New York.

Website: www.scenariodna.com

LinkedIn: Ufuk Tarhan

Faculty Page: Tim Stock 

Ross Dawson: Tim, it's awesome to have you on the show.

Tim Stock: Great to be here.

Ross: So I think people need a bit of context for our conversation in understanding the work you do. A lot of its trends are around culture mapping. So, Tim, tell us what is culture mapping.

Tim: It's a culture mapping, really has its roots in understanding what is going on underneath the surface that people aren't paying attention to. So I searched essentially when we speak to whoever I need to explain cultural mapping to, it's to help companies understand how and why culture is changing, and how to use that information to make better design and business decisions. And so a lot of those kinds of real changes in culture are not obvious. They're not things that we can ask people about. So they're the weaker signals. And so culture mapping allows us to be able to map the relationship between what is the broader culture and subcultures and understand the relationship between those and how they develop narratives within society and cultural change.

Ross: So where's the state of today? So what are some of the signals you're seeing in the Euro cultural mapping work?

Tim: Well, I think that  , we're in a particular moment where we're shifting from one kind of age to another in terms of, especially in terms of how people do work, how we understand our relationship to identity, there's a growing nihilism, I would argue that's going on. And I think that  , when people say a lot of when we're talking about things that are happening, the negativity that people would say, is coming out of the pandemic. But again, if you see, from a cultural mapping standpoint, those signals were there already, but those externalities of the pandemic just really exacerbated them. So things like, sort of issues around how we see work, and how we understand our relationship to work, has a lot to do with how technology is changing, it has a lot to do with the kind of work that needs to that is the kinds of skills, all of these kinds of affordances that go along with that. And essentially, culture is always trying to catch up to that particular change. And so at this particular moment, I say, we're kind of stuck. There's a moment where we haven't found our voice yet. And so it's the reason why we see a lot of this kind of there's political dysfunction. There's, there's issues in terms of I mean, we're, we're at a moment where there's a lot of unrest, and there's a lot of language around that. And so essentially, I see us trying to as a society, trying to find that trying to find that voice.

Ross: So, yeah, there's a couple of directions for this, we're looking at the role of generative AI, one is from a cultural response. Another is, I suppose a deeper level is to our understanding of what is our relationship with generative AI.

Tim: Yeah, I mean, it comes down to what do we do? And I think that that's that  , that nihilism is emerging from well, what am I supposed to do? What is what caught the, we, we've almost coped, we've co-opted a lot of these words like intelligence. So what is left for humans to do? And the state of AI, I would say, is that you would see that there's a lot of replacing, and mimicking human actions. As sort of, we get sort of things that look like they're created, the word creativity, for example, has been co-opted, and sort of like so. But, we're at a point where we need to be asking what is creative. I mean, creativity is a human action, human intelligence emerges differently, the machine intelligence, machine and child, that, machines don't see ghosts, machines don't understand, machines can't believe in, in conspiracies in the same ways that humans can humans, see in between. And it's how we access information learning, learning as children, how we acquire language, and how intelligence is so tied into narrative in that particular way. But right now, what we see is a lot of replacing things that we normally would do. So the question is, is it especially if you're a young person today, you'd be like, Well, what should I learn? What are the kinds of skills there's nobody to tell you? Because it hasn't been framed? To be able to understand, so in a way, there's a lot of shift towards the individual. We're seeing sort of, in every area, from medicine to education to work, we see the individual now having to take on much more responsibility. And I think that causes a lot of anxiety. And so, that's the moment we're in until we sort of figure out well, how are we going to use these tools? So that sort of more collectively? I don't think we figured out the collective part of that.

Tim: Yes, yes, last year, I created this mapping intelligence framework. And my first response, as we started to look at chat DBT was that we don't know what we mean when we say intelligence, essentially, that humans are the reference point. So there's this debate where we happen to come up with a phrase and you use the phrase artificial intelligence all the way through whereas, if it could be a production of history, you might have called it cybernetics or some other phrase, which will give us completely different frame but because it is art or artificial intelligence, its objective has always been to copy and to replicate or try to be the same as human intelligence, which means puts us in a challenging situation. Now, when these are tools we can augment us, Doug Engelbart talked about augmenting intelligence instead of intelligence, artificial intelligence, intelligence augmentation instead of diligence. So this in a way requires cultural and linguistic reframing so that we can move to the amplifying condition types as reframes. 

Tim: Well, it's a culture mapping, it's a culture mapping exercise because it means so much of what is currently broken in, in AI in generative AI is that I mean, it's been programmed, it's, it is one framework, it has walls, and it is defined this language is defined by the engineers that defined it. And the rules that it lives by are it's sort of what, what we consider intelligence is what an engineer would consider to be intelligent, which would be to be able to replicate or, and so forth, speak to an artist or speak to a poet, and you get a very different answer. And you almost get this, we're seeing this very ideological kind of battle going on, and sort of reclaiming kind of ownership over how these, these models are actually trained and so forth. But  , more simple than that, what do we consider to be intelligent? When I think of chat, GBT, I like to think of it, it's the perfect way of sort of measuring intelligence, because, what you should be saying, we all kind of go through this one, we'll put something into check GBT, it should be like, caught that stupid like that. Why? Why isn't that better? That's our ability. The problem is that we're too accepting of these answers. And I think that we sort of end there is that part of us as human beings, I mean, the difference between machines and humans, and this is key to cultural mapping as well. True intelligence is a social function. It's about so-called social cohesion. And so like intelligence happens in groups, it does not happen in individuals. So I think that we have this this idea that individuals kind of create this great, great things in our society, it happens in groups, it happens in these subcultures of groups that create that kind of knowledge, machines can't do that. But the problem is, we also can be, we can be sort of swayed towards accepting that information, it's how conspiracies happen. And so like, that's the moment we're at, where we have to be more aware. We are, in a way, using when I think of augmenting our intelligence, I think of it's always interesting, bringing up gender bringing up AI and sort of certain cases. And it's like the case in terms of dating, the idea of being able to sort of date versions of up-to-date versions of yourself to stop you from making the same mistakes and so forth. It's like learning throughout this process of vamos helping you to be the person that you want to be. 

But the thing is, most of the things in society allow us to sort of copy and sort of, sort of double down on these bad habits. And I see, in a way so much in AI is doing that the Internet didn't have that problem. Generative AI is too quick. It's almost like the arc of adoption is so fast. And so with the internet, it allowed us to stay in the subculture space much longer. So it was able to kind of get that kind of life the whole cyberpunk and all. Have these other kinds of surveillance culture and everything, and there are so many cypherpunks and everything emerging with AI now, it's almost like it's an immediate meme. It's all it's like, it goes from zero to cottage core, and in two seconds, and it's like, then everything looks the same. And we go, Well, isn't that okay? We go, Well, if I said, No, will I be wrong, and people don't want to be wrong. And so they're like, we're all kind of affirming a lot of these sort of negative aspects of it right now.

Ross: Yes, yes. So I'm saying a lot these days that the biggest risk with AI is over-reliance, where we sort of say, Oh, that's good. And we just leave it at that, and we don't exert our cognitive capabilities or stop when it starts going not being as good as it could be. But I think that goes back to your point that intelligence has been a social function. And one aspect of that is that many people find it very useful to be in dialogue with AI you can refine your ideas and have a useful conversation sometimes from an emotional perspective, sometimes from a just idea generation perspective. But it is still essential, it is not a true counterpart. And I don't think you can have the same dialogue you can have with a single group of humans,

Tim: And you can't and the other part of that key to this Ross that's different is the fact that we're having a dialogue and we're the dialogue you want to have with that it's sort of like I want to, I want to get better, I want to improve something. But the thing is, that normally within social behavior is that you do that. And if you then begin to disagree, then you create other subgroups of people that believe your particular idea. And you develop a whole group and ideology around that. And that becomes its area of development, you get tools, you get other kinds of technologies that way, the problem is, we're almost creating this very linear, and it's, there's no, there's no divergence. 

I mean, the key to intelligence is divergence. It's not to affirm what's there, it's actually to push away from that and to do something that is diversity is key to biological you to our biological health, and like, it's the same in terms of intelligence is that you, human beings, by nature, tell them to do something and that they'll do they have this in, ingrained in them to move away from that if we make it easier to be like everybody else. If we are, we're diluting and diluting and diluting our intelligence with that process. So in a way, the internet was a form of AI, because it was all of the subgroups working this collective intelligence in that way, understanding how to sort of come back to that in some, some some way with with these augmented these hugely sort of transformative augmenting tools like generative AI, artificial intelligence is that would be that would be what we would need to be moving towards.

Ross: Yes, well, yeah, I love what you're saying to me intelligence is, is diversity or is grounded in diversity. And that is, it can be one of the useful functions. I mean, asking for diverse perspectives during the VI on particular situations is one of my favorite tools and says, Oh, I hadn't thought of it that way. And that is something additional, but it is complementary to my cognition not to, it's not the machine doing the intelligence, it might come up with a random, useful and useful but it Yeah, and I'm the one who's passing it. But it is useful to get those perspectives. 

But this in a way, it comes back to how we can do this as well as possible. And to your point, I think there's conceivably homogenization of thoughts of some kind, which is potentially emerging from this. So how can we use these tools to augment to amplify or perhaps even a better word is extend thinking, as opposed to having a channel into narrower, narrower conduits?

Tim: Well, the other I mean, the key is that sort of one fundamental sort of step that can be taken is recognizing what tasks to give AI to do and recognizing that means, intelligence. AI can be incredibly valuable because one of the things that human beings have biases. And so in a way in terms of decision making, they will begin to believe things that are actually against what the decision that needs to be, needs to be made. And so like, you can almost And then we almost sort of affirm that is that there are parts of this rash, this rational area for AI that allow us to sort of what are these rational things that we need to do when we start mixing the rational with the creative and so forth, we kind of mix these two things together, we should be focused on the interpreted the deciphering part, and up upping our game in terms of our deciphering our analysis ability, as opposed to taking what AI is giving us as analysis because it isn't, it's just an output of whatever might be flawed in our existing analysis and the input that we put into it. So we need to be a bit better at that.

Ross: So how do we get there? What I supposed to get better at, that's kind of a long analysis. 

Tim: It's a topic that is very much hot within the intelligence community, which is  , they have this thing called structured analytic technique, but nobody uses it. And the idea is that, if you have structure, I mean, the problem is, is that we have these  , the the intelligence community has a methodology of structuring analysis and being able to say, what are we doing, what are we sort of putting off to kind of machines and computational analysis and so forth. But then we kind of go back and say, Well, I trust my gut on this, and I'm just gonna go with my gut on this. And we have to now recognize that so much more, how much faster, things can go wrong. If we don't, if we don't slow it down, structure it. So structure, I would say Ross is critical, which goes back to I would say, culture mapping, which is essentially a way of structuring, structuring language to say if I were to if I, if you were to talk about any one concept people would say, Oh, well, I know what you mean, culture mapping is to say, wait a minute, no, there are many different other implied meanings to what you think that I'm talking about. And understanding that structure is important because then we start recognizing why certain things go wrong in society. And I could give you one example right now: we have an existential threat of climate change. And we have over the last 10 years developed programs like ESG, all these different kinds of language around this. And when we've done that, we've created a counterpoint. 

So right now, there is a, there is as much of an ideological movement against all of what we've created towards dealing with the sustainability issues that relate to climate change, and we can't, and you can't battle them, like there are people who fight against 15-minute cities, or there are people who voted fight against vaccinations there, I was just reading today anti-fluoridation is back in the United States, it's sort of like any science, any that the irrationality of human beings come because, and it's, it's not wrong, it's the fact that what you've done is you've led something that should be very structured in an emotional way you've packaged it, and you've expected that everybody would believe you, and everybody would come around, but actually, change happens socially. And you have to understand to be able to deal with the future is understanding the weight, all the different ways, things are going to change as externalities change all the different contexts as the context changes, and people are going to have different kinds of ideological responses, you need to be able to have a structure to say, I have some scenarios for how that what, how that might how that might particularly play out?  , we saw a lot of these particular signals in the recession. And then it was really clear, but nobody was paying attention to them. And so if they were there, and then the pandemic hits, and then it's like, then they're so obvious. Now they become obvious. But it's the structure that you need to say, how do we then put solutions to understand what you're dealing with? Because you're dealing with people and different groups of people. Everybody isn't the same. Everybody's not going to believe what you believe. And you have to deal with that kind of variability within society.

Ross: So pulling us towards collective intelligence or group intelligence. I suppose you're a part of the subtext here being Intel Jen says she says a social function is not far, far more than an individual function. And, with these kinds of existential challenges, or more and more complex challenges that we have, we do need to build collective intelligence that is superior to individual intelligence that has to be the path to our collective future. So in that, guys, what are we, this new intelligence that we have? How can we best build the best of humans, particularly human group intelligence, augmented by or supported by some of these new tools? Well, I think I mean,

Tim: For us, it's that it's being able to see those subcultures that are in that makeup, the society, we sort of think of society as a monolith, or we think of society as being demographically based. It's sort of divided by age, or it's divided by ethnicity, or divided by, it's divided by these cultures, these particular relationships to the subcultures, whether you're directly related to them or not. And for anything that's changing within, within society, there's always this point of culture mapping is that there's always this process between affirming codes of society. And then there's always this, this counterpoint that's always happening. So as language becomes sort of static, and becomes the rule and the law, there's always this counterpoint. So being able to understand and invest and understand what that response is not capitalized on, not commercialized it, but understand what it is, I'll give you an example that works that we did back a little bit over 10 years ago, and it was that  , looking at, at bicycles and cities, and  , bicycles as a machine was is was understood as being a leisure product that is sold everywhere. But, these subcultures of cycling that had been sort of living under the surface, were telling you how cities needed to be planned. They were telling you how they needed to function, how adaptable they needed, and not just sort of mobility but also things like food, it's sort of you start getting other codes of, of behavior that go with that. There are different researchers during the pandemic that studied skaters like skateboarders and made connections between understanding stents, understanding skateboarders, and how to help people age in place. 

One of the biggest issues globally is that we're living longer and that it's very difficult for people if they're living into their 80s and 90s to live in the home that they're in, because the city is the town that they're in, isn't planned that way. Do you understand? Adaptability? Do you understand? Do you understand how   how things need to function, you need to look at those parts of the culture that are telling you how things need to change, there was the same thing with our mobile devices.  , we didn't have VPNs. But some subcultures were telling us that privacy was an issue with technology while everything technology companies were telling us, what are you so worried about? Why can't we put a camera on everything? And why are you so freaked out about that? Well, subcultures were telling us, and even things like the right to repair. 

I mean, right why can I fix my foot? All of those things are there all the time, but we don't pay attention to them. And we have to understand, we have to sort of, in a way a collective intelligence embraces the full, the full range of what society is, and doesn't sort of, sort of force it to conform to the king of the model that we have, which is currently what we've done. I mean, demographics, sort of like everybody sort of  , fits within a certain box. We've had that model. Since then from the 20th century, it's sort of like it shapes polling and shapes, so many decisions that society mates, but it does, it's, we see it's giving us less and less good results.  , it sort of gets it wrong, more and more and more of the time why? Because people now don't  , they don't fit nicely in those boxes. And they and the speed of change is so fast it's like and how they're influenced by the range by which people are influenced so it  is so much broader because of technology. We have to understand the full breadth of society to be able to do that. That's what a living foresight model is. What is collective intelligence to me?

Ross: Well, I think, to your point, what I take from there is in a way the traditional framing of collective intelligence is you put a bunch of individuals together, and you architect ways in which together, they can be more intelligent. But it's a well, perhaps the units that you are working with are subcultures. And so you might have a group of people that think a particular way. And then another group of people think quite a different way, another group of people think it in completely different dimensions. Linking together those subcultures that represent a frame of the world, or a way of perceiving things or a way of sensemaking is bringing together those cultures out of which true collective intelligence can emerge, rather than looking at it as this aggregation of individuals.

Tim: Yes, I mean, I think we tend to think we think and we think in these boxes too much, I mean, for example, I mean take everybody's talking about the future of work right now. And they're really practical issues related to that because it's like, it comes down to what is an office for what, like, we have all this real estate that we suddenly the pandemic sort of whacked and you go, like, oh, well, what am I going to use that for? Oh, well, we're going through this nihilistic phase where we're going to force everybody to come back and we're going to surveil them, okay, good luck with that over, I'll give you how much time underneath the surface, I'm telling you, there are these other ways in which people are sharing intelligence and solving problems. It's why I like there are different I mean, it's before I've been talking about this, there, there are many different kinds of companies and that has sort of tapped into the intelligence within, within video games, for example, because and how people, and  , and they've even sort of brought that into and brought that into how tasks and sort of problem-solving is done within a company, but you start dealing with, those are the essential issues of dealing with even more abstract issues, like the relationship between physical space and virtual space and realizing there is no, there isn't physical or virtual. Now we're dealing with this emergence of something called fourth space, which is this, where we're digital and physical at the same time, and who understands that, first, who could give me a framework for that? 

Well, I need to be able to tap into those particular groups because that's going to tell me how to make what an office should be, it's not going to look anything like what we currently have, it may be a park, or it might be a, it might be a mall, or it might be   because people are going to are, what do they do when they go to work, they communicate, they and more and more of work is becoming kind of a grazing, more than it is sort of the idea of a meeting, we sort of do shorter, it's shorter kind of creative kind of conversations, and then we go and do our tasks. One of the things that the pandemic taught everybody is what the hell is nine to five? Why do I have to work five days a week, if I can get all my work done in two days, or whatever. So the idea of time has changed. So used to getting all of these kinds of concepts are constantly changing, and society, that we're not keeping up with the cultures that define what that meaning is. And the subcultures are those groups that are ahead, and we need to sort of understand because then the rest of society pulls that in like they did with privacy because it's sort of like they the average person didn't understand that privacy is very abstract, but they kind of go, who has this, who's ahead on this, and they start pulling in those behaviors, and then they start becoming normalized, and they become habit, and habit becomes culture. And that's the issue. So studying that kind of relationship between what is the general culture and subculture is what's, really, really, really critical.

Ross: So is there anything which you would finish off with as advice or suggestions for listeners based on what your work or what you're seeing or what you do?

Tim: Well, I would I would say that there's a lot of there's a lot of opportunity within with generative AI and I mean,I also teach a I've been teaching a course in in trend analysis for going on 20 years now, and which I have integrated generative AI. But it's recognizing how we can integrate these tools that so we do not repeat lacing what we do so in a way we should be what we should be right now, what I'm hopeful for is that there's a great opportunity for kind of a renaissance in, in, in education and sort of defining what kind of skills that we need. And I think that I think these tools can be incredibly valuable in doing that. So I would say, like, recognize kind of what the potential is, and don't forget that we should be raising the bar as human beings in terms of what we consider to be intelligence, what we consider to be creativity at this particular moment. Yes.

Ross: I 100% agree. So where can people find out more about your work, Tim?

Tim: You can go to scenario dna.com. I have a blog that's related to a class that I teach called analyzing trends.com, which I have not posted as much lately, but there is that as well.

Ross: Fantastic. Thanks so much for your insights and all of the work you do. Oh, great.

Tim: It's great talking to you, Ross.

The post Tim Stock on culture mapping, the culture of generative AI, intelligence as a social function, and learning from subcultures (AC Ep44) appeared first on Humans + AI.

– Ufuk Tarhan

Ufuk Tarhan is a prominent futurist, economist, keynote speaker, author, and CEO of digital agency M-GEN. She has worked as a senior executive and board member in a number of prominent technology companies. She is author of two successful books on the future and has received numerous awards including Most Successful Innovative Business Book Award, Most Successful Businesswoman In IT, and various lists of top social media influencers, and was the first female president of the Turkish Futurists Association.

Website: www.ufuktarhan.com

LinkedIn: Ufuk Tarhan

Books

Ross Dawson: Ufuk, it's a delight to have you on the show.

Ufuk Tarhan: Thank you. It's my pleasure to see you again and to hear you again.

Ross: I think the concept of amplifying cognition is central to your work. You've described to me this concept of T-human, and love to hear this concept and how you've shaped that and applied that in your work.

Ufuk: Yeah, thank you. And you were one of the very first ones who picked it up. I'm so happy to explain it. T indeed, I was aware of T-shaped skills. At one of IBM's conferences, I heard that for the first time, many years ago, more than maybe 20 years ago. Afterward, in years, I transform it into a model, a personal transformational model, to adapt ourselves to the needs of the future. And the first application is, of course, made on me. Because I've worked in the IT industry for more than 20 years, as a top manager or CEO. After 20 years, more than 20 years, I decided to change myself, and I decided to reshape my career, my life, and everything. While doing that, at the core, there were future, futuristic studies and thinking about the future more and more, and the technology. I decided to give consultancy services to people and corporations, to teach them or to let them be aware and apply future planning effectively. But at that time, I was a single mother, I was working in a very high-level company, and I needed to earn the money I needed. I couldn't leave the job immediately. So I needed resources. Then I tried to find a way to develop my knowledge about future studies so that I could form my own consultancy company and give consultancy services. 

I remember during University times, I was waking up at 3 am to study for exams. I said that I could do it again, maybe and I could do it. I started to wake up at 3 am three years ago. And then I worked on my today's knowledge or future studies to increase my knowledge in that area. I was going to work, my daily work and I was a CEO at that time. I was working very seriously Of course, and I was coming back and at 3 pm I was working as a futurist, etc. So I realized that it was a hybrid mood indeed. I had to run to life altogether, the future life, I was preparing my future version. At the same time, I was working on my actual work. I just decided that there should be hybrid moods for everybody. Because we cannot quit our ongoing responsibilities and jobs, we need to earn money or we have other responsibilities. So we have to find a way to run them together. And while I was doing that, I realized that I have to learn so many new things, so many so many new things. I discovered this autodidact learning technique. And, I saw that I'm learning everything almost by myself. I'm digging into every source to get more information, knowledge, etc. So, I said that this is an autodidact, learning it is mandatory for everyone in the world right now because we all of us have to transform ourselves. And we have to create a new version of ourselves. So that's autodidact, that's another mandatory thing to learn and to apply. 

And then while I was doing that, I realized it again. I deducted many things from my life. I put many things out of my life, people, habits, time, everything, and, I sold them and became a perfect trader, I was the creator of my own life. So I put all this together. I said that to be able to have a successful, successful, sustainable job because I'm mostly concentrating on the sustainable job, topic, and area. And because every one of us needs to work and have a job not only for earning money, but it's life. So I said that whoever wants to have a sustainable courier life, and job life should apply this T shape including this hack, hybrid autodidact, and the creator moods. And I put all of them together and that T-shape model came out. Just

Ross: Just reflecting a little bit on what you've just done. I've always thought of myself as an autodidact. I have a reasonable amount of formal studies and some postgraduate studies. But essentially, I've taught myself almost everything over my guitar, I taught myself guitar by having guitar and just working it all out and taught myself other instruments to teach myself, and the vast majority of what I've learned, I've taught myself, and I think that's important. And as you think about this wonderful framework around the hybrid, in the sense of, yes, you do need to be continuing the work, which you are doing now, but at the same time to be renewing yourself. And that's, of course, a frame for organizational leaders, as well, in the sense of saying, we need to look at our sustainable current business model, we also need to be building new business models. 

I think the same applies very much to individuals. And this, you know, the idea of curating as you were saying, selecting, I think is a very important framework. But one thing that sort of, I suppose the question that arises out of that is, all of these take intention, it takes people to take control. Of course, I suppose in a way, what you're suggesting is that people need to be making their own decisions to choose what to learn to teach themselves to be able to make those selections to be able to continue to renew themselves. 

Ufuk: For sure, for sure. I cannot imagine any other way to be successful or to satisfy yourself that you're doing something useful for humanity or any society or anyone. So I think it's mandatory to have the responsibility of choosing.

Ross: So you're saying that, so going back to the T-shaped bottle with this, these elements of the hybrid learning they ordered Act and the curation. So what are some of the other frames that you'd put around this T model for us to renew ourselves in a time of change?

Ufuk: Thank you. I'm very happy to talk to you because I don't understand why people are still not so aware of that model, intentionally. So it's very practical. The first inspiration for why I worked in that T-shape model after transforming myself, of course, I was not saying at the beginning of my transformation period, oh, now I'm hype, I'm in a hybrid mode or no, I'm learning as an autodidact or so on. So I trained later, ah, that was this and that was it and the model came out. 

The inspiration of this, of creating that model was that everyone is saying the world is changing. The future will be such that and the future is not a secret anymore. The future is knowledge, like history. If you have to know about our past, the future is history. To create a better future. We also have to know about the future and that is knowledge so everybody first should accept that or become aware of that. So that was my first point. So, with the help of social media and this internet connection, more connectivity everyone is veering off about the future even the little kids know what will happen in the future in every area almost. So, people feel that people feel that they have to change themselves. they have to adapt themselves. Yes, that is no I accept the future will be like that, and I have to change myself, but how somebody should tell me, what I do to transform myself, in your book as you mentioned, we are overloaded with everything, especially with the information. So, among all this information, the fast-moving phase and everything is fast. Everything is too much and I have to change myself. I have many responsibilities, and how am I going to do this? Someone should tell me. And I started that, at that point. And I said that I would tell people how I transformed myself and they should apply this model template to themselves as well. That's the starting point of T-human and T refers to of course, the golden ratio of the people who human who have sustainable job, sustainable career.  T means on the vertical axis, we have to go deep and deep in one area for instance, its future for me, future studies, especially for business life for work-life, for career. 

On the horizontal axis, we have to use this knowledge this expertise in every area, we have to merge this knowledge to industries to areas, and everywhere in any condition. For instance, I can work with organizations in the automotive industry, fashion, education, whatever, whatever it is, so the future of everything, future of anything.

In the vertical axis, we have to go deep and deep, it's endless work, endless learning and experience and on the horizontal, we have to connect this in all areas to all areas. This is the shape and the understanding, but to be a strong T we need other components, other layers and it could it has three Ts inside. That means to be a strong T-shaped human, we need to be tech-savvy, we need to use technology very very strongly and deeply. We have to be very good team players. But in this team, of course, I don't mean only humans. I also mean the digital facilitator, robots, artificial intelligence whatever it is. Team for me forming is not consist of only humans and also we have to be a transformative or design-thinker, Tinker person. We have to have these three Ts inside this main T and once we have this we have to advance our competencies by using this hack model – hybrid, autodidact and curation, we have to make this stronger and stronger by using this hybrid auto deduct and other Ts curation. 

So, this is a complete framework, which brings us to a point where we have a very successful sustainable job because we are transforming ourselves in a very disciplined way and very concentrated way so that we become a people having 5Cs that means we have capability of doing something; we are becoming competent on this area by using 3T and hack. They are certified by society by, by people, by corporates saying that, ‘Oh, you know this, you are certified.’ Once you have been certified and authorized in one area, that brings you high responsibility for creating more and more new things that make you creative. Once you make this you are becoming a changer, a game-changer, which means you are creating new things, which is the most required condition according to Darwin's survival. 

It says that survival depends not on strength or intelligence, but on adaptability to change. So this shapes this frame, and this, whatever you call this template is served to people, if you become a successful, sustainable job player, then apply this to yourself. That's it. I am suggesting and modeling a concrete model. People who are asking ‘what shall I do?’ You do this to yourself, you reshape your future yourself.

Ross: I think it's a very, very strong and very useful model for a lot of people. So I'd like to dig into a few things there. And one of the ones which perhaps not surprisingly, stands out to me is that of the team player in the sense of, as you say, team player, not only with other humans and machines. Now, machine capabilities are advancing very fast. So, the nature of how we are a good team player changes. So how can we think about being a good or better team player with technologies while AI, for example, continues to progress?

Ufuk: Thank you. It is a very sophisticated question indeed. I was thinking what will he ask me to reply or will he want to discuss? Of course I expected such a question and I thought and I found this today and I'm using it for the first time here. I saw that we live in an LS fondue land because we are realizing that ‘yes we are humans we are the strongest species on earth we think so.’ But now we're much more stronger than ever because we are creating artificial intelligence, robotics, everything — even if we are weak, we use those strong team plays let's say so. We are in Wonderland. We can do everything deep fakes and other things — whatever we want we can create and I thought that we are in human-IT Technoland.  Technoland is the rabbit hole with AI, so it makes everything so it is a very oxymoron time. I think we can use this metaphor of Alice in Wonderland and humanity technoland. 

I like to play with letters and words as Alice's Human-IT. Yeah and Technoland is Wonderland and the artificial intelligence is the rabbit itself. So when we think about this, if we call the story itself, we are in the same situation. But if we don't know what to do, we will not be able to respond to this question. Because we know that it has many sides – bad and good. We are living in an oxymoron world — everything is all together and we are in an eclectic world. We are putting on top of it another thing: we mobile, we stand here, and so on. So it's such a complicated and complex stage right now. 

Indeed, I don't agree. Yes, we are saying that ‘we are in a very fast-forward mood and everything is so fast.’ No, it's not fast enough. Not even fast, because we are in standby mode. We know what we can do. But to be able to activate all these things and use them everywhere; first, we need to solve this green energy problem, then we have to have the internet faster and achievable every time.Then we have to create this blockchain web three environment, infrastructure to be able to operate on that. Tthen we have to change the finance system from scratch by putting those crypto assets or whatever it is. Then unless we change or we have these sources, these infrastructural elements, it is impossible to get faster. So we are just in Wonderland, and we are rushing around. We are just assuming that they are not reality and AI is at its very primitive level. These are not AIs yet. Oh, we are scared, of course, we have to be scared of the ethics and the other things that can happen. If we don't structure them correctly. We can imagine everything, but we are not acting yet. Because we don't have enough sources. At the top of it, it is energy. We don't have energy. We don't have an internet connection yet. We don't have blockchain web three and crypto assets. And there is a third war in the world. But that is not a war among Russia, the United States, Ukraine, or the other parts of the world. It is the war of the old system, the war between the old system and the new system — who are the old, we know, who are the news, the tech giants, the youngsters who are protesting the way of work, let's call it capitalism or an ongoing system, and they are trying to create a new way. And I call it sustainable capitalism. 

So, you asked me to integrate this human and technology? Yes, we are trying to integrate ourselves. So they and that's for sure we can do it. And we started to work on that. But we are not even at the starting point yet. Let's solve these listed resources and infrastructure requirements.

Ross: Fabulous. So, just the first thing that strikes me about the Alice in Wonderland metaphor is that Alice went down the rabbit hole because she was curious. So it was a story of curiosity. And in going into this marvelous land where we are discovering and learning the unknown. So perhaps that's part of what we require as well today.

Ufuk: Yeah, yeah, for sure. We are wondering, really what will happen and we are, of course, at the change phase. The thing is that it is fast, but it is not . We are just in a loop and we are just waiting by working and rushing and scaring till we solve this problem. This is at the top of this energy problem indeed. That's the major major problem of humanity. 

What I was thinking about this integration…I read a book years ago, most probably you’ve also read the books of him, the name was Juan Enriquez, I think and he was mentioning the future chasing you or something like that the book name and he was describing he was analyzing did power always enhance of societies who create the alphabet because the alphabet is transforming the abstract to the real world. The first alphabet was in caves. And then this specialized alphabet was created by the Chinese. And at those times the power was in Chinese society or in that region, and then the Latin alphabet came with ABC. We have seen who had the power with that because the power of distributing you're collecting the information. And then, the ongoing alphabet is 101010, a digital one binary, who has this binary alphabet will have the power to transform or to govern the world. And the last, but not least, I guess, the last alphabet is synthetic biology — alphabet genetics, the A, P, C, G, whoever designs this in the A, P, C, G, alphabet, that means get gene editing. And China is declaring that we will be the gene editor, state of the world, the leader of the world. So the last alphabet is not binary, the binary is getting transformed to qubits, which is again another alphabet. This synthetic alphabet will be qubit and will be formed by qubits. And the biological alphabet ATCG will be formed by ATCG. 

So, we are just entering the era of not just simply merging humans and machines or robots for artificial intelligence which is much more serious than that. We are indeed just at the gate of creating a new kind of species, that kind of thing, maybe. We just discovering this stuff? And that's the one I think we should concentrate on while thinking about any AI G and general, artificial intelligence, attics, and so on. So we are squeezing our attention into very closed areas. But we have to open our eyes and open our minds a little bit larger, wider than this. So when I think about this integration, merging, convergence, synthetic information, and logic, I see some other big, big things in front of us.

Ross: Yes, indeed. And so if we think about cognition as making sense of the world, taking in information makes sense of that. So humans and some animals have done that. Well, now AI and as you point to it, it is, what do you call transhuman or certainly beyond, current existing species, which will be doing that in quite different ways.

Ufuk: Yeah, yeah. That's so exciting. Indeed. Although we are exaggerating our excitement without thinking about all these important topics. Anyhow, we are going forward and we're I like this oxymoron very much. For instance, when we say virtual reality, it is itself an oxymoron. And when we say artificial intelligence, it is again, a very big oxymoron. So I repeat this again and again because we need some meeting points. We are under stress to catch up on something, and we're in really scary mood most of the time. FOMO is our, I think the major illness for all of us. So we have to think that, yeah, we're oxymorons, we have to live in a physical environment and we accept it. And somehow, maybe at some point, we have to accept and we have to be in a more steady mood. And the reason that latin phrase about this, Festina Lente, hurry up slowly. We have to be in that mood, hurry up slowly will help us to go a little bit imbalanced mood, otherwise, it's so difficult to maintain healthy.

Ross: I think that's a wonderful note on which to. So how can people find out more about your work?

Ufuk: My name is Ufuk Tarhan, they will see that in the description part. When they put ‘.com’ at the end, it is the place where I share everything, but in Turkish mostly because English literacy is very, very poor in my country. That's another starting point, I said that there is information about the future, and almost all sources are in English, and my people are not able to read them or reach them. 

I started to create or produce the content in Turkish. But now the translation of the AIs helps people to read everything, in every language. So I can start to create the content in English. But in any case, even if they are Turkish, they are translated and vice versa. So that's my address..

Ross: Fantastic! Thank you so much for your time and your insights. Folk. It's been a wonderful pleasure speaking to you.

Ufuk: Thank you very much. It was a big pleasure for me. 

The post Ufuk Tarhan on the T-Human model, being an autodidact, oxymoronic technologies, and teaming with humans and AI (AC Ep43) appeared first on Humans + AI.

– Shikoh Gitau

Shikoh Gitau is CEO of Qhala, a digital innovation company with clients across Africa. She was previously head of Safaricom Alpha, the first corporate innovation hub in Africa and worked for African Development Bank helping governments adopt information technologies. Her numerous awards include being the first African to win the Google Anita Borg Memorial Scholarship, and Africa's Most Influential Women in Business and Government, Technology. She sits on numerous boards and holds a Ph.D. in computer science.

Website: Shikoh Gitau

LinkedIn: Shikoh Gitau

Twitter: @DrShikoh

AI (Artificial Intelligence)

M-PESA

Mobile Money

Wall Street Journal

The Economist

The New York Stock Exchange

The Pathology Network (TPN)

Gemini

Ross Dawson: Shikoh, it's wonderful to have you on the show.

Shikoh Gitau: It is wonderful to be here after going through every other challenge, but we are here now.

Ross: So you have spent all of your career amplifying people with technology. I would love to just hear your perspectives on how it is we can amplify humanity, and amplify ourselves.

Shikoh: I love the word ‘amplify’ because it sets a very good tone for this conversation. So one of my mentors Kentaro Toyama wrote a book at the very beginning of my career. And I remember him giving the talk before he did the book. And he kept saying that technology is an amplifier of human intent. At that time, he was a Senior Director at Microsoft Research in India. And his goal for going to India was to help Microsoft build these technologies to enable human flourishing. I think after years of doing this, he realized that technology builds technology so much, so much to do something, but eventually amplifies a human act, a human intent, a human habit. And that's what I love. I love this conversation because it set me on my career path. And my career path is you have to. I started looking inside how technology amplifies my intent. I want to be able to change the world. I want to be able to increase thriving and economic emancipation in Africa, how's technology going to ‘Hey, help me achieve those goals.’

But more importantly, how is technology going to help other people around me and on the African continent to be more specific, be able to achieve their own goals? And that is how I got my career started in technology. So it was very interesting when I saw this. I'm thinking oh, amplifying cognition is part of human humanity and humaneness. For me, that is how I'm jumping into this looking at it from like, not just like an AI perspective, because AI is just another technology. And when I say that some people take it personally, I've been working in technology. I've gone through so many fats and buzzwords and hypes of technology. So I know AI. Well, it is a significant technology, it is one of the other technologies. For me, I feel like one of the technologies in Africa is a mobile, mobile phone. The mobile phone did change our lives. Yeah, to be totally honest. It changed how Africa works. And if I was to choose between, like, we are back to whatever Dark Ages and I was to choose between AI and mobile devices, I'll always choose mobile devices. So I've seen this hype, I've seen it happen. And I've seen the amplification part of it. So I am, I am riding the hype, but I am very conscious that it is just amplifying what we as human beings want to achieve. 

Ross: Yeah, hello, I love what you're saying, typically around this idea of intent. That's the first thing that really struck me about generative AI is that what it doesn't have is intent. That's what humans have intent on. And I think this point around you saying that the mobile phones, essentially Africa leapfrogged. So it led to mobile payments because it had the mobiles. And that's what people had. And so it did lead the world and these technologies. I’m interested in thinking about other things with any other technologies now, where Africa could leapfrog in the same way that it did with the applications of mobile phones.

Shikoh: So specifically taking Mobile Money, right? We always say it’s like a nice cliche that always says that a necessity is the mother of invention, right? So for us, having Mobile Money was not innovative. It was not, innovating for the sake of being on the Wall Street Journal, The Economist, The Times, or being listed in the New York Stock Exchange, because that's many of the founders when you meet a founder, in Silicon Valley in New York, in Florida these days. You will find somebody just wants to be listed and their goal is to be able to create this company that is then listed on the New York Stock Exchange. That was not the intention. We intended to solve a very painful problem. At the time M-PESA. was coming in. That was 1997. We only had a 2% penetration rate in mobile, mobile, and mobile financial services, that is somebody who has a bank account, people who are able to save, people who are able to access credit, people who are able to access insurance, we're just a measly 2%. And those are the 2% that were employed with formal jobs. Right now we lead the world with, like, 98% Mobile Money, we flipped the numbers. Why? Because my mobile phone is my bank account.

Every time I go to the US, at least in the last two years, I've seen things changing in the US and Europe. So every time I go to the US, I'll just be carrying my mobile phone around. And every time I needed to pay, I'd remove my phone, and I realized, Oh, my God, they don't have M-PESA. Here, I need to go and find my card. So I had to carry cash and cards. I don't carry cash in Kenya, why? My Mobile Money, my mobile phone is my Mobile Money. And that's how we reflect. So in the last two to three years after COVID, especially when I saw the larger adoption of mobile payments in the US and Europe, I realized that we've been experiencing this for more than a decade as there is no surprise here, you're talking to us. And in the same sort, again, amplifying human intent. Our intent was to solve this very painful problem around financial services access. In the same way, I strongly feel and after saying, I mean, I have been in Europe for the last couple of weeks, I realized that we are going to replug even in AI because you see there is no urgency in Europe to adopt AI. Zero, like I was thinking, zero urgency in all this conversation. Everybody's like, Yeah, but things work. Why should we make them faster? What I mean, is that things actually do work for them. There is no need for interest, what is it called? Efficiency, because efficiency is there. Like AI is for the many people that I was speaking to another additional chore.

For us, AI is a necessity. It's the difference between life and death. And always give this example if we work with one of our startups called TPN, The Pathology Network. Let me give you and let me just put it in context. There are 3000 pathologists on the African continent, which is 1.5 billion people. Do you do the numbers in terms of ratio? Just do the numbers in terms of ratio, in terms of GP, this one to 3500 people, one doctor to 3500 people. It should be one to 50 people, I'm just putting it in context for you. So when AI is coming to bridge the gap of I'm a pathologist, I'm able to solve this by uploading pictures online, getting that quick initial diagnosis, getting connected to the right medicine or the right treatment plan. We will use it because we are solving a problem that is actually killing us. Yeah. When you're being told, if you're able to install this an A&P and see and follow the following instructions listen out, check out for these take a picture uploaded of your what is it called of your plants and to see if they are doing well if there is a disease, what to project to how much production utility of land is going to have? You're going to use it, you're going to use it for your kids. I hear everybody in the US call screen time. I'm saying screen time is the only time my child can be able to access this information when I'm at home on my phone, I'm going to give them my phone to go and learn.

We are adopting this technology to be able to take us to that place where everybody in the world is about efficiency and increasing productivity. It is nice to have. It's a chore. I'm using the word chore because I remember somebody saying you know AI is just another technology etc. Things already work here. And I'm thinking to myself, ask us if we're using it to replug ourselves. So the world is still figuring it out. Africa is going to lead in AI. I am more than confident of that because I have our team that is actually working really hard to solve some of those problems that the foundational problems and our meet. I'm meeting all these amazing innovators across the continent who are saying if we have this, we will be able to solve this. So our goal is to be able to work with partners to solve this, the foundational aspects of it. So I know like, let the world continue fighting about regulation, fighting about what is the role of AI? Is it going to take up humanity, we are going to show them how it's going to be used to solve our problems.

Ross: That is awesome. That is absolutely fantastic. I can just imagine, as you say, the scope, you know, all, as you say, relatively speaking on the margins in the sort of the highly developed nations, but in terms of the incredible difference that these tools can make, and you're just giving me a fantastic example there. It's staggering, in terms of the potential, so very, very excited to see that and how that can be applied at scale. So one of the impacts, and somewhere do you look at as in terms of sovereignty, and technology, particularly given that has been dominated by big tech, which, because it's big, it dominates? It takes power from us in many ways, I suppose. And whereas there's, of course, technology has always had the potential to give power to individuals, it says somehow, we often haven't taken that. And so perhaps those again, Africa can lead and point away to where the individual can be the leader. In a world where technology holds the balance of power.

Shikoh: Yeah, when I think about sovereignty, I think about sovereignty of the individual from technology, but also think of sovereignty of geographies and entities, apart from others, they mean, loosely defining a serenity, it means the ability to define your own future and your own destiny very loosely, like, if you read all the definition, it goes to that. It's that ability and capability to be able to define that right? Sometimes that ability is taken away from you. Why? Because somebody else somewhere else is sitting trying to define what your future looks like. Sovereignty means that I need to be in charge of my destiny, I need to be able to be in charge of my future. Yeah. And that means being able to understand the context that you're operating in. Yeah, and not letting other people define your context for you. And that's part of the foundational work we've been talking about on AI in Africa is saying, Guys, when you go to Europe, they're saying, oh, we need to regulate, we need to regulate, we need to protect ourselves, we need to do this. But when you go there, everything is working for them. Everything works. Yeah, they have a super supercomputer, they can switch on a flip chart as a flip switch, right? They have been collecting data for ages, our data has already been moved from one person to the other. Yeah, they have talent, they have mechanisms and processes that they are working on, and they have the luxury to start talking about regulation. And for them, as I mentioned, it is just an updated technology. It's not for many of the people that we spoke to, it's not adding anything, it's not adding any further efficiency to what they do. The things already work for them. The pace might meet, maybe wanting, but it's still working for them, they don't need to do any other thing. But for us, we need to first challenge ourselves, which means that we have to define many of these things for ourselves, which means we have to define things for ourselves. Because for them, they are in no hurry to make anything work because things already work. Things are not working for us. And they don't understand that things are not working for us. And that's the whole sovereignty part is understanding the context and defining it for yourself, and defining that future for yourself. And that is the ability and capability of doing yourselves. 

So it is building these capabilities on the continent to be able to do this for ourselves to be able to define and innovate for ourselves because we understand our problems further. If I went to Europe, and I told them, I have to Uber for pathologists. They'll be looking at me and thinking what do you mean, you have to Uber for a pathologist? I'm saying this because there's only one pathology in my whole county of 10 million people. And they're like, what, what? Exactly because it is not something that occurs to them. I cannot just walk into a medical center and get health care, then they have a right to get health care and then fight to the state later for payments. We don't have that luxury not because a state cannot pay, we don't have the doctors, right? So we have to be able to see that the syringe I'm thinking about is the serenity of mind and mindset. Knowing that we are solving for ourselves, we are solving for things that are very, very African. Yeah. And other people are coming from totally different contexts. My totally different place, and their idea of the world. And the less they're looking at the world is very, very different from us. And then accepting and acknowledging that the lens we look at the world is extremely different from theirs.

Ross: There's lots I want to dig into there. So first things you talked about, you mentioned the luxury of regulation. And I think you've just flown back from Europe, where there's got the most intense regulations around AI and technology and data and so on. So you're suggesting that Africa can flourish better where there is less regulation? Because some of the regulation is, of course, trying to avoid over-concentration of power? When you look at regulation, or its potential role in allowing Africa to flourish through technology, how do you envisage that?

Shikoh: To be totally honest, after being in Europe, I've stopped hating on them, because I was just thinking, Why are you pushing this down our throat, literally, but being there and seeing for them, there is no, like, whether AI works or not, whether it's regulated or not, it will not affect their life. Yeah, so I totally get it from them like, we have the luxury of saying, let's regulate everything, completely. And then let's try it out slowly by slowly if it works for us. Let's try everything and see what works for us. Not if it's what works for us, which are two distinct things. Yeah, for them it is if it works for us, and that's well and good. If it doesn't work for us, that's well and good. It will not affect them. For us. Whatever works for us, works for us, gives us a big step change. And the difference between If and What is huge. It's miles apart, because, for us, we are looking at how we can innovate around this technology to meet and close our gaps. Yeah, for them. They're like saying, how can we use this technology for efficiency to just make life a little bit better? But if it does not exist, it is fine. It does not exist, we don't care. And for me, that is the difference. And for me, my mind shift was changed from actually not needing AI to survive. We do need AI in the same way. I mean, in the same way, they fought everything from mobile phones to cloud computing, everything. We don't have enough computers, we will rely on cloud computing, right? We don't have a computer in every school. We rely on mobile phones for that, right? We don't have teachers, we don't have these things. So we have to rely on innovation. Yeah. We have to constantly and consistently be innovating around technology. they are just small, you know, in innovation, they say they're these dramatic big innovations. And these, what's the word for that innovation that we call it, it is called Step CI, that's the top step change innovation. If you read the book by Christian Christensen, you will see the differences in innovation. For most of Europe, it's like small, small innovation, that is helping them just a little bit. It is not 10% incremental in terms of difference for us. It's a 500% difference if this thing works. And for us, it's what can you do for us? Not if it can do for us? 

Ross: Yeah, one of the things you mentioned there is around education. And I think that's yeah, that's one of the, as you say, completely transformative in quality education personalized to all African youth. That's that, you know, the impact of that is absolutely incredible. But one thing you said earlier was about somebody the fact of the unique African way of thinking from mine, which is very distinct, and for many reasons for the rest of the world, and without trying to define it because of that. That's too big a question. So well, how does that inform your vision of Africa's potential? So what is the as we see Africa as the continent of extraordinary potential for so many reasons today? How does the uniquely African perspective way of thinking and way of shaping the potential of the vision for what Africa can become in the coming years?

Shikoh: To explain how Africa is different I always give this example because people don't understand these until they are able to be practical in how they do. So in the US, when you put Siri or Google Drive or Google, any of these mapping technologies, they only say, drive straight, turn left, turn right, go straight, turn left and right. We agree on that, right? There are places on the continent, where people will say, walk north, then turn south. Yeah, the other people will take a walk straight up, when you see a tree of this kind, tan on your left. Those are different ways of thinking, right? And what happens is that we are put in a box to turn left and turn right. Every time somebody says turn right, I have to lift my hand and say which hand I used to write. And then that is my right hand. And then the other one is left. So that's how I in my head, I'm able to figure right and left. It's not automatic in my head. For many people, it's left and right is very automatic in their heads, right? And now you can imagine, when, as a community, we are a very community-based communal culture across Africa. It's not a Kenya thing. It's not a South African thing across cultures, we have a term for it called a Nguni Bantu, but it's everywhere. Ubuntu is everyone's continent where we are because you are, there's no individuality, which is another condition around AI that I always argue against, right? It's about not budgeting time around us as a community. 

So when I'm giving directions for somebody to go somewhere else they go when you get you when you see this house, that is so and so's tree, so actually is named after a person, then you turn left, then you see so and so's bridge, then you cross the bridge. So we've personalized all these things around our community, around, our heritage, right? And that is what we are bringing to the world, right? And you're bringing this idea of like, my humanity is not based necessarily on me as an individual or my intellect. So every time I ask my friends about having costumes with my US friends, they say, You're not scared about AI, I'm saying no. Why? Why? Why should I be scared about AI, it's going to take away our ability to be like, unique, and individual. Saying my uniqueness is not formed by my intellectuality, we recognize that there are other intellectual beings in the universe. As part of our growing up. We were taught that intellectual beings can be innate and animate beings. Yeah, we are taught that in many, many African countries, not just Africa. I mean, like, even in Asia, they have a lot of these beliefs that our intellectuality does not make us unique. Our ability to think and reason is not just unique to human beings, what makes us part of a community is that that is what makes us human, our ability to talk to each other, to be able to have compassion and have kindness, you know, to each other is what makes us as a humanity. But when you go to the US when you go to Europe, everybody's by themselves, they go to their small apartment, they don't know who their neighbor is, as good as every like every holiday like those are two holidays these last few days they eat, whether you're Muslim or not, you're partying.

During the Christmas holidays, everybody's partying with their neighbors. Why? Because that's how we were brought up in the community and I'm saying not putting up. I'm going to my neighbors if the whole community comes together, we bring our food together and have a really good time. Right? And for us that communal way of thinking that I don't think of as myself alone, I think of myself in the context of other people is what makes a difference. Yeah. When I think about sheep, I don't think of sheep as an individual. I think about Chico within the context of my family and the village I came from. Yeah, so anything I do impacts my village and everything in my village does impact me. It is two-way, it's a two-way conversation and people don't understand that. We have deep roots here. We have a deep association with ourselves and that is something that does not translate into many of the AI models we are pushing like you know paper I don't if you don't read our paper, we have something you're calling Data Sets and Data Systems. Data sets are what the world knows. Data Systems is how that data is being used in our context. And that is what is important for us not to lose sight of what Africa is about. 

Ross: So I mean, that goes to, I suppose to the point that mobile technology being such a transformative tool for Africa, and for many, many reasons, but including the fact the reality is that families are different places, and people are connected, and so to connect, and that's a very obvious supportive community. So I'd love to hear how you see AI, being able to amplify community or the relationship between how AI in an African context might be, you know, relates to the reality of community and the ability to support and to grow, grow the community.

Shikoh: So let me backtrack. So I feel old when I say this. I was amongst the first five people at most, who studied the impact of the internet on the continent through mobile phones. So my research actually my PhD research was around, what the internet looks like for the billions of people on the continent on their mobile phones. That was my world. My whole dissertation was about the very unknown days of the Internet or the mobile Internet on the continent. So it was very, very, very early days, no smartphones yet, right? And in the middle of doing my research, Facebook became publicly accepted, acceptable on the continent. So it stopped being like a university only. Products become and everyone can be able to access Facebook products. At that time, I saw a switch even in my research, because at that time, the internet, Facebook became the internet. Right. And that equating Facebook to the internet changed. Why? Because I'm able to connect like I was not in the country. So, people who I have not seen for many years while coming onto Facebook, I was able to connect to them and link back to my childhood, link back to many, many things, right? Think of it from that point of view. So Facebook became the internet for many, many African people. Yeah, so we have to credit Facebook for that. And then it became even better with WhatsApp. Yeah. Now I am able to create these tight-knit communities within WhatsApp. Yeah. And for many, many, even my grandmother, my mother, everybody, my extended family is on we have a WhatsApp group for the extended family and magazines and every level of community that you can think about to my siblings, right? And that has connected us as families like people, we will I mean, without the internet, who never would have lost love made lose if they like I wouldn't have traveled out of the country. I couldn't connect with my cousins. Now. We were very, very tight. We could talk every day until midnight. It's a nice, nice community and family is what is called family is not child on these platforms, right? When I think about AI, AI has the ability to do that even better. Yeah, the ability to engage a baby's ability to enable people to know patterns, not being able to connect, find, find help for each other, close family for us is not only for connecting but also finding help and supporting each other in very, very dark times. But also like I'm having a baby. Does anybody want to come and sit with me for the next two weeks? Right? So that is where we are, we are looking at the internet being an intricate part of us, again, amplifying our intent to be a community and helping us critical communities across the board. Yeah, and for me, that is what is critical is being able to create those bonds, using agents using understanding using our understanding about each other, like agents understanding all the single each other and be able to notify us if something is not happening or not being able to seek for health help, both in health, financial education, any type of help that somebody is able to do that. But most importantly, helping us bond better. Yeah. Because once I have a better understanding, I'm able to bond better. 

Ross: Fantastic. So to finish up, just like to get a few words from you on the potential for amplifying the humanity of Africa. So you know, I think that's your mission with extraordinary other people on that journey with you. And you know, just love what were Where could this go? What is that? What is that vision for how Africa's beautiful humanity can be amplified to the fullest?

Shikoh: Before I can amplify our humanity, AI needs to accept and acknowledge that Africa actually exists. Yeah, we just wrote an op-ed a few weeks ago, where like, our hook was the Gemini debacle. Right. And everybody liked it, it was hilarious for me. Because as an African woman, and uneducated when I was in March, I've been erased. Yeah, over and over again, I do not exist, the number of times I receive an email saying, Dear Mr. Gitau, all of these things, because nobody bothered to Google and find out that I'm a woman, right? So it was very hilarious to see these threads upon threads of conversation around Gemini racing. White men, because it is post-primary around white men more than anything. And as I was telling my friends at Google, you need to give the guy who made that bag and give them a race, because, they brought to life, what we experience every day. Yeah, but when it's flipped on the other side, then yeah, it is actually quite painful, right? And we need to be able to acknowledge Yes, that it was a bug. And or not, I don't know, I only see some bugs that are normal technology. So I understand it cannot be a bag. But for me, what's exciting about that is being able to showcase that this can be undone. the narrative of humanity can be undone. So it is a very conscious thing that people actually do. Yeah, in the same way, you can arrest somebody, you can decide not to erase them. 

So acknowledging that the African continent is a continent of 1.5 billion people, we are a huge landmass, and not minimizing us to something small in the middle of the globe, we are bigger than the biggest continents of the continent. We are the largest continent, but every time we teach geography, we are minimizing the place of Africa in the world, right? We are minimizing the intellectuality of African people of black women of African origin. Acknowledging that AI can help us acknowledge that Africans can do the rest. Because right now what we are fighting is we are fighting bias, barriers, and hurdles to get acknowledgment, one acknowledgment, acknowledgment is there, and we will do the rest. We are not asking the world a favor to do us, we are saying, can we stop believing that Africa is this small thing in the middle of the continent? That is a nuisance to the world? Africa has a lot to offer to the world. That is my closing remark.

Ross: That's fantastic. Yeah, as you say, you'll be able to do it for yourself and you already are. And I think that it's not that ignoring will be fading away as Africa makes a bigger, bigger impact of duty on you and so many other wonderful people in the continent. So thank you so much, not just for your time and your insights today, but also for all of the wonderful work you're doing to amplify humanity not just of Africa, but better the world. Thank you. 

Shikoh: Thank you so much.

The post Shikoh Gitau on amplifying humanity, Africa’s AI leadership, technology sovereignty, and the power of community (AC Ep42) appeared first on Humans + AI.

– Tom Hope

Tom Hope is Assistant Professor and Head of the AI Research Lab at Hebrew University of Jerusalem and a Research Scientist at Allen Institute for AI. His focus is developing artificial intelligence methods that augment and scale scientific knowledge discovery. His work has received four best paper awards and been covered in Nature and Science.

Google Scholar: Tom Hope

LinkedIn: Tom Hope

AI (Artificial Intelligence)

LLM (Large Language Models)

GPT-4

Claude

PubMed

Simulated annealing

Swarm optimization

AlphaFold

Semantic Scholar

Google Scholar

People

Nicholas Carlini (DeepMind researcher)

Nicky Kittur (from CMU)

Joel Chan

Daphna Shahaf

Ross Dawon: Tom, it's awesome to have you on the show.

Tom Hope: Thank you, thank you for having me.

Ross: I love the work which you are doing. And I suppose the big frame around this is how we can use computation to accelerate and augment scientific discovery. So,  just love to sort of start off well, what are some of the ways in which computation including large language models can assist us in the scientific discovery process?

Tom: One of the main ways I currently look at this is using large language models and more generally, AI to tap into huge bodies of humanity's collective knowledge, scientific corpora, as a great example, millions of papers, over 1 million papers coming out in PubMed, every single year. Of course, you have patterns, you have many other sources of technical knowledge. And these sources of knowledge, potentially our treasure trove of many millions, if not billions, of findings, methods, approaches, perspectives, insights; but our human cognition, while extremely powerful, and its ability to extrapolate and be creative, pull together all kinds of diverse perspectives, it's still very limited in its ability to explore this vast potential space of ideas, this combinatorial space of all the different things you can combine and the different things you can look into. 

As our knowledge continues exploding, so obviously, there are going to be more and more directions to explore as a result. So this problem keeps accelerating, with our knowledge accelerating. So the unique ability of AI and LLM recently to reason over complex texts, and complex data suggests that there is a future where the systems can help us humans, find those pieces of information that help us be more creative, that help us make decisions that help us discover new perspectives. By taking out problem contexts, the current thing we are interested in and working on a decision we want to make. And then somehow representing that in a way that enables retrieving these different nuggets or pieces of knowledge from these massive corpora, synthesizing whatever was retrieved into some sort of actionable inspiration or insight that helps us make the decision. And potentially, even automating some of these decisions and some of these hypotheses that we make as part of our process, there's still a long way to go there.I guess we'll talk about that right now.

Ross: Yep. Well, in one way, I'd also love to dig into some of the specifics and the details of the strategies for that. And also, just to start off, just actually pulling back to the big picture. I mean, how do you envisage the complementary roles of human cognition? And let's call it AI cognition in this process of scientific discovery? Where might that go in terms of those complementary roles?

Tom: So, we are living in quite revolutionary times in this area, right? I mean, things keep changing very rapidly. So to prophesize on what the ability of AI is going to be in a year from now, or even in a week from now, is a risky business, right? We can talk about what things are currently look like – currently the ability of MLMs and this new like as the representative of state of the art, AI, the ability to extrapolate from what it's seeing, it's massive training, like the entire web or the entire corpus of archive papers, let's say. The ability is quite limited. In our experiments and experiments by others is a nice quote I like from a Deep Mind researcher, Nicholas Carlini, that working with GPT4 is less like having a co-author on a paper, more like some addition working with a calculator. So a particularly strong calculator, right? But still, it's calculated. So if you wanted to come up with a new direction or creative direction, which as a scientist or as a researcher, that's a lot of what we do. So currently, it's quite limited. To give you an interesting example, I just yesterday tried to prompt GPT4 to come up with a creative new idea for mining scientific literature for generating new scientific hypotheses. It's kind of a meta kind of question. Because you're asking, it's how it could use itself to come up with a new scientific direction. I told it to be non generic and to be technical and go into details, etc. And what I came up with was, use predictive analytics and natural language processing to find new trends and directions. Okay, so then I tell you, well, GPT4 that's a bit too generic. Can you please be more specific? And then it's okay, so let's use quantum natural language processing and quantum predictive analytics. So its ability to do this test is very limited at the moment. It will either kind of go for these generic suggestions or recombine all kinds of popular concepts and software we want from an AI scientist. 

So currently, as a short answer, based on the current state of the art, and again, not saying what will happen in a week or in a year, it's time. Currently, LLMs can be our extensions, to scale up the way we search for the relevant pieces of knowledge, and potentially search for inspirations. Because, we're currently limited in our ability to see very narrow kinds of segments of human knowledge. Even in our very own specific areas, we're kind of losing the ability to keep track. So it could be that even if we've slightly extended out of our narrow kind of tunnel vision, will suddenly the kind of gold nugget that great inspiration we're missing will be out there, right. So LLMs can be that sergent. But the ability to synthesize a creative idea and to reason over it, and extrapolate into proposing something new and solid and reasonable. Currently, that's where humans are still needed.

Ross: Yeah, absolutely. And for good times to come. 

Tom: It looks like. 

Ross: So what I love about your work is that you have found ways to architect or to use LLMs and ways that are far more effective than out of the box. So for example, just ask GPT4, or Claude or something, it might give you a decent answer, or it might not. And even if you've broken, poke and prod a bit at it, whereas you have discovered or created various architectures, we're bringing these together. And so for example, in your literature based discovery, or in multi agent review processes, or, indeed, in your wonderful recent paper on scientific inspiration machines optimized for novelty. 

So, we'd love to just hear. I suppose the principles that you have seen work in how you take LLM is beyond just a text interface, towards  where it does create better, more insightful, more valuable complements to scientific understanding and advancement.

Tom: Yeah, sure. So, one core principle goes back to what we just discussed: the ability to retrieve useful inspirations. Okay, so we need to think about what an inspiration is, right? An inspiration is something that stimulates in our mind some sort of new perspective, or some sort of novel way to look at the problem – that's, let's say, one of the main ways to think about inspiration. And now you want to be able to give the LLM the ability to retrieve useful inspirations. That is, problems, let's say or potential solutions from somewhere around the design space of the problem you're currently looking at. So problems that are not too near but also potentially not too far. There is some sort of sweet spot for innovation, right? So if you want to be able to translate what I just said into some technical notions, you can embed your problems, and embed the solutions in some sort of vector space that enables the LLM to search for these inspirations. Then, prompt the LLM to consider those inspirations, synthesize a new direction, and then reconsider its idea in light of what's out there already. And that's when it's in the specific context when you're trying to innovate. Innovation, from the novelty is directly tied in to comparing to what's out there and expanding. And extrapolating out of what we currently know. 

So the second design principle is to have the LLM reconsider its ideas by comparing to existing work. And that is, again, a form of retrieval. But it's a different form of retrieval. Whereas, in the initial retrieval I mentioned, we want to be able to retrieve kind of structurally related partially related pieces of information, not necessarily more like things that are in the immediate neighborhood of your problem, but things that are kind of slightly outside of it. In the second phase of retrieval, we want the LLM to kind of be very accurate. And given that it's an idea, we wanted to now find the closest matching ideas out there, kind of like what a reviewer would do when considering a scientific paper. When a reviewer considers the scientific paper they want to know — Okay, here are five papers that are the closest to what these new papers are proposing, how close are they? Is the idea that's being proposed incremental or not? And the LLM needs to be endowed with this ability to find the most relevant work, and then compare and contrast it and kind of iterate over that. So those two design principles we implemented in that paper you mentioned of innovating, of scientific inspiration mentioned machines optimized for novelty. 

Ross: Just one question is, do the major large language models have sufficient corpus of scientific research? Or does this require fine tuning or retrieval, augmented generation are other approaches to ensure that you're addressing the right body of work. 

Tom: In my experience, it definitely requires retrieval augmentation, fine tuning could also help — that's a different story, because our ability to fine tune GPT4, for example, does not exist, right, because it is not open for fine tuning. And it's quite a big leap over other state of the art models, you know, Claude 3 is now getting close, but also we cannot find that. And retrieval augmentation is crucial for multiple reasons. First, you know, while the language models have been trained on, as we said, the entire web and probably have seen many of these papers out there; that does not mean that we can directly access that knowledge and get the LLM to access that latent knowledge with some prompt. If you just ask it to, let's say, come up with a way to relate to the work that's closest to some input problem that you feed in, it may well hallucinate a lot. And also kind of tend to focus. And this is rough intuition tend to focus on the more popular common areas that it's seen during training in less and less exponentially at the kind of tails of the distribution of, let's say, scientific papers and see and this is kind of very hand wavy, because no one knows exactly what's, how to quantify what's going on there when it's retrieving knowledge on this latent parameter space. But intuitively, that's probably what's happened. Right? So by retrieval, you can get a much finer level of resolution control when you're able to retrieve the exact scientific papers or sentences of nuggets of information you want the LLM to consider when it's coming up with a new idea.

Ross: So I was very interested in what you said earlier about finding the ideas that are sufficiently far away, but not too far away, as it were. And so how can you architect that, as you say, given that, the LLM probably is not really familiar with those concepts within the body of work that it has. 

Tom: So the way I think about this is via structural similarity, structural connections. To give you one of the most concrete examples, analogies. I've, in the past, and also fairly recently worked with, for example, Nicky Kittur, from CMU, Joel Chan, Daphna shall have on computational analogy, which is this kind of long old idea in the eye, where given some input, you can find abstract structural connections and analogies to other inputs. So for example I like, let's say you have some problem in optimization, you want to optimize some complex function or objective. Where would you get inspiration for doing that? Right? So if you use this kind of standard, let's say, search over a big corpus of technical problems, and solutions, you'll find many other optimizations, maybe you'll find some sort of other pieces of knowledge on mathematics and operations, research, etc. 

But can we go further and find inspirations from let's say, nature, from physics from, from how animals cooperate, right? So that is actually something that humans have done in the past, right? So humans have used inspiration from thermodynamics to come up with what's known as simulated annealing, right? The same sort of analogy between how thermodynamics behaves and and metals and mental heating and cooling, etc, to come up with some analogy for the energy of an objective function, or swarm optimal optimization approaches – optimization approaches based on multiple agents, let's say ants, searching some complex space, and then gradually converging into the local or global optimum points. So that's something that with standard search, you're not going to be able to find, but with structural kind of abstractions, being able to match on partial aspects of a problem or partial aspects of a solution, you can certainly get the retrieval to go out outside it's kind of initial local bubble and find more diverse perspectives.

Ross: Structural structures or problems and if you can find similar ones, that's immensely valuable. So how specifically do you get the LLM to be able to identify structurally similar problems or challenges? 

Tom: I'll give you one example that we kind of pioneered a few years ago, where we break down and input text, let's say a description of a past idea in a scientific paper, we break it down into two fundamental aspects: problems, mechanisms, the relations between the mechanisms and the problems right. So which mechanism was used for which sub problem connections between mechanisms etc. And given that you have this breakdown, you can now build a kind of a search engine that finds you ideas that share similar mechanisms.

Ross: To what degree is it humans or AI, which are doing that structural mapping? 

Tom: AI does the two main kind of heavy lifting of this pipeline. The first is going over millions of let's say papers or patterns, etc. and automatically extracting these aspects, the purposes, the mechanisms, etc. And then, as a second step, when you have some sort of input, let's say you want to find inspiration. So you conduct automated retrieval. You find inspirations with similar mechanisms, but very different problems.Then you can start by embedding these different aspects, you can come up with all kinds of similarity metrics, that consider partial matches partial matches by matching on certain mechanisms or matching on mechanisms while constraining the domain or the problem space to be distant than the inputs. And in that way, you can, for example, given some sort of problem on designing materials, you can come up with inspirations from biology, some of those real examples we've seen, or we've helped researcher, who is having some problems with discovering connections to between graph mining and, and some whatever their application domain was, I won't go into those details right now. But discovering some connections between that into decision theory. So by kind of conditioning on certain mechanisms and problem key phrases, but not others.

Ross: So one of your papers you looked at using a LLM to provide review feedback to scientific papers. And I suppose the basic idea was that if you just asked you that LLM didn't do a particularly great job. But you built a multi agent structure, which created a far better, more incisive, more useful feedback on the paper. So the thing about multi agents is the architecture as in how the multiple agents combined, in order to be able to create better insights, I would love to hear how you have structured those multiple agents to create that better review feedback on a scientific paper.

Tom: So just to connect that to what we're saying, right, the ability to review and an initial idea to review a scientific paper, it's kind of fundamental, if you want to automate the process of coming up with better hypotheses, right, because a reviewer agent can then refine an initial idea. And the most basic form of review is finding related work. And the contrasting to it, which, as I discussed, is something we've already done. 

But now, in the paper that you're just mentioning, we tried really hard to get GPT4 for you know, against state of the art to, to give us better feedback on a manuscript. And when I mean, a manuscript, it's a full PDF of, it's not just an abstract or a few sentences. And a main issue we saw is in terms of specificity. So when we asked GPT4 for or even with a lot of prompting effort to generate some sort of critical review of the paper, they often came up with suggestions like, you should consider more ablation studies, or you should consider adding statistical tests, etc,. And when you think about it, those are nearly always correct, right? I mean, it's pretty rare to have a paper that shouldn't consider more ablation studies, or do more mystical tests. So if you just evaluate the accuracy of that, well, it's probably gonna get you very close to 100% because it's pretty much always correct. But is that really useful?

We've also seen some previous work, also very recent on using LLM, like GPT for generic reviews. And they seem to have promising results. When you dig deeper into them, we find that a lot of the so-called promising results are because of that, because they generate kind of generic suggestions. So to make LLMs more specific, what we found to be the most effective currently at least, is that multi agent architecture you mentioned. So to get multiple LLMs to each one focus on a very specific aspect of a paper or in a very specific aspect of a reviewing process, right. So it'll focus only on the experiment section or want to focus only on clarity to focus only on the aspect of novelty compared to previous work. And then to get them to orchestrate, right? So you can think of the orchestration of an idealized metal reviewer, right? So I meant to review, unfortunately, at least in our area of AI hasn't had the bandwidth and time to kind of coordinate between reviewers and to have them kind of focus on specific aspects, you sometimes see that in journals, and kind of high quality journals not flooded by so many 1000s of submissions every month, that the editor will kind of reach out to expert reviewers each one focusing on a specific aspect of related to their expertise. And then coordinating between. So this kind of orchestrated LLM can take that role.

Ross: So are there any, are there any specific aspects of that orchestration in terms of how you guide the LLM to do that.

Tom: Our focus was to break the task down into multiple LLMs, each focusing on specific aspects. And then the orchestrator wasn't something far from what you'd imagine in the basic implementation of it. So it would take kind messages from each one of the reviewers, consolidate them, and pass other messages back to other reviewers so they can consider other contexts from other LLMs. Part of the reason we did this also was because at least when we were conducting our experiments, using one large language model to take in a full scientific paper was outside of its reasonable ability in terms of the context window. When I say reasonable ability, I mean that got added the ability to take in 128k tokens toward the end of our experiments cycle. But even with that, there's a lot of work on what's called last and the last in the middle effects are the ability of the LM to reason over complex, long documents kind of diminishes quite rapidly, even with fairly easy questions. And this is a very complex question, requiring kind of back and forth reasoning and comparing different parts of a paper and seeing if one claim is supported by another, etc. So that's why we needed to kind of break down into the multiple agents, the orchestration was fairly standard. In that sense, the main component here is how to break down into different aspects of reviewing.

Ross: So there, we've talked about a few different structures, the multi agent, the analogical reasoning, the other ways to be able to find structurally similar problems on. Are there any other high level architectures that you point to in your work that enable LLMs to accelerate scientific discovery.

Tom: So in terms of analogical reasoning, you can think of zooming out of that as a specific design choice for falling under the more general let's say building blocks of creative thinking, such as associative thinking or divergent thinking. And analogies that, let's say, as a fundamental and kind of wide reaching function for achieving those rights. So a different way to think about this would be to let's say, Forget about the aspect of analysis, but just diversify the inspirations that you're looking for – not necessarily in terms of analogies, but just diversifying your retrieved nuggets of knowledge. And this is something we've also been exploring recombination, tightly related to analogies, but not exactly the same, when we're trying to kind of recombine concepts. And when we try to recombine these concepts, the question is, how do you select the right one you want to recombine things that have not been not too close to each other, have not been recombined in the past or their nearest neighbors have not been recombined in the past? But also you want some notion of feasibility, right? You want to be able to kind of maybe predict the outcome of what's going to happen when these two concepts are merged together. Is it going to have some sort of no sort of impact that you can anticipate is the combination based on historical combinations is this combination likely, in some ways, you have this kind of very challenging balancing act of novelty likelihoods as a feasibility impact, we've started scratching the surface on some of these, right? So, and then work for just one example. And I can elaborate if you want more in Simon, the assignment paper, we also have fine tuning experiments. And fine tuning allows you to learn from past combinations of ideas. And when you're fine tuning, you're essentially optimizing for likelihood, right? The likelihood function, and LLM is what you're optimizing, the likelihood of seeing a sequence of tokens given the input. And in our case, that translates to the likelihood of proposing some idea given a problem, right. And if you can learn from past examples, you're optimizing for the likelihood, which corresponds to a different notion of what you want the LLM to do when it's coming up with ideas. But of course, if you're optimizing only for likelihood, you're kind of converging into the mainstream, like into the writer, and you want to balance it with novelty. That's what we've started to do inside.

Ross: So to round out, I mean, you're on the edge of this idea of how we can use AI to accelerate scientific discovery. So what is now the frontier? What are the research directions? Where do we need to push against to take the ability for AI to potentially vastly accelerate our scientific discovery process?

Tom: So it's important to note and obviously, we're not gonna have time to discuss those. It was important to note that scientifically, discovery is not only about, let's say, hypothesizing creative directions, right? I mean, alpha fold, as a kind of leading recent example for protein structure, and then leading to protein generation is a great example of an AI that can help boost scientific discovery without necessarily being creative in the sense that we think about it at least. 

So there's a lot of tasks that fall under the process of making scientific discoveries – designing experiments, conducting the experiments with some sort of agent that can actually issue commands to a robot, let's say in a lab in a wet lab, for example, and then a feedback loop that can kind of help the agent decide what are the more in kind of promising areas in the space of ideas, and then some of the some research groups working on that, as we speak.

Another process is finding the information you need, not necessarily inspiration, the information you need to solve a problem. You're currently having some problem in your experiments on optimizing some part of your device or process, etc. How can an AI agent help you understand your current context, your current problem, and then find information that's needed to solve it without necessarily being created?

So there's a lot of different aspects that go into science. And all of them need solutions. I think of it as kind of zooming out and thinking of one kind of big answer. The big question is to break down the scientific process into these major building blocks, components, modules, having agents, whether LLM or some other future architecture that may magically emerge and focus on these different modules and components. And do all of that work, while somehow understanding our human contexts or human objectives? Right, what we're trying to achieve is our preferences, very kind of ill defined, but you know, innate, fundamental human concept and human experience. That's very hard to convey to LLM by just you know, seeing let's say, your code or your Google Docs is not necessarily capturing what you want to achieve, what are your preferences? What are your subjective utility functions? What's your career goal? For example, right, or why a certain combination of ideas is something that appeals to you more than some other combination of ideas, because maybe it aligns more with your values or your ethics, right? So all these different considerations then how to translate those into kind of specific commands, the specific MLMs that can perform actions in each one of those modules that form the scientific process. So that's the kind of the biggest, let's say, frontier, how to build systems and models that can do that.

Ross: Yep. And clearly we are whilst, your work of you and your colleagues has taken us quite a long way, there's a massive amount still to go. And this is, but it's still such an important domain. I mean, the application of this could be transformative, and everything from healthcare to saving advancements space travel to who we are, and all this understanding. So it's an incredibly exciting field. 

So Tom, where can people go to find out more about your work?

Tom: So please check out my semantic scholar page. And of course, my Google Scholar page. Owork as a semantic scholar. I'll also mention my Google Scholar page. And check out more broadly the fascinating work done at AI to unscientific discovery. And of course, colleagues from other institutes also have my website online, you can quickly find me on Google. And please feel free to reach out if you found any of this interesting.

Ross: Fantastic! We'll provide links to all of your significant papers and also the related areas of interest. Thank you so much for not just your time and insight today, Tom but also your very important work.

Tom: Thank you very much for having me and have a good evening.

The post Tom Hope on AI to augment scientific discovery, useful inspirations, analogical reasoning, and structural problem similarity (AC Ep41) appeared first on Humans + AI.

– Céline Schillinger

Céline Schillinger is Founder and CEO of We Need Social, which works with organizations globally on engagement leadership. She is the author of Dare to Un-Lead, which was Porchlight Leadership & Strategy Book of the Year and on the Thinkers50 Best Management Booklist. Previously she worked in senior roles in the pharmaceutical industry across many countries and continents. Her extensive awards include Knight of the French National Order of Merit.

Website: www.weneedsocial.com

LinkedIn: Céline Schillinger

Sanofi

Network Activation

Employee resource groups

Community Studio

Book

Dare to Un-Lead: The Art of Relational Leadership in a Fragmented World by Céline Schillinger

Ross Dawson: Celine, it's a delight to have you on the show.

Céline Schillinger: Thank you so much, Ross. Thanks for having me.

Ross: So you work a lot with organizations and amplify their capabilities. And I think the really interesting starting point was, how is it that you think of what organizations are and how they function? What are the underlying principles that guide you?

Céline: Yeah, you know, this question came to me quite late in life. And actually, I started my career in small organizations in a very entrepreneurial kind of setting. I was working in Asia at the time. I moved to Asia, quite young, on my own to look for a job, look for adventure. And I started to build my career there, and I spent years in Vietnam, and then in China, and then I joined a large pharmaceutical company returning to Europe after about 10 years. And that was a shock for me to discover this whole new world of large enterprise. It had a different language that I did not understand. I thought I was already sort of a seasoned professional with 10 years experience behind me, but I did not understand this new language. It was talking about frameworks and metrics processes, and I wondered. I did not even understand the job description, I was off the job I was responding to the job offer is so funny, I asked someone to help me decipher this, I think, but that's part of organizational culture, to have this their own language and references and acronyms and all those things and ways of doing of course, so I discovered the large enterprise.

And for a while, I did not question or did not even wonder how it worked. Because I was all in on the pleasure of discovery. It was all about experimenting and meeting new people, and it was great. And then progressively I started to realize that, yeah, there's there are…how can I say principles ways of working, which do not necessarily emerge from which are kind of a religion kind of in a way – they do not emerge from the field or from common sense or the ways of working are prescribed and determined by habits, beliefs, and not necessarily by what would be needed, by customers by efficiency and so on. And I thought of, I had, maybe this kind of ethnological view coming from outside coming from a very different world. I started to question this, and question my role in perpetuating role models, behaviors that made no real sense. What was my role in maintaining that? Could I contribute to changing them a little bit instead? But what could I do on my own? So probably nothing. But then, about 15 years ago, I joined forces with other colleagues. And we formed a network of people wanting to bring about positive change, not wanting to protest. No, so I didn't join any union. For example, I joined a network. I co created a network. And that was when I remember the surprise, the puzzled look on the face of HR, HR did not understand what this thing was about. ‘An employee network. Well, what is it?’ It was before employee resource groups became popular? And then it was really weird for them, some of them. I remember somebody asking me who's the boss of your network, I would say, we have no boss, it's a network. But they felt like it was impossible to imagine another way of organizing than the one they were accustomed to. In the organization. A pyramid with a boss with a senior leader or the top, people reporting to him or her – often it's a him and we created a bit by chance originally was a bit of came a bit of a surprise for me, but we created something new a new way of delivering value, delivering value by connecting people around something they want to achieve together. There was no hierarchy, no one giving orders to each other to anyone else. There was a common desire, I was fueled by this willingness, this desire to create change, create an impact. 

Ross: This was around 15 years ago within the organization?

Céline: 2010, in a big pharma company that I was working with, at that time, called Sanofi. We created a new space for freedom, a space for creativity, where we sort of realized we empower ourselves. And we sort of realized that criticizing, blaming people or organization culture or the company for problems, what leads you to a better place, what may lead you to a better place is to actually roll up your sleeves, connect with each other and do something about it. Right? 

Ross: Absolutely. 

Céline: We had no idea until we started this and did it and, and it was amazing to realize that we had more power than we thought. And we didn't need a roadmap created by somebody else, we didn't need an order by or a job description or whatever, for other forms of prescription to create, and to innovate. And, and so we did that. And to me, it was a whole new world opening up to the whole new world of agency connection, and community building.

Ross: Originally, I mean, I think organizations are networks intrinsically, they always have been. 

Céline: Yes, you’re right.

Ross: So kind of, that's been harder to image, you know, given the traditional hierarchical structures. And yeah, the first thing that started to shift us more towards the realities of networks was actually email. So anybody could send an email to anyone else in the organization. And so that's the flattening of the organization's ability to connect. So the networks have always been there. That's the reality, all organizations are networks, they function as networks, but that recognition, and giving it the name, and to give me as you say, framing in the way, what you've done, just gives enormous power to the ability to create values. I think this idea of, you know, whoever's in the organization, be able to connect them with where they can create the most value, solve a challenge to see an opportunity. And so if you can have that fluid, network enablement, that creates an extraordinary value in the organization.

Céline: Definitely. I remember the time, I was very naive, and I remember drawing the org chart to newcomers who wanted to understand what this company was about and who we were…and I said, ‘Look, let me draw you the organization chart. This is how we work’ – how naive was that, right? Now, in hindsight, I am like, oh, this is just a symbolic representation. 

Ross: Yes, it is not the reality. 

Céline: Exactly. It is not the reality.  It is far from it, right? It does serve some purpose, including ego boosting purpose, which is not the most useful thing for business. But yes, we definitely need to expand people's views to other forms of representation. And one of them is something I've been working on lately. I call it “network activation” using visualization tools. So there are plenty of them on the market. And some of them can be extremely useful. Using some of those tools to make people look at themselves as a network and realize visually that they are a network they are connected by so many more things than they even imagine. And it's very often an aha moment for them. To see that what matters is not so much who is where in which position. How long are they still going to be the boss of this or that, but what matters is the density and the quality of our connections.

Ross: So, are you using digital trails or survey-based or how are you discovering what the networks are?

Céline: Yes, yes. So survey-based, is quite simple and very powerful. Because then you involve people in the responses in the process, right? You explain to them

Ross: They’re also thinking about it.

Céline: Exactly. 

Ross: I mean, I always love one of the best questions…there's a number of wonderful questions in network surveys. And one of them is, you know, who helped you the most in doing your work. And often, it's not the boss, or the person reporting to you; someone different in the organization. These kinds of things, and people start thinking, ‘Oh, well, actually, who is it that I draw on when I need help?’ And that's the start, you know, that's already a way of awakening that awareness. 

Céline: Yeah, you're right. These questions are not always easy to answer. But other questions are easier. For example, what do you know? Which country have you worked in? What? Those kinds of questions about personal professional experience, history, skills, aspirations. Then on a map, you realize that these things are actually common with other people that you didn't even know existed. But now you have a reason to go and talk with them, or to create something or sometimes you realize that there are potential nodes that can become communities of practice, for example, which are a fantastic way to further an organization — to connect the system to more of itself. 

Ross: So thinking about this, this idea of amplifying cognition, or just thinking about amplifying organizational capabilities. So this is, as you described, this is a wonderful network activation tool. So what are some of the other approaches that you use with organizations to be able to amplify the capabilities of individuals or the organization as a whole?

Céline: So you see, we are here on the podcast.  Podcast, I think, is a super interesting tool as well, to bring to the world of organizations. So I've also been working with a partner Lila North, on the Community Studio — some it's an offer, we've, we've been implementing successfully in several organizations by which you get a group of volunteers, create together an internal podcast, with a series of episodes and the volunteer group is gets renewed after each season, each podcast season. So you amplify the group of people that the podcast community, the internal podcast community, through, it's really important to have community engagement there so that the guests can become part of this community. And you have this community, this community grows progressively and becomes a sort of not a platform in the sense of a technical platform, but an opportunity, a group that enables cross entity cross level conversations. And that creates a habit or an openness to curious conversations. It's really hard in organizations today to have curious conversations about each other. We're so focused on our rules on milestones or deliverables and we're still enclosed in this hierarchical structure very much. And this pushes communication habits with. With these kinds of things, internal podcasts, the studio is run by a community of volunteers, and from which we collect insights in order to create meta conversations. Were able to open up I wouldn't say change dramatically, culture, this is I don't think this is possible anyway. But at least open up new possibilities. And whether people seize or not depends on them. So we always remain very aware of the freedom we need to let people act otherwise if they act upon order from anybody else or upon our suggestion or if it's not their own. You like this ownership piece that makes it sustainable.

Ross: Most of the best ideas come from conversations — the best thinking, ways, and perspectives. So what you're doing is basically having these conversations in public so they can be heard by the organization. I love this idea of being able to distill that into meta conversations. But I'm interested in some of the practicalities of that. I mean, you've got networked people who are interested in that. But how do you disseminate that to make that people listen to it? One of the ways in which you helped him propagate this through the organization?

Céline: You have to make it interesting. So we equip volunteers with good question, interviewing skills. And it's fascinating to see that they become better and better at interviewing people. First, in the first episodes, they sort of follow the script, you know, the questions we've written together, it's very, very scripted. And progressively, you see them evolving, and actually paying attention to the responses of their guests, and asking follow up questions. And that kind of thing. It's really fascinating to see it develop skills, but also it creates better episodes, more interesting questions. Yep, sessions. 

And so with promotion, and engagement, it's also part of the work that volunteers get involved in. And so we equip them with that kind of skills, we help them become engaging leaders, rather than just makers of something. You know, it's about engaging colleagues creating connections, and then connections, and conversations over those. Those first conversations, it's really interesting to hopefully see it grow and expand throughout the organization, some departments or less, for example, it's very easy to get salespeople interviewed. Those people are used to talking and you know, bringing their points. And for some others, it's more difficult for people in maintenance jobs and technical jobs on the front line. But that's the challenge. That's part of it, it's precisely what we try to bring volunteers to do more of, you know, go and have those. Reach out to these. These people who do not have a voice, try to build rapport, create the conditions for them to come in and talk and express their views because we don't hear those people enough. 

Ross: I couldn't imagine to see the…you'll get a far richer flavor of the organization, usually just speaking to the people you deal with in your current projects and your work. So to be exposed to, as you say, technicians or maintenance or other far, far flung parts of the organization that would really make you feel more belonging. So this is internal only. So just available on the intranet.

Céline: Yes, this is internal only because you have freedom of speech. It's already not easy for people to speak openly on a podcast on an internal podcast. So if it was external, it would be really way too challenging. But you know, I'm thinking of another example of amplification, which is a piece of work I was involved in, back in 2014, to ‘18, when I was an internal change agent, I did not work on my own at that time yet. But it was really an interesting piece of work involving volunteers as well as quality improvement. And for many years, the company had tried to establish pharmaceutical company factories, enormous industrial challenges. And for many years, the company had addressed those challenges. Were through the quality department — a small group of experts, professional people, highly dedicated to their mission. 

But this was not enough. And the outcomes were not great. It was only when we amplified this work by involving volunteers, by involving people from all over the company, not just quality professionals, but anyone, anyone who wanted you can be you could be, I don't know, a legal expert, you could be an admin, you could be a technician, anyone was welcome to participate in this movement. And there are ways of creating a movement. And it doesn't work. All the time in this particular case, it worked beautifully. And we engaged I think, around 5000 people instead of the originally, I don't know maybe 250. You were involved in creating quality, improving quality and by having many more people, but also many more viewpoints, many more a greater diversity of perspectives. And also by making this work, not just an intellectual work, like, how do we solve problems, but an emotional work, too? How do we connect around solving problems? How do we make it engaging? Interesting? How do we create enthusiasm? How do we make people? How do we create desire? Right? So this is what made it work.

Ross: So extending your ideas, I mean, I'm not sure to what degree you think of yourself as a futurist. But I'd like you to cast forward to, you know, this, these the ideas you have around the sorts of organizations that are truly effective. So let's say you know, 2030, whatever in the years to come, we have many unfolding forces. So what are some of the ways in which you would point to this very successful organization of the future, and how that can be enabled?

Céline: You know, I think I'm not a futurist at all, I'm a presentist. Because these things already work now. So you don't have to wait until tomorrow to put them in practice. They do work already now. And I would say, I would recommend three key practices or lines of thoughts, right? That I have found for myself and my colleagues, and my clients now immensely useful, the first line of thought is around agency, creating more space for creative freedom. Instead of trying to enclose people further and further into narrow job descriptions or scripted courses of action, instead of trying to transform them in a way as in robots, we will never be great robots. So it's about removing those expanding territories, in which people can first get back this thinking capacity that is often lost under process in organizations. So recreating space and time to think and it starts with ourselves, right? What do I maintain from this system that would deserve to be changed? How can I be authentic to my n and really walk the talk and what I do, and what I think and so agency creating more space for people to to act for impact, to decide to negotiate to create sense making opportunities and so on. The second line of thought is around networks, creating, removing this, we talked about this pyramid, hierarchical thought pattern, I think the hierarchy will not go away. I think it is still useful in many ways, but removing the patterns of domination and submission that it entails, will be immensely useful. So that information can flow faster and be more readily accessible throughout a network. That's why we need to bring in. 

Ross: What are some of the enablers of that? I suppose we want to create more networked organizations that often say that the successful organizations of the future will be very effective networks. What are some of those things that enable that?

Céline: think volunteer networks, think digital networks, enterprise social platforms, think communities of practice? Think network activator with a network visualization, I mentioned think those kinds of the Community Studio amplifying stories in a peer to peer mode. The possibilities are infinite. As soon as we move away from this pyramidal thought pattern, and try to look at networks and what could enable them then we realize the possibilities are limitless. Now we need to find, I would say the most practical and simple solutions to put that forward but creating a volunteer network around an opportunity. Something that really matters to an organization is a good way to start. If you create a volunteer group around something that is not really valued by the data – that is not that important for the organization will not produce much impact. But if leaders, leaders, if the company gets really serious about it, let's involve more people. 

Let's change the type and the nature of our conversations. I remember the clients I worked with a few years ago who had decided to create a new technology that they wanted to roll out. And I suggested that instead of rolling it out, they create conversations around with people who would be affected by this new technology, and create a volunteer group with people who wanted in order to address that issue, the technology and more broadly, the digital future of the company. And by doing that, it was a very simple move. But by doing that, we transform people from victims of a change to co creators of a change. And we formed networks between these people, and between these people and their leaders — their leaders and titles.

Ross: So that was the third point too?

Céline: Yeah, to community building exactly, let’s third line of thought is creating community, bringing the network together and making it stick together. So that it doesn't go in all directions, and sticking together around a big opportunity, the vision of a better future that is CO created by those people, not just by the executive team in a boardroom. But that involves at least a representative sample of the organization. So that a diversity of perspective is already present from the very start of an initiative. And then there's a lot of effort to be made to reinforce the value of this community so that people do not default back to a purely functional role, a vision of their role.

Ross: So, to round out, I'm interested in just you personally, in how you…you have expansive ways of thinking and experience in ways do you apply that? So I'd love to hear anything you would do personally, to be able to amplify your own cognition and thinking and capabilities.

Céline: I've been using digital networks a lot myself, and that has been a huge enhancer amplifier. enabler, I've been able to connect with people to learn about new ideas, new thoughts. I was an avid fan, that's the first thing. So connecting with people, and I'm very sad of what Twitter has become, because I don't like it anymore. But what it was in the past was a really amazing blessing. So I'm very grateful for that. 

And the other practice that I've used personally was to write. And it took some effort for me. At first I felt absolutely unable and not even legitimate. I thought, you know, what could I write about? Why would people even read anything about me? Or about my ideas, and I was pushed gently by friends of mine who said, ‘Yeah, you should.’ And actually writing is a fantastic way to organize your thoughts, to expose them to others, to be challenged by others, to grow. Now I look as I look back at some of my older posts and think, I wouldn't think this way anymore. But it was a necessary step in this process of sense making, really. So I think this is a great thing and making them public so that you can share and connect and learn from others.

Ross: It creates a feedback loop. It networks of thought and I think one of the largest networks of people is networks of ideas and thoughts. And when you put things out there, then that starts to catalyze, these different connections, these different ideas, the different possibilities, so, absolutely,

Céline: yeah. And participating in podcasts, like yours is also another way. That's why I'm extremely grateful for the opportunity.

Ross: Where can people go to find out more about your work?

Céline: They can find me on LinkedIn. They can find me also on my website, weneedsocial.com

Ross: Fantastic. Thank you so much for your time and your insights, Céline.. It's wonderful work you're doing.

Céline: Thank you so much, Ross. Very grateful.

The post Céline Schillinger on network activation, curious conversations, podcasting for connection, and creative freedom (AC Ep40) appeared first on Humans + AI.

– Charles Hampden-Turner

Dr. Charles Hampden-Turner is a British management philosopher, business consultant, and co-founder of consulting firm Trompenaars Hampden-Turner. He is the creator of dilemma theory and the author or co-author of numerous influential books, including Maps of the Mind, The Seven Cultures of Capitalism, and Mastering the Infinite Game. He is received many awards, including Guggenheim, Rockefeller and Ford Foundation Fellowships.

Website: www.thtconsulting.com

LinkedIn:

Charles Hampden-Turner

Fons Trompenaars at TROMPENAARS HAMPDEN-TURNER

Facebook: Trompenaars Hampden-Turner 

X (Twitter): @FTrompenaars

YouTube: Trompenaars Hampden-Turner

Freud's ID and Superego

Jung's Collective Unconscious

Mobius Strip

Yin and Yang

Conscious Capitalism 

People

Mitchell Beasley (Publisher)

Gregory Bateson

R.D. Laing

W. Edwards Deming

Ray Anderson

Paul Polman

Books

Natural Capitalism by Paul Hawken, Amory Lovins, L. Hunter Lovins

Maps of the Mind: Charts and Concepts of the Mind and its Labyrinths by Charles Hampden-Turner

The Seven Cultures of Capitalism: Value Systems for Creating Wealth in Britain, the United States, Germany, France, Japan, Sweden and the Netherlands by Charles Hampden-Turner and Altons Trompenaars

Mastering the Infinite Game: How East Asian Values are Transforming Business Practices by Charles Hampden-Turner and Fons Trompenaars

Ross Dawson: Charles, it's an honor and delight to have you on the show.

Charles Hampden-Turner: Well, good to meet you. And if I can help you let me know.

Ross: Thank you. So I first came across your work when I was in a bookshop in Geneva, Switzerland 1981 or 1982, it must have been, and I saw on the table, this book, which had maps of the mind, and it was immediately resonated with me, because what is, you know, the basic exploring all of these different models, what the mind is, and how we think and be able to not just explain those, but also to have a visual representation to show us what they were, and I've still got it, I still refer to it. And it really influences my thinking. It is so useful to have these maps of the mind to help us understand the way we think to bring that to life. So I'd love to just hear the genesis of maps of the mind and just some reflections back from quite a few years later on, on those wonderful projects you did.

Charles: Well, I knew Mitchell from Mitchell Beasley, and he was always producing encyclopedias, including the joy of sex and other things. And he said he wanted to do something on the mind. So I approached him and said, I could create 60 visions of the mind, all of which I, I loved and asking myself, why did I love them? It's because they were consistent because they had something in common. I hadn't in those days worked out what they had in common. But once I finished, I began to see what they have in common. What they have in common is that all paradoxes starting with Freud's ID and superego are about as different as you can get and Jung's collective unconscious and libido etc. And if you go all the way through the book, you'll find every map has a duality. And every map has a reconciliation of that duality. But I only realized that in retrospect, and I longed to add a chapter, explaining that the whole book is often a piece is part of an overall pattern.

Ross: I think your selection of the models in the book actually reflects that as in many of them are quite explicitly about paradoxes, such as Gregory Bateson or artie Lang or others that you chose. So I think that framing and the choices you made already, implicitly suggested that you had the pattern in your mind already.

Charles: Yes, I did. But you don't know what your subconscious is doing.

Ross: So you've written many books on management, cross cultural leadership, around the you know, essentially what it is that drives the value in organizations. And more recently, you are working on a book which ties you've said to me all of your life's work together. And so how, what is what is how, how could you tie together or pull together all of your threads of this marvelous work through your life? 

Charles: I was thinking about the German mathematician and Mobius. And he created the well known Mobius strip. And you give just one twist to a paper loop. And when you give one twist, suddenly, the sides disappear, there is no there is just one side, you can take a pen and draw all the way around and you will end up where you started. And there won't be any part of the loop that doesn't have a line on it. Or you can take an edge. You can follow the edge run with your finger and you will finish up where you started. It has gone from one strip with two sides and two edges to a noop. With one edge and one surface. And essentially you're moving in between. It's like yin yang, which has been Chinese folk wisdom for Centuries, and is probably partly responsible for economic progress. And, and it's also by its ying yang, but it's also like to you, you, you'll see red, you see green, but you'll never see red, that is not about to become green. And you'll never see green that is not about, but to become red. And if you turn opposites into contrasts, and constantly move between the contrasts, then there is almost no problem in the world that you can't solve. I mean, it still takes great skill, a genius, but at least the problem is solvable. You are tough on problems, and tender on the people who have the problems. And you can take almost any dichotomy you like, and give it one twist, and it becomes one continuous process.

Ross: So want to come back to how this applies to businesses and value creation, the economy, but I mean, the apps one of the most obvious questions to arise is that we live in a world of polarization, particularly of political polarization of your certainly, economic polarization of polarization of wealth, and augmented by a whole series of factors. So this model of the Mobius, in perhaps integrating or bringing together paradoxes or polarities, how can this be applied in resolving the polarities we are experiencing in society today.

Charles: But it's already been, it's already been applied. And with great success, W. Edwards, Deming visited Detroit and tried to get them to use his error correcting system, but they all turned him down. So he went to Toyota, and taught them and Toyota now produces almost a million cars a year, and is larger than the entire American automobile industry. Nearly all thanks to W. Edwards Deming. So you, inevitably when you try to do something, it's imperfect. And so you start with an error, and then you correct it. And in other words, you give it the middle, upper half, twist, and glue it together, and you get errors, really leaving corrections. And now comes the important part leading to continuous improvement. So the more errors you make, the faster and quicker you get better and improve and improve and improve. And if you don't think you're making an error, then raise your aspirations a bit. If you raise your aspirations a bit, you'll soon start making errors again, and you'll correct those, and you get better and better and better. But this is true of nearly all businesses, this is true of the invisible hand. You want to make money for yourself. And you can do that if you satisfy customers better than anyone else can. So you succeed in you, you succeed in communicating and cooperating by competing, competing and cooperating are really two contrasts like the lights of a dog, and together, they spell safe driving or good business.

Ross: So as you say, you're applying this to value creation, as in value creation being mutual, you can't, the only way to create value for yourself is to create value for others. 

Charles: That's why Quakers made so much money. Yes, because they were constantly helping other people in their society, and they weren't allowed to go to Oxford or Cambridge. They weren't, they weren't allowed to. They were barred from major professions, and they helped each other. And Quakers in Britain in the 19th century, created 40 times more wealth than their numbers allowed 40 times more wealth than their numbers, personally, and that's by helping other people, and thereby helping yourself by guiding yourself.

Ross: So over the last couple of decades, and moreover, there have been some very powerful forces which are reshaping business globalization, of course. The Internet and the connected world give the ability to create value across domains, the rise of platforms and other things which have come from telecommunication. So, do these change the nature of value creation today? What is where we sit today in terms of value creation and this model?

Charles: Work is important because it means that customers, not only contact the supplier, but they contact each other. And, things go viral. In other words, suddenly things catch fire, because customers are telling each other how good it is, and they trust each other more than the supplier, the supplier is going to make money out of them, but they trust each other. So anyone who's not on the internet is losing a hell a hell of a lot of money, it is a great accelerator of business and everybody needs to be on the net. So the net effect is what happens when customers talk to customers and talk to you.

Ross: So using this metaphor of the Mobius strip to know a world where businesses are more global, more and more value happens across organizational boundaries, more partnerships. I mean, how does this guide leadership and how organizations are structured in terms of how politics are structured?

Charles: You mentioned the word metaphor. And the metaphor is itself a paradox a reconciled paradox, because a metaphor is the likeness of unlike characteristics, okay? The ground was blanketed in snow. And in some respects, snow is very like a blanket. But if you snuggled down under some snow tonight, you wouldn't be very comfortable. So snow is both like a blanket and a blanket. And metaphors are heuristic devices that there are ways of finding out reconciliations. I'm all for metaphors. Scenario planning was when it was invented in the shower by never gonna use the metaphor of a southport rider. You see the waves coming towards you, you get ready for them. And you don't know which wave is going to hit you. And you get ready for each one.

Ross: Yes, yes. So speaking of that, scenario planning has been a thread through much of your work, I believe.

Charles: Yes, I joined Shell in 1981 and worked as an in-house consultant for three years. And it made such a great influence on me. And there were 60 or 70 people in the department, you won't find a planning department in the world with 60 or 70 people. 

Ross: one of the frames which you bring in thinking about Mobius’ leadership and that of conscious capitalism. So how is it that we can both, you know, apply these models to be able to make the capital which we apply to create value or, you know, the choices we make to be more conscious as an organization?

Charles: Well, the irony is that, you know, everyone thinks business is selfish, and we celebrate selfishness. But the truth is quite the opposite: capitalist countries have better education, better health and a lot longer life. Capitalism bestows upon us enormous advantages over the non-capitalist countries. And so we have to, we have to face this ridiculous dilemma that business is all about self interest and selfishness. And yet, it's probably one of the greatest sources of benefit to humanity. 

And the reason is actually quite obvious. And the reason is that when you do something for somebody else, they reciprocate, and they come back to you and do something for you. There are many examples, UPS hires young teenagers to drive its trucks. And in America, it gives money to college scholarships to every mile. The kid drives. So if you've driven 20,000 miles, you get $5,000 towards a college scholarship, you think, how on earth can UPS afford to do that? I mean, it's ridiculous, you, you can't give away the store, you can't start giving away? Well, can't you if you have a choice between services between different courier services, and you knew that your son or your daughter, or your friend's son or daughter, or that your nephew or niece had been helped by UPS? Which courier service would you choose for the rest of your life, and which courier service would the recipient choose for the rest of their lives, they get a college education, they're going to get a good job, they're going to use about 700 careers during their life, or more. And so when you appear to give away you don't give at all, people reciprocate. And you get it back and you get it back in spades. And there are lots of examples in a book called What's it called? Conscious Capitalism, that if you, since you do good by accident, why don't you do good deliberately look at the accidents and start doing the accidents on purpose.

Ross: So one of the challenges of capitalist structures is, so called externalities, where people can be part of a system and they have mutual value creation. But there are things which are outside the system outside the system that is measured, or outside, people looking at so most obviously, in terms of pollution, or carbon impact, or other ways in which, it's not customers or people that are directly there, but they are, external to the system was as it is accounted for today. And so that's where the conscious part comes in. I think we're in a way, as you're describing some of these benefits can be raked, you know, as you say, just by accident, as it were, or beginning to be conscious about.

Charles: You have to make the externality, internal. Economics can't ignore virtually every value and that makes life worth living. And we live on a hospitable planet that has its blessings upon us. And in 200 years, we've got we've come close to wrecking it in certain respects. And if when you do something, you improve the environment, then that makes your work worthwhile. It reconceptualized your work in something important and something in something moral. Suppose you make carpets there's nothing to be ashamed of in making carpets, they're useful. And interface carpets in America, the bed carpets and Ray Anderson reached the age of 60 and read a book on Natural Capitalism and suddenly realized that he was wrecking the environment. The carpets are made of nylon, nylon comes from oil, etc. So he pledged zero emissions by 2020. And suddenly his people had a new idea to work for. Is it better to make carpets or to save the environment? If while making carpets you save the environment. This makes your work far more worthwhile, far more exciting, and you have something to leave your grandchildren. I've left your world that is still beautiful because I helped make carpets in a non-polluting way.

Ross: You've worked with many leaders of large organizations over long periods. And I'm sure that there are some who are more and less receptive to these kinds of ideas. So how do you engage with some of the leaders who are perhaps more skeptical? When you start to discuss this kind of concept, how do you start to shift their thinking? What's that journey?

Charles: Well, I start with a crisis, I start with something that is going wrong for them. And after all, they wouldn't be talking to me unless they wanted something. And then if a business faces a very serious crisis, and comes to me and says, This is going wrong, people are lying, the thing is corrupt and things like that, then I think I can intervene, but problems vary from person to person. But if you start with a dilemma, or you start with a crisis, it will start with some I want this, but this gets in the way, then you're, then you're onto something.

Ross: So there's nothing like a good crisis.

Charles: But in every crisis, there is an opportunity. If the world is indeed, if we're on the edge of a tipping point, when the world is no longer hospitable, where people try to migrate, other people stop them with guns. And we already have a crisis, and people crossing the channel, and people trying to make political hay out of that, we could be on the edge of something very, very dangerous. Anyone who saved us, that is not simply making a useful product. God dammit, they're saving the world, saving us, saving all of us, they're saving themselves, they're saving their children. So any crisis gives us vast opportunities to be of help.

Ross: So you mentioned that came from a dilemma, which of course, you know, is a paradox. And from when I was younger, I was thinking, is this idea of making the paradox more extreme? So rather than trying to start from the start to resolve the paradox, you actually push the polarities out in order as a mechanism perhaps to be able to as a path to resolution. 

Charles: That's quite a good technique. I use that a lot. When I'm consulting. Somebody says, the trade unions really got it in for us. Yeah, they're giving you trouble. So we should walk, we should wreck them, we should undermine them, we should take them to court, we should. And you make this longer and longer with more and more data threats. And he will say, ‘Well, I can't do that. So you can't do that.’ So you're going to have to solve it. And then you give your suggestions. So yes, if they see a dilemma, if they exaggerate the dilemma, then you go along with them, and they will want to take pills, oh, well, I can't do that. I will end up in prison.

Ross: So which goes to this idea of the mindset, how can we prime our minds to be better at resolving paradoxes? I mean, we have always lived in a world of paradoxes, as you've suggested, I mean, even more today, the world is full of paradoxes and polarities and dilemmas. So can we nurture a frame of mind that enables us to move more readily towards resolution?

Charles: Well, I think so I, what I do is give lots of examples. Things we already believe in our paradoxes, for example, the marketplace, people who worship the market. When things get scarce, prices go up. And when prices go up, you can make more profit by supply. So buy supplies arrive and the prices go down. So in the marketplace, prices go up and down and they are self regulating. And lots of people see this as a sort of Calvinists God, punishing the slow fall and rewarding the thrifty. And, but it is a simple paradox. So you just increase the number of examples of how we already use paradoxes. In our understanding, the invisible hand is a paradox you, as I mentioned earlier, you follow your self interest. But in order to do that you have to satisfy a customer better than that, and other people can. So paradoxes are everywhere, you have to make people excited and interested in paradoxes, at the moment that they are afraid of them, because they're afraid of contradiction. And they're afraid of appearing irrational. So everyone wants to be rational. And to be rational is to decide what comes from that in de se do, to cut off. So you have either or, and you cut off, what we're talking about is a choice combination, that either all or both and, and coming together, they solve all manner of problems.

Ross: Yes, the So you're suggesting is becoming more familiar with the fact of paradoxes in our world, and the fact that they include the resolution within the paradox. And as we see that, and make that visible to ourselves more, that makes us more able to see that and other paradoxes that we come across.

Charles: Very much. Agree. 

Ross: So what's the core of the impact you want to have from here in terms of taking this thinking forward? How can that be disseminated to take them forward into the, to the leaders or to all of us in being able to understand and to engage so that we can?

Charles: Well, I think we have to create a few heroes, and somebody who creates wealth, meaning, substance and aspiration, and success for a company needs to be admired. And we put all this we put all this time to intuition. But paradoxes are the logic of intuition. And I keep trying different things. But what I'm trying at the moment is to get Paul Polman. At the moment, he doesn't answer emails. I mean, he's so well known, he probably gets 100 emails a day, and he doesn't reply to hundreds of people he doesn't know. And I don't blame him, I probably have to do the same thing if I was that famous. 

But I would like to get Paul Polman to say yes, I was 10 years with Unilever. And this is the logic by which I proceeded. And so he's not just a brilliant intuitive leader, whose memory will fade with time, he is the author of a new logic of management, or a new logic of leadership. And that's one way to get it across. And the other is to just to keep writing or to create a podcast creatively filmed, create a video to it's highly visual. So if we, I sent you some some PowerPoint presentations, you can see that it's easier to visualize than it is to talk about, yes, it is a helix and I have a dot that goes up a curly helix that moves first in one direction and the other direction and first, and knits the two together. So you create a helix that joins the two together.

Ross: And indeed, yeah, that's which takes us back to the maps of the mind where the visuals help us to understand. And yeah, there are some wonderful visuals which have been used to illustrate how particular paradoxes are resolved with Mobius structure in your visual, so I very much look forward to seeing that come to fruition in the book in another work. Good. Thank you so much for your time and your insight and all of your life's work has been an inspiration to me throughout.

Charles: I hope it helps. I hope you are attracted. I'm running out of time. And thank you very much for talking to me.

The post Charles Hampden-Turner on Mobius leadership, reconciling paradoxes, dilemma strategies, and conscious capitalism (AC Ep38) appeared first on Humans + AI.

– Philipp Schoenegger

Philipp Schoenegger is a researcher at London School of Economics working at the intersection of judgement, decision-making, and applied artificial intelligence. He is also a professional forecaster, working as a forecasting consultant for the Swift Centre as well as a 'Pro Forecaster' for Metaculus, providing probabilistic forecasts and detailed rationales for a variety of major organizations.

Website: Dr. Philipp Schoenegger

LinkedIn: Philipp Schoenegger, PhD

X (Twitter): @SchoeneggerPhil

Artificial Intelligence (AI)
Large Language Models (LLMs)
ChatGPT
Judgment and Decision Making
Superforecasting
Philip Tetlock
AI Augmentation
The 10 Commandments of Forecasting
Alibaba
Claude (Language Model)
Palm (Language Model)
External vs. Internal View in Forecasting
International Energy Agency (IEA)
Metaculus Platform (Forecasting Platform)

Papers

AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy

Wisdom of the Silicon Crowd: LLM Ensemble Prediction Capabilities Rival Human Crowd Accuracy

Ross Dawson: Philip, it's wonderful to have you on the show.

Philipp Schoenegger: Thank you so much. Thanks for the invitation. It's great to be here.

Ross: So on your website, you have this very interesting diagram, which shows that your current research is around the intersection of judgment and decision making, and applied artificial intelligence. So what is that space? And how have you come there? What's, what is that that pulled you to this particular space? 

Philipp: So I think what, what really motivated me to work in this area is what motivated many other people to jump into AI. And this was just the release of ChatGPT, in late 2022. So I hadn't been working in artificial intelligence before, I have a social science and humanities background, having worked on charitable giving, and political philosophy before. But having seen ChatGPT, I think it took 10 days. Until I had my first research project, we've caught up slowly. And our first idea was, how can we? How can we kind of mimic social science participants with artificial intelligence? And so what we did is we ran a bunch of studies that had been replicated in humans with the text DaVinci, free to see the early ChatGPT model. And ever since I've never looked back, and I've pretty much only wanted to do more AI stuff. It's way too interesting. At this point, pretty much almost all the work.

Ross: I think we're pretty aligned on that. And it's just like this intersection of human intelligence, artificial intelligence is so deep, so promising, so much potential. And so now it's wonderful to see the work that you're doing. So speaking of which, we recently, you were a lead author of a paper.

AI Augmented Predictions, LLM assistants improve human forecasting accuracy. So first of all, let's just describe the paper at a high level, and then we can dig into some of the specifics.

Philipp: So the basic idea of this paper is, how can we improve human forecasting. Human judgmental forecasting is basically the idea that you can query a bunch of various interests and sometimes lay people off about future events, and then aggregate their predictions and arrive at surprisingly accurate estimations of future outcomes. So this goes back to the work on Super forecasting, the Philip Tetlock. 

There's a lot of different approaches on how one might go about improving human prediction capabilities, absolute maximum training, as it was called the 10 commandments of forecasting, how you can better forecast out or there might be some, some, some conversations where different forecasters talk to each other and exchange their views. And we wanted to look at how we could think about improving human forecasting with AI? And I think one of the main strengths of the current generation of large language models is the ability of the interactive nature of the back and forth to have a highly competent model that people can interact with and query whatever they want. Really, they might ask the model, ‘Please help me with this question. What's the answer?’ They might also just say, ‘Here's what I think please critique it.’ And so this opens up for human focus, like a whole host of different interactions. And we wanted to see what the effect of this might be on forecasting accuracy.

Ross: That's fascinating, I suppose one of the starting points is thinking about these forecasters. So I suppose just so people could be clear that human forecasting in complex domains is superior to AI forecasting, because they don't have those capabilities. So they're saying humans are better than AI alone. But now the results of the paper suggest that the humans augmented by AI are superior to either humans alone or AI alone.

Philipp: At the current amounts of papers that I have published, yes. But depending on when this airs, there might be another paper coming out that adds another twist to this. But yeah, so in early work, we find that that just a simple GPT 4 forecaster underperforms, the human crowd and on top of added underperforms, just seeing 50% on every question, but in this paper, we found that if we give people the opportunity to interact within large language model, which in this case was TBD for turbo, and we prompted it specifically to provide super forecasting advice. 

So our main treatment had a prompt that explained that the 10 commandments are super forecasting and instructed a model to provide estimates that take care of the base rate, so you look at how often is it that things like this have typically happened, that quantifies uncertainty that identifies branch points in reasoning. But then we also looked at what happens if the large language model doesn't give good advice? What if it gives what is called bias, that is, if I'd be more noisy advice. So what if the model is told to not think about the base rate, so not think about how often things happen to be overconfident to basically give very high or very low estimates to be very confident. And, to our surprise, we find that actually, these two approaches similarly effectively improve forecasting accuracy, which is not what we expected.

Ross: I think that this is a really interesting point, because essentially, this is about human cognition. So it is human cognition, taking very complex domains, and coming up with a forecast of a probability of an event. So or a specific outcome in a defined timeframe. So in this case, the interaction with the AI is a way of enhancing human cognition, that they are basically making a better sense of the world. And I guess one of the things which is more distinctive about your approach is, as you say, you could allow them to use anything anyways of interacting, as opposed to a specific dynamic. So in this case, it was all human directed. There was no AI direction. It is AI as a tool, with humans, I suppose, seeking to augment their own ways of thinking about this challenge. 

Philipp: That’s right. And of course, being human has asked me to, like make at least a sizable amount of participants just simply ask them the question, right. I just said, well, once the question will be the closing value for the Dow Jones at the end of December, and I just copied it in and just saw what the model did. But then many others did not. And they had their own view. And they typed in, ‘Well, I think that's the answer. What do you think?’ Or, you know, ‘Please critique this.’ And I think these kinds of interactions are especially promising going forward, because there's also this whole literature on the different impact of AI augmentation on differently skilled participants, differently skilled workers. 

In my understanding, the literature is currently mixed. So studies are finding different results. So we didn't find a specific effect here. But other work finds that when the model just gives the answer, low performance typically tends to do better, because you know, they'll know the answer. And the models are probably better than them. But if the model is instructed to give guidance, only low performers tend to, you know, not be able to pick up on the guidance and follow it. But I think there's still, there's still a lot of interesting work to be done before we can pin this down, because there's so much diversity in which models are being used. What's the context?

Ross:  Yeah. But I think that's a particularly interesting outcome in the sense that humans are mainly not very good forecasters. And it's only a relatively small proportion of people who are good forecasters. So it would have thought that there would be some kind of differential because it's almost like, people who have some kind of understanding of what forecasting takes, and others who doubt the kind of basically, you're guessing, but it showed some similar improvement. I think that's a very interesting outcome.

Philipp: Yes. So of course, it might be that the reason we see similar improvements is that we group the same population into high and low skilled groups based on a different test, and that the effect might be vastly different if you pick a random subset of the people. And then people who do forecasting for a living like truly high skilled forecasters. 

I think it's very plausible that the effect here is different. But most studies just take the same batch of employees or workers or study participants, and then divide them by some type of criterion, which is what we also did. And yeah, we did not find an effect. And similarly, we didn't find the disparate effect on question difficulty as well. So we expected that, maybe participants are more likely to just defer to the AI on hard questions and easy questions to do it themselves or something different. And there was also no significant effect as well.

Ross: Right. So you mentioned before that the subjects would use the AI very differently. So it may not have been specifically part of your research. But do you have any indications of the types of ways in which using the AI created the most value or augment the decision making or forecasting the most? 

Philipp: I wish I did. I looked into it. I just struggled to come up with a very strong and defensible method, especially after having seen the data. So I typically like to write down exactly what I'm going to do before I see the data to kind of avoid a contamination of what I think I should do with the results. But I think at least on some questions, people just didn't seem to benefit from getting an anchor. Some of the questions are really difficult. There were like, you know about Bitcoin hash rates, commercial flights on a certain day. That's not something one has a type of intuition about. I don't know how many flights are there, globally, or at any given day, especially at the end of December, I could be off by orders of magnitude. 

And I think one of the most simple, helpful ways to model can help is just give a prediction that is within one order of magnitude most of the time. So it's a starting point. Yes, that's right. I'm talking 10s of thousands, or millions, or like, what are we talking about? And I think, especially on like, harder questions, those were, I think, harder questions generally, like, ‘How many AI papers will be published in a given month,’ It's difficult, difficult to know if I was researching. And I think one big improvement here is simply the speed where, of course, people could go online, they could search the terms, they could try to find a source, they could double check it, but it will take half an hour to an hour. What was the simple interaction of the model in seven seconds?

Ross: And is one of the other ways as you said, interrogating the AI? So in the sense ofI suppose a couple of frames. One is, you know, ‘This is what I'm thinking, Do you have any other ideas?’ And the other one is around identifying different criteria, which may affect the outcome which the person may not have considered?

Philipp: Yes, absolutely. We didn't see this in the majority of interactions, but there are definitely people who did use it like this. And I think especially once you move to more sophisticated contexts, where people have, like a higher investment in the outcomes, I think this will most likely be the other kind of margin, at least, the most successful way to be augmented by to have the back and forth of one's own kind of use and points but and also take the outside opportunity to see like a model prediction and get feedback on one's own arguments.

Ross: So you've already spent a moment describing the super forecasting prompt that you use for the research. And so it's quite a long prompt, as you say, it mentions the 10 commandments of super forecasting, and provides quite a lot of detailed guidance on how to interact and describe probabilities and so on. So I'd be interested to know how you came up with that particular prompt? Did you try many types of prompts? What was the kind of testing to be able to provide this as an optimal super forecasting prompt?

Philipp: Yes, so it's clearly not optimal, so that they didn't run independent and critical analysis to make sure this is indeed, the most optimal. But I think the first step that anybody who tries to interact with is about forecasting, especially a couple of months ago, really was that many models just did not want to give forecasts for the future. And they had, like an aversion it's unclear at which point of the model pipeline this was introduced, but an aversion to providing probabilities about future events, they were generally very hesitant to give probabilities, or even specific quantities as a study. 

So the first part was simply drawn from a previous paper where we spent a lot of time trying to figure out how to consistently get GPT to give a forecast. So this has to work, you know, simply asking, just like naively, that doesn't work. And then we basically drew on the literature of super forecasting and tried to supplement that approach with what we thought in humans would be the most appropriate and most promising approach to think about future events.

Ross: So have you tested a variety of different types of prompts?

Philipp: We've tested a variety of different types of prompts on the outcome, complexity and helpfulness, not accuracy. The main idea here wasn't to get the most accurate forecast. The main idea was to get if you're, if you're prompted, respond in the way that we would like a super forecasting assistant to respond, right? If you're in for a prediction? Do they give you a prediction, and also give you the reasons for and against if you ask them now to explain. Do they give an explanation if you give your forecast to take the ticket? So yeah, we will get like a trial of a bunch of different prompts to see which one most mimics the type of assistant behavior we thought would be most useful for our treatment.  

Ross: I am just interested,  have you tested the other – the major, large language models to see if there's any differences in their propensity or ability at forecasting?

Philipp: This might be I don't know when this episode will air so the paper might be out by then. But there's a paper where we do exactly that. So what I call the Wisdom of the Silicon Crowd paper, which is where we try to mimic human cloud forecasting, via 12 distinct language models that are very diverse that are interconnected Quinn, seven B, from Alibaba, and of course GPT4 and Claude, and Palm and everything else. And we have every model give several forecasts on over 30 questions, and then we aggregate them. And then we actually find that the crowd of MLMs matches human forecasting performance. And this is the first time I think, this is salt has been found that if the large language models themselves form a cloud, they can they can hit the gold standard of a human forecasting tournament, which is even higher, because it's very interesting experience people forecasting there, 

Ross: Extremely interesting. So in that case, was the aggregation a simple mean, or what is the structure for aggregating the different models predictions?

Philipp: This is one of my favorite findings in the forecasting literature generally, is that like, yes, there's many ways to do really fancy aggregation methods. But a simple median is extremely powerful. And this is literally So the median is not the average. It's just taking the mean, right? Yep, just a median. And this is just extremely powerful across different contexts across different deviations from ideal scenarios.

And that's also what we use here. And, of course, there's massive heterogeneity. So diversity and how well models do some models do really badly. Don't call it the worst one. But so some models are very prone to forecasts of 99% or 1%. Right? They just think, like things happen or don't happen, whereas other models are more in the middle. And we also find that across the large language models, what is possibly something like what's called an acquiescence effect, which is the effect that whatever the question is, the model is more likely to say Yes, than No. It doesn't matter what the question is. And we find that the cloud overall is more likely to be on the side above 50% on the forecasts, despite the fact that less than 50% of questions that solve positively. So there is really a bias in that response. But nonetheless, the cloud effect still matches human accuracy and exceeds a simple benchmark of just giving 50%.

Ross: Extremely interesting. So, you mentioned this in some of your papers, but I mean, just what we've already touched on them in a way, but what are sort of short term and medium term research directions in the space of forecasting and decision-making and augmented with AI? 

Philipp: So that's, that's a lot of things happening right now. Many, many people are working on this. I think what I'm most interested in working on is currently the AI plus human and human plus AI interactions. And to see like, these first papers were like a stab at it to see. Yeah, the effect is real, it works. But now I think there's a lot of work to be done to more closely and more specifically look into what exactly is it that improves these performances. So for example, we like the paper we discussed first, it was humans being augmented by AI. In the second study on a different paper, AI is being augmented by humans in a way. So we have AI that predicts a bunch of different effects. And then they are being told what the human set says about these topics. And then they're being told, ‘Well, here's new information for you. A human tournament gives this 45% chance, you are now free to update however you want.’ And we find that actually the AI predictions get significantly more accurate, after learning. 

Ross: Fantastic afternoon.

Philipp: But there's a small caveat. This effect of improving the accuracy is less effective than if one had just taken the machine forecast and the human forecast and averaged them. Right. So there's still a bias in the model towards, towards I think its own views, and it only updates somewhat towards the human. And it doesn't properly distinguish between when the humans might be better than them and when they might be better to be, you know, relying on their own predictions. So you know, there are improvements, but they're still not above simply averaging. And I think just getting that right getting the what read improves model performance from humans, what improves human performance from AI? Is it numbers or is it maybe maybe it's not numbers at all? So I'm especially interested in seeing just very fancy and complex rationale. So reasoning for forecasts without having the numbers to see if that can improve performance, because that would eliminate the worry that people are just copying the new model. 

Ross: Yes, yes. There's a lot of rich aspects to that, including what are the mechanisms and structures for bringing together human and AI insight and sequencing and structure. But as you say, it's often the simplest that can come up with the best result. 

So switching gears a bit you are, amongst other things, a professional forecaster. So you're one of these humans, you're at the high end of the spectrum, in having skills, techniques, capabilities and performance, which exceeds others in being able to forecast extremely complex events. So what do you think about this? How have you developed this capability?

Philipp: That's a good question. I think the first caveat here really has to be that having worked with so many other experienced professional forecasters, there is no one size fits all answer, I think, we share some characteristics and backgrounds and methods. But I think in many ways, we are quite distinct from each other. So historically, the way I started it, I just got extremely interested in forecasting after the COVID pandemic, because I was so good at forecasting the very beginning and so bad at forecasting the middle, that I had two data points of like a great success and a great failure. And I was like, well, am I good at this, I'm really bad at this. And so then I signed up to one of those platforms, metaculus to get basically a track record. 

And so I went on, hundreds and hundreds of thousands of predictions at this point. And I think the main thing that really helps, is the distinction between outside and inside view. So if people haven't heard that before, the inside view is basically one's personal opinion about things. So if I just think about the chance of Donald Trump winning the presidential election in November, I might have my own kind of use. But then there's the outside view, which is a map view polls, better people that might be accurate forecasts. That might be track records that might be how likely a person is to be a president who didn't win the election, likely to win the election after that. And I think the most important thing in forecasting is not to not stick too hard to both ones inside view. And also not to always defer to the outside view. So I think the main challenge here is to find this balance between where, yes, actually, I probably should just defer to what other people are saying. And also to be on this kind of point of view, I think you have an advantage that really adds to it. And I should stick to my guns and be like 10% above or below? What would be the you know, the baseline expectation.

Ross: I've taken, for example, Donald Trump being elected president at the end of this year, so how would you go about it? I mean, is there a sequence of things where you consider the different factors or you take an external input? So or do you build a sort of a structured process to be able to start and then get to a point where you have a forecast? 

Philipp: For exactly this question, a second in response to the previous question. First, one feature of experience forecasts is also knowing when not to forecast, and I was invited to a project on a Donald Trump election. And I chose not to forecast on this, because I thought I didn't have a good enough procedure to add, like a lot of expertise and forecasting accuracy. And so actually, for questions where one probably doesn't have a good edge, but doesn't have like an additional part of knowledge, or a good track record, or a really high level view of all the information, it's probably best to not forecast, like it's an easy way to jumping into things that I'm probably isn't best suited for.

Ross: So let's say it is a subject. So whatever subject is something which is in your area of expertise, or you feel you have something to add. So at that point, do you have a process or approach in working through this challenge?

Philipp: Perfect, yes. Step one, for me, is to always get a very broad view of what everyone else is thinking. So part of the reason I enjoy forecasting so much is because one gets to work on so many different topics from Chinese coal consumption to climate change to financial markets. So the first thing I do is to try to read as much as possible, and to get as many forecasts predictions and rough numbers of base rates in this context as it's possible so often, when the thing one might want to forecast isn't quite the same that what other people are working on, but just to get a rough picture, and then to basically kind of construct what would the number be like if it just continued as usual? So what's the actual trendline? And then to ensure that deviations from this trendline have to be justified to myself, quite specifically? 

Because, very often the future is definitely like the past, of course, sometimes it absolutely, it's not. And those change points are very hard to forecast. And also, most experienced forecasters build a track record in environments where yes, the future is, is something like the past, we can sample, we can't sample the AI revolution a thousand times and see who gets it right. We can only sample repeated elections and economic indicators. So I think my bias here really is, you know, trend continuation as a first step. And then try to identify biases in individual people who might hold that trend, or who might argue for deviations of that.

Another kind of thing to look at is when evaluating sources, try to go back to those sources, previous views and predictions, so often, they're in there, but something like the International Energy Agency. Sometimes they do publish the forecasts, and I forgot what the type of graph is called. But it's a very, very striking graph for interest rate predictions, where the actual interest rate and the predictions at every year five years out, are shown and and I think it's called, like a hair haircuts diagram, something like that, where the predictions get it wrong, almost every time and to basically try to identify where those biases exist, and in what direction they are. They're optimistic about climate change, are extremely pessimistic, and then try to kind of account for the underlying bias in the trend. And that I think gives me a first kind of basis. And this can be done via…you can do your own time series, modeling some machine and stuff. But it can also be purely judgemental, just like in terms of numbers, especially where there isn't much data to go on – one can fit a model to something where it has like, three data points.

Ross: Yeah, a couple, I mean, a couple of points there. One is that if you are starting from other people's forecasts, I mean, basically, there aren't very many good forecasts out there. They're either commercially biased, or they're just not what many people are really trying to do and publish as forecasts. So there's not actually a lot out there. But it's another jittering point saying, you start with other people's forecasts, rather than sort of starting from the inside view, Full Movie, the outside view. So I suppose at that point, what you're trying to do, as you say, is to find what are all the failures of the existing predictions, you can add some value. That's right.

Philipp: That's right. And that's, of course, other people I've worked with, will do quite the opposite, they have to have their own view about how the world works. And then that standard inside view, and then use the outside view to supplement it. But I think this is, this is not the way I operate. And of course, they've been successful in doing so. But I really think that one can learn a lot from actually reading all the data and getting all the insights from all the different areas, especially on most projects. I work on it for 20,40, 50, 100 hours only, for a whole kind of context. And I think one will miss a lot by just going on intuition, because, you know, my intuitions are in my current contexts, decent probably. But there's a lot of things I don't know about. And I think it's very good to continue to be humble, and to just try to get a trendline going and stick to something like that.

Ross: So let's say you're speaking to somebody who just needs, you know, be useful to make better predictions in their work. Leader – business leader, startup leader, whatever. So what will be our advice, what are just a few things that they should start doing, which will make their predictions that better than they used to be.

Philipp: One advice is just don't think you're too special and get a view of the base rate of a trendline. And the second thing is, try to find a way to get experience forecasts with good track records to help you. So this can be by a business that offers like, like a swift center I work for, but this can also be internal. So this could just be an internal forecasting competition on the stuff that really matters to your business. Keep a track record of this and try to identify over months and years who's like the three best people we have on this and then make sure to draw on them going forward. 

Of course, this is very risky, because these types of internal competitions can end seniority hierarchies very quickly, if a junior analyst who stayed out of undergrad just turns out to be better than everyone else. But I think just identifying the people who actually could do this consistently. And in the context you cared about, I think it could be very useful for most businesses that have at least medium size, what I can kind of think about holding an internal competition like this.

Ross: Yeah, I think that's a really good idea. I did recalls, one of the early enterprise crowdsourcing examples was Google used for sales forecasting. So essentially, using a crowd, they found that, significantly better than all of the sales forecasts they had.

Philipp: So there's a lot of interesting work right now on our prediction markets, better forecasting tournaments. And I think, you know, many people might get sidetracked with what should focus too much on this, but as the recent work just shows that actually, the most important thing is just getting the experienced forecasters, it doesn't really matter. If it's a prediction market or forecasting platform with just just a monthly survey. I think the biggest bang for the buck really is identifying the people who were most equipped to forecast that event rather than the context of any business organization.

Ross: Fantastic. I will certainly be following your work closely. I think it's fascinating; a really interesting paper. Sounds like the one which is about to come out. I'll definitely be diving into that. So love your work. Thank you. Is there anywhere people can go to find out more about what you're doing?

Philipp: Thank you so much. So I'm mostly on Twitter, on social media and also my website. But I think Twitter and websites are probably the best. The best spot.

Ross: Right? Fabulous. All those will be in the show notes. Thanks so much.

Philipp: Thank you so much.

The post Philipp Schoenegger on AI-augmented predictions, improving human decisions, LLM wisdom of crowds, and how to be a superforecaster (AC Ep36) appeared first on Humans + AI.

– Bryan Cassady

Bryan is the founder and director of the Global Entrepreneurship Alliance, a foundation with a mission to coach or train 1 million entrepreneurs by 2027. He has built 8 successful companies in 6 countries. Bryan has taught innovation and entrepreneurship at numerous leading universities around the world, and is author of the book CYCLES.

Website: www.bryancassady.com

LinkedIn: Bryan Cassady

X (Twitter): @bryancassady

AI (Artificial Intelligence)

Idea Generation (Ideation)

Idea Evaluation

Market Scoping

Jobs to Be Done (JTBD) Framework

Brainwriting Technique

Book

CYCLES: The simplest, proven method to innovate faster while reducing risks by Bryan Cassady

Ross Dawson: Bryan, it's awesome to have you on the show.

Bryan Cassady: Thank you for having me. I'm glad to be here.

Ross: So you dig deep into AI innovation? So what's the premise of AI innovation? What's it mean? And how do people start on that journey?

Bryan: Innovation is really, really important. It's what makes the world turnaround. And the question that comes up for me time and time, again, is how can you be more effective in innovation? How can you be better, faster, do it easier. And I came across AI about a year and a half ago. And I'm just amazed at how much better things can become. And my personal take is trying to find the facts that back it up where it works, and where it doesn't work.

Ross: So let's say as a starting point, go to an executive team, they say no, right? Innovation is important to us. We've got some processes now, what would be the first steps in being able to introduce AI into their innovation process?

Bryan: I think, you know, for me, the challenge is to have two bits of knowledge, one knowledge around innovation, what is innovation? What are you trying to do? What are the facts? And then secondly, its domain expertise in terms of AI, and you have to bring the two of them together. And the first and foremost, most important bit of knowledge around innovation is AI works when the system works. It's a system driven process, we tend to look and say, no, no, it's those people. They're not being creative. They're not doing what they want. But if innovation is not working in your company, it's what you as management is doing. It's not what your people are doing wrong. 

And the second thing is to look at innovation as a process. I mean, it's, you have to do a lot of things, right. It's not enough to do one thing, right? And everybody seems to focus on idea building. But idea building is actually the easiest part of the innovation process. And I think what I would look at is ways that you can use AI to get aligned better, that you can communicate better, that you can pitch better, that you can evaluate ideas better. And there's lots of cool stuff that can be done there. And I hope we get a chance to share some of the cool stuff we've been doing around this and in the next few minutes. 

Ross: Absolutely. Well, I want to dig deeper on a lot of levels. But I mean, to ground this, can you give a specific example of how AI has been introduced in being valuable in the innovation process of an organization?

Bryan: Well, you know, everybody looks at ideation as the core of innovation. So I'll start there. That's because that's what people are interested in. And there's a lot of research showing that if you take a typical person, you give them AI to use to become more creative and more effective. And the impact is biggest on the lowest or lowest performing people in your organization. And sort of it's an evening out factor. But what people forget in the process is you have to use it effectively. And we just completed some research, we evaluated 5400 ideas generated by humans and AI. And we found a few things that came out. One is AI, on average it doesn't build better ideas, AI builds a lot more ideas. But when you take AI and humans intelligently together, your hit rate goes up, and your hit rate goes up amazingly. 

And hit rate I mean is what is the percentage of really good ideas. So if I look at an average human, we hit one out of 100, out of the park. If we look at a human's plus AI, without any training, they hit about one and a half to two. But if you get humans plus AI, and some good training, you're hitting somewhere between five and seven. So this is five to seven times more ideas that are really good ideas. And that's a big difference. 

Ross: You said something to the effect of bringing together intelligently, humans and AI, and also about the training. So what are the ways in which you can bring humans and AI together? What are the sort of steps or processes or where that's done? And if what is true, what's the training, which enables people to do this better?

Bryan: So I think if I was going to start out a training, I would look at the place where innovation fails most often. Innovation feels most often because you're solving the wrong problem. And in fact, AI can help you an amazing amount of trying to identify what is the right problem and defining the problem correctly. It's and the advantage you have with AI is it can give you sort of a naive viewpoint on your problem that you didn't think of before. And you know what I see there is an incredible operator to start at the beginning. And just to put a framework here, I talk about innovation in terms of the ABCs. It's ABCs, because it's easy to remember, maybe your listeners remember, it's to align, build, communicate, check, systematically improve. And at the end of this, you've got, you'd have to pitch your ideas for. And what I look at is ways that AI can be used in the whole process more effectively. 

And the first part is alignment, figuring out where you want to go. The second part is building ideas. And by the way, this is the easiest part of the innovation process, but AI can really knock it out of the park. And for me, you know, the superhuman characteristic of AI right now is the ability to evaluate ideas. I just talked to you a second ago about a study with 5400 ideas. Imagine what would happen if you went back to your company and said, you have to evaluate 5400 ideas. AI is not going to complain, your people are gonna complain a lot. And the other thing that comes up within the innovation process is if you want to get ideas moving forward, you have to present them effectively. And I actually do a really good job of taking human ideas, and making them better structured, and easier to understand for other people. And then lastly, you know, the part at the end is, how do you get these things in the world? How do you look at where the weaknesses are? And AI is very good at identifying system problems, to where things don't fit together? So that was a long answer to a simple question, Ross.

Ross: Well now, what is the structured one so that it makes it cognitively tractable? So the first step was a line. So the way I put it is framing. So when you start off with a human plus AI process, the first step is framing your objectives, challenging the context. So, when I usually say that is mainly a human role, because the human understands the context and the objectives and the frame? And so what, how do you go about the Align process is that mainly a human in terms of defining the terms of the engagement as it were, or AI and humans work together to be aligned to ensure that when you are building ideas, and assessing them, and so on, that they are in the right context?

Bryan: I wish I could answer this completely. It's sort of a moving target right now. I've spent the last six months building an AI for innovation toolkit. And within that AI for innovation toolkit, there's 40 tools around alignment. And those tools are typically around market scoping. You know, what is the market we're in? What are the needs? Which are there? What are the jobs to be done? And what you're trying to identify as, first of all, are you pointed in the right direction, creating a product that somebody wants? And then at the end of the process, are you communicating where you want to go in a way that your team can understand? And one of the challenges that comes up, I can only identify qualitative ways that AI is helping in this process. But I'm finding it really hard to identify quantitative waste, how can you prove that the team is more aligned. And if you have some suggestions for me, I'm all in for it. But I think what's important here is what you're talking about is the framing of the problem. If you solve the wrong problem, you get the wrong solution. And AI can be your friend and identify what solution you should be looking at. And how do you get yourself pointed in the right direction? So I mean, to answer your question here, it's hard. The alignment heart is a heart heartbeat. And I do think that remains the domain of the human. But the domain of the smart human is to use AI to augment what they're doing, and to think more effectively. And, you know, for me, typically, when I do an alignment project with a company, it was taking me a day and a half to two days before. And now with AI, we're doing it in a half a day. So it's not just qualitatively better it's quantitatively faster along the process.

Ross: So as you said before, around the ideas and the idea of filtering, you can generate an unlimited number of ideas. And you can use AI to assist you in the filtering and you know, some kind of ranking or assessment. They're also in that aligning price as you've talked about things like market scoping, or adjacencies. And there's a whole lot of useful prompts or tools which you know, really are about looking at what is the position of the organizations? What are adjacent opportunities? What are different ways of extending current capabilities, for example, but again, we start to often be quite verbose, or again, proliferation of ideas. So let's say we've got a strategy team and the building, you might want to bring more people into that process as a broader brainstorming ideation process. But in the aligning, it's probably a bit more efficient to have a strategy team or executive team. So how might you then use some of these tools to generate some of the scoping or alignment or framing issues, and present them to a strategy team so that they can efficiently and usefully get a better understanding of where they are? And where opportunities might lie?

Bryan: I think, within any team, what you need is diversity of viewpoints and different perspectives. The one thing you can't ask a CEO, at least not very effectively, is why is your company going to fail? Or why? Why is this project going to fail? And one of the things that I see very powerful in using AI is to be that questioning that Doubting Thomas within the room and saying, ‘Look, are we pointed in the right direction, find me holes in what we're doing.’ And nobody wants to kill the AI. But if this was a human person in the meeting room, they certainly would have all the weight of the world on their shoulders that they're talking about all the reasons that things would fail. But I would say, for me, the role of AI, it's sort of like, you know, when you were a kid, you had this telescope, and you're sort of trying to get it more clear view of where you're going in, it's what you're trying to do is bring in more light. And as you bring in more light, you know, the direction you're going becomes clearer and clearer. It doesn't give you the answer, because you're still left with a human to interpret what you're seeing. But it certainly adds clarity and depth, and features that you might not see within your typical management team. And the answer here is using AI just to be smarter than you were before, to question better than you did before. And, you know, I guess for me, one of the things that came up is my story with AI. I just didn't believe it could do these things. But the proof is in the pudding. When you actually start using it, you said, wow, wow, we can do this that I didn't expect. So, you know, I think the process of using it is to add clarification and ask the tough questions that a lot of humans don't want to ask.

Ross: That's really interesting. So I've, I usually frame the devil's advocate, the red teaming challenges and so on at the end. So you might have a strategy or you might have an investment decision or some kind of decision, which you can then you know, AI is extremely useful, as you suggest for being a, you know, hopefully not emotionally – It doesn't create emotional responses, but can actually, you know, quite succinctly and well articulate the reasons why a particular decision might be wrong. But I mean, you're very interestingly raising this idea of saying, well, take where we are now, in our current position, and to challenge that as a starting point, rather than simply being able to look at that as a set of challenges at the end of the process. 

Bryan: One of the things that I've seen is very effective use of AI is, you know, if you have a management team, and for example, you have them do brain writing. So, you know, in a very effective part of the innovation process to have people get ideas off their chest, and you just have them all type games, what are we focusing on today? What are we focusing on today? And then you feed that all into the AI. And you say, Aha, try to imagine what is John thinking, what is Mike thinking, and you look for the common themes. And it's, it's amazing the speed which you can go through, and it's this process of interaction where you have people doing deep thinking and then the AI summarizes and synthesizes it in a way so that people can even think deep. So it's not letting me think for you, but it's helping you in the process.

Ross: Yeah, I've often used recording so basically get a group together, record it, and then immediately transcribe it and do an AI summary. And so people can basically see what it is they were talking about live or potentially pass that on to other groups to bounce back. And so that actually if you're doing that live, it can be a very powerful tool for, for being able to, you know, distill that but one of the other points coming out of what you're just suggesting is this idea of clustering or categorization. So classically, you stick to posting it up on the whiteboard. And you work out where you know how they make sense together. But that's actually something which AI can be very useful for, as well as be able to say, well, here are different ways in which we might cluster these different ideas and how they might emerge with some kind of sense of the idea of scope we're generating.

Bryan: I think, the clustering of ideas, and also the identification of divergent ideas. Because if everybody's thinking the same thing, a lot of people don't need to be there. And what you need to do is capture, what are the divergent ideas in the room. And, you know, go down that path for a bit and see where you're headed. And everybody gets smarter along the way.

Ross: So you're, you're suggesting that AI assesses how divergent ideas are as part of the process. 

Bryan: I think one of the things that comes up is there is a degree to which you're looking for convergence among people. But you're also looking for the outliers. And you're trying to say, who has a different point of view? And then you ask, ‘Ross, why do you have this point of view? Why are you seeing this differently than the rest of us?’ And, you know, it might not come out in a typical conversation, because we lose some of those nuances when we're listening to one another. And we tend to hear what we want to hear. And yeah, AI has an ability to actually hear what we do here, what we're really saying a lot of times more effectively than our human ears do.

Ross: You one of the tools I really like doing with boards is to use live polling, to identify the range of opinions. And you can see, well, are they clustered? Or who's on the outside? And then you can sort of say, well, who, who has that very different opinion to everyone else and why and that's a really great way to surface conversation.

Bryan: But I think, you know, what happens is I see innovation as a process where you go from step by step, and the starting point is alignment. And alignment is defining the problem. And if there's one thing that's proven within the innovation process, and proven with the creative process, the most important part of getting good ideas is defining the problem. If you don't define the problem, right, you get the wrong solutions. And finding the problem, right, actually can define half the solutions that you find. And I think, using the clarification, using polling, using different things that bring ideas out, you get much better than you would just sitting there by yourself. 

Now, the challenge that comes up there as a human, especially if you're the consultant in the room, I mean, do you really want to give away the power to AI at that point in time, because you want to be the smartest person in the room? Or at least I think most consultants do. And, I think there's a certain degree of humility, and realizing, in fact, that this buddy in your room, this extra team partner, which is AI can do certain things that you might not be able to do more effectively.

Ross: Absolutely. Right. And so I actually say I don't do consulting, I facilitate. And so I facilitate people and now facilitate humans and AI together. So it's facilitating the process. Yeah, as you say, it is a process. And I think, the role for people like you and I and our peers is to how do you enable that process of smart people together with extraordinary tools to flow in a way because you can't map it out mechanically you know, it's an enabling process.

Bryan: So Ross, I know you're the one asking the questions today. But I'd like to turn it around for just a second, where have you seen the biggest impact in AI getting teams aligned?

Ross: Getting teams to work together better?

Bryan: Working together better, but also just to decide what they're working on?

Ross: Oh, well, in terms of working better together? I haven't actually seen that. That's a really interesting use case. I mean, actually, there was an interesting study recently where I was generating, basically being contrarian and arguing other cases. And for some people, it was useful. And it's a little bit similar to the cognitive abrasion, creative abrasion with people where it can be useful to have different opinions. And I think there may be a role for AI in that, but I haven't personally experienced that. But the way, I suppose the starting point would be seeing that I do start with that human frame from the beginning whether humans do the frame and start off by saying this is what we wanted to embark on. This is our context. And even just a preliminary definition of that. And that's when you start to get other opinions or perspectives. And actually one of the most useful ones I found is that creating diverse perspectives from AI, so to generate a whole array of different personas that are relevant or possibly even, not even relevant, to be able to create things and that just being part of that journey, but my I practice has always started with the human, and then be able to get in those other perspectives or other ideas or framing from AI.  

Bryan: While you were talking, I was thinking, Where do I see the biggest impact? I see it as management usually thinks what they're presenting is very clear. And one of the things that I see is very positive is AI can come back and say, This is what isn't clear. And to give them and feedback in a non confrontational way, to make sure that what they're explaining to their teams, or their marching orders have to be very specific and clear, if they hope to get people marching in the right direction. And I think that might be, at least in my opinion, where AI can be used most effectively. clarification.

Ross: Well, that goes to the point of AI as interactive agents. So you can have a lot of single prompts, lots of people, you know, sharing all this prompt for this and prompt for that. But you know, often, if you guide the AI to be interactive to ask, keep asking questions in detail until it has enough information and feeding that back. That is one of the most useful things and it can feel frustrating, you don't want to set up an AI which asks you these questions which take forever to answer. But I think this dialogue partner has been able to say, these are the missing bits of information, can you feed that in? And that actually the human thinking that through to answer those questions, as you say, it can be one of the most powerful engagement techniques.

Bryan: I think the multiple prompts are actually quite smart, because it forces clarification on both sides. And, I think a lot of times, we don't think deeply enough. And what you're trying to do is trying to find ways to think more deeply and more effectively. And this gets you there quicker and easier.

Ross: So you mentioned before this idea of evidence, looking for evidence and variables to find the best path for AI innovation would love to know what you're thinking or processes or approaches to be able to build evidence into creating better. 

Bryan: That's a good lead in because that leads me to three things that I'm doing right now. One thing we did, is we did a lot of work generating ideas with AI and humans ai, ai alone, humans alone human plus AI. And what we found evidence based, is in fact, there isn't much difference in terms of the average quality of ideas. We found out in fact, humans are more novel than AI is. And you'd say, well, you know, at this point in time, you should give up on AI. But in fact, what we found is something else, which is more important. And in fact, in the innovation process, what you're looking for is outliers, you're looking for great ideas. And what we did is we rated all the ideas generated and identified top 5% ideas, and what is the percentage of time that an idea falls in the top 5%. And so we have evidence based information showing that in fact, humans plus AI in an interactive process, like you were talking about, generate somewhere between five and seven times more top ideas. That, for me, is an important finding. 

The second thing that we looked at in terms of evidence based work is a mount. For me, everybody looks at the hero as building ideas. But when ideas become really easy to generate, the hero now becomes your ability to evaluate ideas. So if you can imagine we have 5000 ideas, who's going to evaluate them? And is the AI any good? So we ran a second set of studies, looking at human evaluations of those ideas versus AI valuations of ideas. And what you can find out first of all, is a lot quicker, that's not a big surprise. But more importantly, AI is much more reliable. So you present the same idea twice. Your odds of getting a different evaluation from a human are dramatically higher than with the AI. But the thing that came up most interesting for me is in fact, if you use a collection of AI personas to evaluate the ideas, the AI personas take sort of a middle ground. And they're actually a better representation of the whole universe of evaluation than any single human. So in fact, the humans are less correlated with one another than the humans are correlated, meaning that AI is not only a fast way to do it, it's a reliable way to do it. But it's also a more valid way to do it. 

And the last thing that we're doing research on and this For me, it's because I'm a teacher, I teach courses. And at the end, of course, my students do pitches. And I've suffered through so many horrible pitches. And my best estimate, we were at a 40 to 50% fail rate on the pitches. And I hope that's not an indication of me as a teacher. But in fact, what we tried to do is we tried to find ways to improve that pitching process. And we've now put AI in it. So we use AI, to help people think deeply about their ideas to make a first version of their pitch. And then the humans have something to work with, it's no longer a blank sheet of paper, they work on it. And then we created a second tool, which evaluates those pitches. And we've mined from a 40% fail rate to a 5% fail rate. Now, if you can imagine companies doubling the effectiveness of your communication, that's an amazing thing. Now the question is, how do you present this quantitatively? And I do a presentation in June, we actually have to present the results. So I have to start writing quickly.

Ross: Fabulous, no, really looking forward to that. So to round out, let's say, Yeah, whoever was listening at the moment already understands AI innovation has embarked on that journey at the start, you know, they're familiar with the ideas we've been discussing at a high level, and they are beginning to get going. So what would be your advice to them on how it is they can refine and improve using AI and innovation?

Bryan: I think one of the things you have to realize is the perspective in which we see things most of the people listening here are really positive about AI. And you have to realize a lot of people are not very positive about AI. AI, people don't want to see AI as the star of the show. And in fact, AI should not be the star of the show. In fact, the biggest advice that we give is to say, look, AI has an amazing amount of limitations, horrible and a lot of things. But when you use it smartly, it becomes an important part of your team. And as an important team member, your team gets better. So I would say my key advice to you is look at how you work together to deliver better results. And don't tell anybody that AI is doing the job. AI is working with the humans to do a better job.

Ross: Yeah, I think that's really important. The cultural aspect of how this works is fundamental.

Bryan: And it can't be the star of the show. It has to be another team. Actually, let me conclude with something that I found was. So before here, I used to work on group dynamics, how do you make a group effective. There is a guy named Scott Page who did a lot of research on diversity and group performance. And he found a really important finding. In fact, it's not the talent of a group that performs group performance. But as the diversity of a group. And to test this, one of the things I did in my academic studies is we took the best students, and we added the worst students to the group. And you would think, what's going to happen, it's not going to help them at all. In fact, when you add diversity, even if it's somebody less smart, even if it's somebody less good, the results go up. So what you can do is say look, AI adds diversity, it's a different type of team member. And what you need to do is round out your team and get the full team together and put another person at the table and everything will go better.

Ross: That's a great way to frame it at the highest level. I think that's really important. So Bryan, where can people go to find more about your work, I think you've got quite a lot of things you share.

Bryan: I share a lot of stuff with you. The easy place to find me is on my LinkedIn. Bryan Cassady, I'm the only one there. In fact, there's two of them there. One is my old profile, which I forgot the password for. And the second one is my site, which is www.bryancassady.com. 

Ross, I really enjoyed the chat today. I learned a lot while you're asking me questions, and I appreciate your time. And thank you very much.

Ross: Now, it's fabulous that you're one of the people out there pushing the potential for AI and innovation. I think we're a growing community. So thanks so much.

Bryan: I would like to end with the last thing we have. We are looking for beta testers for our AI renovation toolkit. If anyone contacts me on LinkedIn and sends me a direct message, I will send you back a link and you can try it out. There's Jordan 65 tools and six apps and it does some amazing stuff.

Ross: Fabulous. Thank you, Bryan!

The post Bryan Cassady on AI innovation, Humans + AI idea evaluation, increasing diversity with AI, and evidence-based innovation (AC Ep35) appeared first on Humans + AI.