The Lindahl Letter

Don’t panic! We are going to take this topic in two different directions this week. First, the obvious path will be taken and we will talk about programming, code, and python cookbooks. Second, for those of you that were hoping this week will be a deep dive into artificial intelligence and machine learning derived recipes compiled into a cookbook you will not be disappointed either. Another way to talk about what we are covering today would be that not only are we talking python cookbooks, but also we are covering the epicurean adventures that could be created. 

Beginning with the first topic under consideration; coding related answers to the machine learning cookbook question are real. What exactly is a Python Cookbook? If you are a fan of the work by David Beazley and the 3rd edition of the Python Cookbook from the publisher O’Reilly, then you might already have an idea [1]. I personally did not have a copy of this book on my shelf or any of the other python cookbook compilations. These are compilations of code brought together like recipes to help people learn how to use and deploy Python as a computer programming language. You can get to all the code examples that David Beazley has shared on GitHub [2]. For those of you looking to dig into some Python code and learn about how it works, a collection of recipes compiled into a cookbook is a decent place to start. You might get lucky enough to find some recipes on the specific thing you are trying to solve. The other book that stood out to me was the Modern Python Cookbook from Steven F. Lott [3]. The code from that book is also published on GitHub for easy access [4].

At this point we are pivoting to the second question about actual recipes. You might remember back to 2017 when Eric Schmidt shared an AI created cookie recipe built out by engineers at Google as part of a real world challenge [5]. You can find the recipe for the “chocolate chip and cardamom cookie” via the link above. Getting to that point apparently took 59 recipe iterations which is interesting. Oddly enough that article also brought in some of the warning elements that Eric Schmidt has shared about the potential misuse of AI. Cookie related AI examples have improved since 2017 and you could check out the Nestle Toll House cookie expert or alternatively the “cookie coach” AI now to help answer your baking questions [6]. It got announced on Twitter and they had a fun video about working with Ruth the cookie coach. 

It does appear that Ruth the Nestle Toll House cookie coach really is still online and working over at https://cookiecoach.tollhouse.com. From Eric Schimdit sharing recipes to interactive chat experiences people have really opened the door to recipe driving artificial intelligence when it comes to cookies. At this point it might be a good idea to move beyond cookies and look at recipes in general. You could go out to the Google Cloud blog and learn about “Baking recipes made by AI” from December 15, 2020 [7]. The embedded video digs into making recipes with ML and is a 7 minute journey of fun featuring Sara Robinson. You may already know that during the pandemic Sara Robinson collected a ton of recipes and used a TensorFlow model to make predictions on future iterations [8]. Between the code shared on the Google Cloud blog and by Sara Robinson you can actually work on a similar cookie effort to make your own machine learning model to work on recipes. 

Our friends at Google are not the only ones that have started to wonder about using machine learning for the practical activity of creating recipes. The team over at Towards Data Science have walked through the good and the bad of the process as well within an article titled, “Using machine learning to generate recipes that actually work” [9]. This article really walks you through the process and it is interesting. People really do seem to be having fun with using ML to create recipes. Another way to take a look at the process would be with an article from KDNuggets called, “Generating cooking recipes using TensorFlow and LSTM Recurrent Neural Network: A step-by-step guide” [10]. As you can tell from the number of sources of examples on how to start making your own recipes with machine learning it is a topic that people really seem to be passionate about. Some of that passion is translating further into action. Consider the article, “Forage: Optimizing Food Use With Machine Learning Generated Recipes” [11]. Within that article Angelica, Elbert, and Brian take a real look at reducing potential food waste with machine learning. It is an interesting way to apply machine learning to a practical real world problem. 

My efforts to find a real cookbook for sale that is made up exclusively of AI/ML made recipes were not successful this week. You can find websites like https://cookbook.ai/ that help you search recipes and build meal plans, but I did not find a cookbook you could buy from a bookstore. 

Links and thoughts:

 “Sparse Expert Models (Switch Transformers, GLAM, and more... w/ the Authors)”

Christina Warren this week reminds us that the Download has moved over to https://www.youtube.com/GitHub 

“More WordPress Add-on Trouble”

Top 5 Tweets of the week:

Footnotes:

[1] https://www.oreilly.com/library/view/python-cookbook-3rd/9781449357337/ 

[2] https://github.com/dabeaz

[3] https://www.packtpub.com/product/modern-python-cookbook/9781786469250 

[4] https://github.com/packtpublishing/modern-python-cookbook 

[5] https://www.cnbc.com/2017/12/05/eric-schmidt-google-used-ai-to-create-the-perfect-cookie-recipe.html 

[6] https://www.mashed.com/348141/nestle-toll-house-just-released-a-cookie-expert-ai-for-all-of-your-baking-questions/ 

[7] https://cloud.google.com/blog/topics/developers-practitioners/baking-recipes-made-ai 

[8] https://sararobinson.dev/2020/04/30/baking-machine-learning.html 

[9] https://towardsdatascience.com/using-machine-learning-to-generate-recipes-that-actually-works-b2331c85ab72  

[10] https://www.kdnuggets.com/2020/07/generating-cooking-recipes-using-tensorflow.html 

[11] https://cs229.stanford.edu/proj2017/final-reports/5244233.pdf 

What’s next for The Lindahl Letter?

* Week 70: ML and Web3 (decentralized internet)

* Week 71: What are the best ML newsletters?

* Week 72: Open source machine learning security

* Week 73: Symbolic machine learning

* Week 74: ML content automation

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Both OpenAI and Hugging Face have teams doing great things with respect to machine learning models. Their delivery models are very different. You can visit OpenAI at https://openai.com/api/ and look around at the machine learning models being sold to consumers like IBM, Salesforce, Intel, and Cisco. Delivering machine learning models via an API is one way to go about publishing and sharing your work. Alternatively, people are publishing models like what is happening over at Hugging Face https://huggingface.co/models where if I’m reading the page correctly you can search for various models from a list of 36,028. Both of these organizations are delivering excellent machine learning content to a world of people looking to operationalize machine learning within their corporate strategies. Deciding to publish a model or to sell it via an API is a major decision to make. Selfishly, I much prefer the open source models that I can play with and download. 

Before we move on to the rest of this analysis please consider my full disclosure that I participated in the OpenAI private beta for both Codex and GPT-3. Both of those sets of beta analysis provided API access and not full downloads of the models in question. That participation may have given me a good idea of how the system works and let me kick the tires, but it did not cloud my judgment or make me want to give OpenAI favorable treatment. I do agree with the original assessment by the OpenAI team that the GTP-3 and GTP-2 models open the door to misuse [1]. Back in 2019 The Verge team noted that, “OpenAI has published the text-generating AI it said was too dangerous to share,” [2]. 

We face a very real possibility that the models could be misused to flood our information streams and that it would become almost impossible for communication to function. Some people already believe that bots and other flooding techniques to AstroTurf and falsely drive news cycles are already breaking a problematic news ecosystem. A truly asymmetric delivery problem exists when the amount of content being produced is massively larger than what can be consumed by an individual. Traditional media has transformed from the highly curated view newspapers both national and local mixed with nightly news broadcasts provided to near real time broadcasting. The level of curation within a 24 hour broadcasting channel is even fundamentally different from the single serving real time publishing cycle that happens online. While the topic of information flooding deserves an entire post of consideration especially related to the mechanics of how it works I’m going to move on to the ethics part of the question.

Ethicists have been debating the potential release of dangerous machine learning models for some time [3]. It is a serious debate that needs to be had probably at a governmental and ultimately international consensus level given the potential influences on civil society as a whole from a dangerous intersection of technology and modernity. You can easily provide a model like GPT-3 a prompt for a topic and it will very quickly spit out content. If you elected to do that over and over again for a nefarious purpose, then you could flood comments, posts, news, and other points of information. It is a truly great tragedy of the public information commons that takes the power of sharing information online and tips it to an extreme. 

Outside of the ethical considerations of these large language or foundational models. We are probably going to see the heavily used and curated machine learning models deployed via the API method of selling and providing accessibility. Reducing the friction to be able to access and use an API which is generally going to be curated by an organization that is handling all the maintenance and training has a certain value proposition going forward. You almost get to set it and forget about the ongoing cost of training, enhancing, and maintaining the machine learning model. Your machine learning return on investment model may very well allow for some additional cost per transaction within an externally sold API to get the benefits of speed to access and ongoing scalability. That creates an advantage for the biggest companies that can provide proven uptime and reliable service. My attention turned to looking at Google Scholar for “machine learning API marketplace” to see what publications surfaced [4]. A lot of the articles felt like pitches or introductions to specific technology. They were describing parts of the landscape, but were missing the bigger picture of what would happen in the overall marketplace.

Links and thoughts:

1. I watched Linus and Luke during the WAN show episode from April 8, 2022. They were super excited about launching a screwdriver, backpack, and maybe getting a second giant product testing studio. It's good that Linus is getting into the product testing part of the review space as that is an area where we need more focus and professional attention for the people to consume in general. Our ability to get independent reviews of a high quality seems to be shrinking over the last few years. We have seen a lot more focus on unboxing and reviews from people on YouTube as people are branching out to try to get unbiased and unfiltered insights into products before buying them increasingly online.

 2. This week in ML News Yannic Kilcher talked to people on the street and we learned about Google's 540B PaLM Language Model and the OpenAI DALL-E 2 Text-to-Image model.

 3. On this episode of Decoder Nilay Patel talks to Chris Dixon about a lot of topics related to web3. It was a very interesting hour of discussion about the future of the internet and what will happen online. 

4. Nilay and friends were really excited this week on The Verge podcast where they discussed a couple of topics including Mark Zuckerberg’s big plans for AR glasses, Google’s apparent lack of interest in AR hardware, and Elon Musk’s Twitter drama.

 Top 5 Tweets of the week:

 

 

Footnotes:

[1] https://www.wired.com/story/dangerous-ai-open-source/

[2] https://www.theverge.com/2019/11/7/20953040/openai-text-generation-ai-gpt-2-full-model-release-1-5b-parameters 

[3] https://www.theverge.com/2019/2/21/18234500/ai-ethics-debate-researchers-harmful-programs-openai 

[4] https://scholar.google.com/scholar?hl=en&as_sdt=0%2C6&q=machine+learning+api+marketplace&oq=machine+learning+API+mark 

What’s next for The Lindahl Letter?

* Week 69: A machine learning cookbook?

* Week 70: ML and Web3 (decentralized internet)

* Week 71: What are the best ML newsletters?

* Week 72: Open source machine learning security

* Week 73: Symbolic machine learning

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead. 

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Things got shuffled around a little bit and instead of hearing about, “Who still does ML tooling by hand?,” this week my attention shifted to writing down my thoughts on NFTs. This happened in part due to a lengthy conversation I had about the future of NFTs this week. Part of that conversation was distilled into this series of thoughts on the subject. This was almost an interview edition of the podcast. Given the length of my current topic backlog the topic being replaced may not surface again or will have to be incorporated into something else along the way as a sort of bonus topic. I’m on the lookout for the first mainstream use case for NFTs. To qualify that statement I’m going to describe mainstream as millions of users doing something (that something is the basis of what I would consider a use case) on a regular basis with NFTs facilitated by a marketplace. 

For those of you who wanted a quick primer on NFTs, then please consider watching the 8 minute video by The Verge team called, “NFTs and the $13B marketplace, explained,” from February 2, 2022 or reading the “NFTs, explained” post they have [1].

Earlier this month Mark Zuckerberg said that NFTs are coming to Instagram [2]. We are already seeing NFTs being added to Twitter profile pictures in a special way [3]. Both of these use cases for NFTs create a use case for the technology or the foundation for where a marketplace could exist with a sizable audience. That is the key element I have been waiting to see within the NFT space. People can 100% get non fungible tokens associated with a blockchain and keep that token in their cryptocurrency wallet. This whole process requires a bunch of things to work which is why I think having some of this background stuff  built into an application like Twitter or Instagram will help create a more mainstream experience. 

What exactly do you need to make an NFT work?

* Blockchain - You need a functioning blockchain that can be accessed to view and confirm ownership of the token. This is a really important thing as if the blockchain ceases to exist or becomes so slow from lack of operational activity your token could become worthless. It is a digital asset and is inherently ephemeral.

* Marketplace - The person needs a marketplace or some method of transacting the exchange of an NFT. This is where I think the mainstream social media applications getting involved will make this easier for everybody.

* Wallet - Somebody has to have a cryptocurrency wallet to be able to store the information related to the token and other elements for you. This is a collection of keys that are necessary to prove ownership of your token or cryptocurrency. 

* Minting - Within the process one of the parties has to be able to generate NFts and bring them to a marketplace or distribute them as gifts. I’m assuming that Instagram will give people the ability to make NFTs from pictures they upload. 

Right now as it stands somebody who is reasonably tech savvy could go out and buy, own, and trade NFTs right now. I don't believe that all of the blockchains being used for these NFTs will survive. I think a bunch of people are going to be disappointed when this happens. However, a bunch of people will probably have forgotten they have had the NFTs on that blockchain to begin with so they won’t be disappointed. Sure people are working on some interoperability between blockchains so you can move NFTs if you think your current blockchain might be on the path to ruin. All of that complexity will fade away when people are just using the ecosystem related to Instagram or Twitter to make the process frictionless. A process for NFTs that includes millions of people within the marketplace and zero barrier to entry in terms of technology will kickstart an actual economy around the content. It is also a lot less likely that Instagram will fade away. People will keep sharing photos in some form now and in the metaverse. 

Soon enough I’ll be focused on building out 8 sections of that machine learning syllabus. My plan for that effort is to really build out and potentially revise that syllabus into something really good. I’m talking about taking it from academic articles all the way to easy to consume lectures from other academics. My goal with that effort is to pull together the very best content from all over the place into a one stop shop for somebody to come up to speed on the subject of machine learning. I’ll put each one of them into a Jupyter notebook format and publish them on GitHub as well to allow people to do push and pull requests against the content. That seems like the most reasonable way to begin the sharing process in an open and earnest way. 

Links and thoughts:

1. You can watch this video about “NFTs and the $13B marketplace, explained” from the team over at The Verge. I enjoyed it and thought it was a good introduction. 

2. This episode of the Decoder podcast really digs into the possibility of what is going to happen with NFTs and crypto. Nilay Patel really gets into the future of things during the podcast, “Steve Aoki on why he’s a ‘crypto believer’” This was one of the most meta conversations I have listened to in some time.  

Top 5 Tweets of the week:

Footnotes:

[1] https://www.theverge.com/22310188/nft-explainer-what-is-blockchain-crypto-art-faq 

[2] https://www.engadget.com/mark-zuckerberg-confirms-nf-ts-are-coming-to-instagram-204435805.html 

[3] https://www.theverge.com/2022/1/20/22893502/nft-twitter-profile-picture-crypto-wallet-opensea-coinbase-right-click 

What’s next for The Lindahl Letter?

* Week 68: Publishing a model or selling the API?

* Week 69: A machine learning cookbook?

* Week 70: ML and Web3 (decentralized internet)

* Week 71: What are the best ML newsletters?

* Week 72: Open source machine learning security

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead. 

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Let’s start with a brief aside. You may have noticed the podcast audio from last week’s episode was a little different. I accidentally pressed the pattern button on the back of my Blue Yeti X microphone and moved from stereo mode to omni mode. The change did not take away from the listenability of the overall podcast episode, but it does change the temperature or color of the overall audio recording feel. Personally, I strongly prefer the results of the stereo mode that have a slight natural reverb and increased depth. You may recall that I tried the cardioid mode as well earlier as it is recommended by Blue microphones for podcasts, but accidental button presses aside I’m going to stick with the stereo mode for recording going forward. I’ll take a moment now to kindly remember the iconic words of science fiction writer Douglas Adams, “Buttons aren't toys!” That is good advice to remember in the future when moving the Yeti X around my desk. 

We can now return to the question at hand related to the title of this essay, “Does a digital divide in machine learning exist?” Yes. A digital divide exists. A world of online content exists, but it has a certain barrier to entry or access. You need a smartphone, computer, or tablet with access to the internet to participate with and access the digital world. It is distinct and separate. To that end a divide exists. It really is a digital divide between technology usage and a normative set of functions distinctly separate. Within the machine learning landscape I would argue that a digital divide exists and we could probably categorize it in multiple ways. First, you have the layer of digital divide that exists between those who have access to technology and elect to use it. Beyond that first layer you have to consider the complexity of machine learning models and the underlying data. As a second layer, you probably have to consider that the digital divide creates an inherent bias where under representation or even complete exclusion exists inside the data being used to train and implement machine learning models. That is probably a structural data inequality that is not easily corrected to remove bias related to a lack of inclusion. 

I did read an article titled, “Exploring the Intersection of the Digital Divide and Artificial Intelligence: A Hermeneutic Literature Review,” from 2020 [1]. The paper is free to download and does look at a lot of literature. You can generally jump to “Appendix A: Digital Divide Research” to see some extra content about the digital divide if that is an area you are interested in better understanding. They had a focus on visible and invisible AI that I found interesting. A focus on AI that is visible to a user and elements that would not be visible. A lot of the machine learning models at work today are not visible to the users they are impacting. 

The metaverse is not visible to most people. For most of us the metaverse is an example where a very real digital divide will exist within two distinct worlds of interaction. I don’t participate in the metaverse. A more pressing example than the metaverse would be access to care between those that can utilize digital access as a vector and people who are unable to use technology for scheduling. Within that frame of reference, I read an article from Anita Ramsetty and Cristin Adams about the “Impact of the digital divide in the age of COVID-19” from the Journal of the American Medical Informatics Association [2]. It was only 2 pages long, but it was directly looking at the topic I wanted to read about. The authors very carefully argued about how underserved communities could exist from a healthcare perspective based on a digital divide. 

Machine learning itself also creates problems based on the efforts required to make it work [3]. A lot of gig workers help train and work with the data necessary to make machine learning work. The article listed above from the MIT Technology Review really was focused on the content the title indicated, “AI needs to face up to its invisible-worker problem.” Within the article a NeurIPS talk was referenced by Saiph Savage [4]. That talk is over an hour long and will make you really think about how the largest datasets got labeled and who did that work. It really is something to consider and understand about how foundations are built within the largest language models. 

Links and thoughts:

1. I listened to a New York Times podcast from Kara Swisher this week called Sway that covered how Elon Musk might shape the future of Twitter. Casey Newton showed up and they dug into the potential changes at Twitter now that Elon Musk is the largest single shareholder.  

2. This week you are getting a second podcast link. This one was titled “Is streaming just becoming cable again? Julia Alexander thinks so” from the Decoder podcast with Nilay Patel. It was an interesting conversation with two people who have obviously spent a lot of time talking.

Top 5 Tweets of the week:

Footnotes:

[1] Carter, L., Liu, D., & Cantrell, C. (2020). Exploring the Intersection of the Digital Divide and Artificial Intelligence: A Hermeneutic Literature Review. AIS Transactions on Human-Computer Interaction, 12(4), 253-275. https://doi.org/10.17705/1thci.00138

[2] Anita Ramsetty, Cristin Adams, Impact of the digital divide in the age of COVID-19, Journal of the American Medical Informatics Association, Volume 27, Issue 7, July 2020, Pages 1147–1148, https://doi.org/10.1093/jamia/ocaa078 

[3] https://www.technologyreview.com/2020/12/11/1014081/ai-machine-learning-crowd-gig-worker-problem-amazon-mechanical-turk/ 

[4] https://nips.cc/virtual/2020/protected/invited_16164.html 

What’s next for The Lindahl Letter?

* Week 67: My thoughts on NFTs

* Week 68: Publishing a model or selling the API?

* Week 69: A machine learning cookbook?

* Week 70: ML and Web3 (decentralized internet)

* Week 71: What are the best ML newsletters?

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead. 

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Ethics should be a part of every machine learning course. It has to be a part of every machine learning journey. Perhaps the best way to sum it up as an imperative would be to say, “Just because you can do a thing does not mean you should.” Machine learning opens the door to some incredibly advanced possibilities for drug discovery, medical image screening, or just spam detection to protect your inbox. The choices people make with machine learning use cases is where the technology and ethics have to be aligned. Full stop. That is the point I’m trying to make today and this essay could stop right here. I’m going to carry on anyway to celebrate the point as I consider it to be vitally important. 

No one really solid essay or set of essays on AI/ML ethics jumped out and caught my attention this week during my search. Part of my search involved digging into results from Google Scholar that yielded a ton of different options to read about “ethics in machine learning” [1]. A lot of those articles were about how to introduce ethics to machine learning courses and about the need to consider ethics when building machine learning implementations. Given that those two calls to action are the first things that come up and they are certainly adjacent to the primary machine learning content being shared it might make you take a moment to pause and consider how much the field of machine learning should deeply consider the idea that just because it can do something does not mean you should. Some use cases are pretty basic and the ethics of what is happening is fairly settled. Other use cases walk right up to the edge of what is reasonable in terms of fairness and equity.

An open access article from Nature did catch my attention by Samuele Lo Piano called, “Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward” [2]. That 7 page article has almost 2 pages of references which was pretty intense as citation to content ratios go in published articles. Within my search I was looking for a foundational article or essay that is commonly referenced. I never really did find one. I ended up moving on to an industry driven essay from the team over at Toward Data Science about, “Ethics in machine learning” [3]. That essay did scale back to the basics of the question at hand in terms of how ethical considerations are applied to building machine learning models.

I wanted to refocus my efforts on the macro considerations related to ethics in machine learning at this point. I remembered that Rob May shared a weekend commentary as a part of the Inside AI newsletter recently about the dark side of reducing friction in taking action with advanced technology [4]. Rob even went as far as sharing an article from one of my favorite technology related sources “The Verge” about just how easy and low friction it was to use machine learning to suggest new chemical weapon builds [5]. That is a very real example of where reducing friction to doing a thing opens the door to very problematic actions that illustrate the need for a foundational set of ethics.

If my call to action and introduction of an imperative to the machine learning ethics space were not enough to compel you to ground your efforts, then please consider a hand curated selection of three videos to assist you in your journey. Maybe one of them will catch your attention and help spread the word about ethics in machine learning.

1. In under 2 minutes Dan Frey who is a professor of mechanical engineering at the Massachusetts Institute of Technology (MIT) introduces this video, “Exploring fairness in machine learning.” This pitch goes back to the Comprehensive Initiative on Technology Evaluation (CITE) from the MIT D-Lab which was launched in 2012 and provides a framework you can use for evaluation [6].

2. If you were looking for something less tactical and more discussion oriented, then please consider this much longer 90 minute video from the New York University (NYU) Stern School of Business, Fubon Center for AI, Business Speaker Series, titled, “Machine Learning, Ethics, and fairness.” The video is from back on Monday, April 15, 2019 by Dr. Solon Barocas of Cornell University and Professor Foster Provost who is director of the Fubon data analytics and AI intuitive and it really digs into the question of ethics in machine learning.

3. Finally the third curated selection for you is a shift to a 6 minute video from an industry leader.  The IBM Technology and the IBM Cloud group shared a video with Phaedra Biondiris whose title in the video is noted as, “Trustworthy AI Leader: IBM Global Business Services.” This video is more grounded and is probably a good place to wrap up this essay.

Links and thoughts:

https://www.youtube.com/playlist?list=PLZHnYvH1qtOYXzWxVdIU1ZDpbLvxbZdyQ 

Top 5 Tweets of the week:

Footnotes:

[1] https://scholar.google.com/scholar?q=ethics+in+machine+learning&hl=en&as_sdt=0&as_vis=1&oi=scholart 

[2] Lo Piano, S. Ethical principles in machine learning and artificial intelligence: cases from the field and possible ways forward. Humanit Soc Sci Commun 7, 9 (2020). https://doi.org/10.1057/s41599-020-0501-9 

[3] https://towardsdatascience.com/ethics-in-machine-learning-9fa5b1aadc12 

[4] https://inside.com/campaigns/inside-ai-31781/sections/inside-ai-commentary-by-robmay-275419 

[5] https://www.theverge.com/2022/3/17/22983197/ai-new-possible-chemical-weapons-generative-models-vx 

[6] http://d-lab.mit.edu/research/mit-d-lab-cite 

What’s next for The Lindahl Letter?

* Week 66: Does a digital divide in machine learning exist?

* Week 67: My thoughts on NFTs

* Week 68: Publishing a model or selling the API?

* Week 69: A machine learning cookbook?

* Week 70: ML and Web3 (decentralized internet)

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead.  

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Maybe revisiting large language models should have been saved for a few weeks from now, but we are going to begin that journey into the foundations of machine learning anyway. My opening question within this chautauqua should be about how large language models in the machine learning space will change society. To that end it might be good to read a post from the Stanford University HAI or Human Centered Artificial Intelligence Institute, “How Large Language Models Will Transform Science, Society, and AI,” by Alex Tamkin and Deep Ganguli [1]. That institute has a mission of, “Advancing AI research, education, policy, and practice to improve the human condition.” While that sounds like an interesting mission statement to attempt to fulfill, it probably ignores the darker possibilities of what could happen. I went out and read the 8 page paper from the post Alex and Deep that they shared, “Understanding the Capabilities, Limitations, and Societal Impact of Large Language Models” [2]. Understanding how large language models could impact the economy and potential misuses are considered in that research which made it a very solid place to start my analysis for this week.

Some really large blocks of content for machine learning exist online. The amount of written work being shared related to machine learning is exponentially growing. It is seriously out of control and beyond what anybody can really reasonably track anymore. One of those blocks of content that caught my attention this week was the ML Compendium by Dr. Ori Cohen [3]. First, this pretty deep work made me wonder about how GitBook works and what other content might be on that platform. Second, it made me wonder about how interactive delivery formats might change the future of textbooks in college settings. A quick search for language models in that collection of links and other content took me to a section on “attention” that included BERT, GPT-2, and GPT-3 [4]. It was not really what I was looking to read about this week and my attention quickly turned elsewhere. 

What I was expecting to dig into was the paper on foundational models from a bunch of Stanford University related contributors noted as, “On the Opportunities and Risks of Foundation Models: A new publication from the Center for Research on Foundation Models (CRFM) at Stanford University, with contributions by Shelby Grossman and others from the Stanford Internet Observatory” [5]. You can get the full 212 paper over on ArXiv [6]. By this time in our journey together you have downloaded that paper a couple of times. Yannic theorized that the paper will end up being a key referenced work due to the number of contributors and the volume of things covered. I can see it becoming a part of curriculums for years to come as it has so much reference material in one place and it is free to download.

I’m going to backtrack for a minute here and let you know that after a bit of review it appears that GitBook was designed to provide living documentation [7]. Teams use it to maintain and share technical documentation for software and APIs. It appears that it is also used for some projects like the one shared above. I really do think that type of content curation is probably the future of academic publishing for coursework. Really large static textbooks will be replaced by interactive content that could survive in the metaverse. Students' expectations for the delivery of content to them will fundamentally change in the next 10 years and courses that demand a rigid reading of chapter by chapter in a textbook will fall out of favor. 

Links and thoughts:

Top 6 Tweets of the week:

Footnotes:

[1] https://hai.stanford.edu/news/how-large-language-models-will-transform-science-society-and-ai

[2] https://arxiv.org/abs/2102.02503 

[3] https://mlcompendium.gitbook.io/machine-and-deep-learning-compendium/ 

[4] https://mlcompendium.gitbook.io/machine-and-deep-learning-compendium/deep-learning/deep-neural-nets#attention 

[5] https://fsi.stanford.edu/publication/opportunities-and-risks-foundation-models 

[6] https://arxiv.org/pdf/2108.07258.pdf 

[7] https://docs.gitbook.com/

What’s next for The Lindahl Letter?

* Week 65: Ethics in machine learning

* Week 66: Does a digital divide in machine learning exist?

* Week 67: Who still does ML tooling by hand?

* Week 68: Publishing a model or selling the API?

* Week 69: A machine learning cookbook?

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead. 

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Thank you for tuning in to this audio only podcast presentation. This is week 63 of The Lindahl Letter publication. A new edition arrives every Friday. This week the machine learning or artificial intelligence related topic under consideration is, “Sentiment and consensus analysis.”

My coding experience probably most closely aligns to this topic. Crawlers, bots, and other automation use sentiment analysis. A lot of my original automated coding efforts were related to trying to understand sentiment analysis. A lot of people built web crawling software that ran on some pretty tight schedules to collect a set of target pages. That content was then scraped to find companies trading on public stock exchanges. Those company names and more importantly stock symbols would be evaluated for settiment. Initially that analysis was brute force or you could say explicitly hardcoded values to see how many words near the stock symbol or name were positive or negative. You can get lists of words like a dictionary with sentiment scoring that make that relatively easy to accomplish. You probably can imagine that information was used to try to figure out if a stock was going to go up or down. I personally only mapped out the system for paper trading, but it was a very interesting technology to build. 

You can go out to Google Scholar and find a ton of articles to read with a search for, “machine learning sentiment analysis,” [1]. You will see a ton of natural language processing and machine learning topics that intersect with the phrase sentiment analysis. Understanding the sentiment of a block of prose is something that is highly desirable for a variety of reasons. The use cases in some cases are very valuable which makes this particular topic something that a lot of different researchers have worked to understand. That Google Scholar search directed me to an academic publication with 501 citations to initially take a look at for this review. It was another Springer publication where they wanted $39.95 to download that single PDF. Those prices for access to academic works is why pre-prints and other open access journals are so popular. Erik Boiy and Marie-Francine Moens produced a work called, “A machine learning approach to sentiment analysis in multilingual Web texts,” [2]. You could go out to Research Gate and get the publication without the paywall as it was shared by one of the authors [3]. The article is surprisingly easy to read and happens to be very direct in the delivery of content. Back in 2008 and 2009 they spent time talking about manually annotating and working with different techniques. Current machine learning efforts have come a long way thanks to sample training datasets where a lot of this type of mapping or annotation has been done already. To that end you can get a lot of off the shelf machine learning models for this type of sentiment analysis that outperform the 83% accuracy they found for English texts. 

After digging into that foundational article in my review I’ll admit my focus shifted to looking at how people use sentiment analysis to mine data from Tweets on Twitter. One of the articles that caught my attention was, “Machine Learning-Based Sentiment Analysis for Twitter Accounts,” from 2018 [4]. That article also very quickly referred to sentiment analysis as opinion mining which seems to be a popular convention at the start of research papers. On the Twitter development webpage you can even find instructions related  to, “How to analyze the sentiment of your own Tweets,” that comes complete with code snippets [5]. Other coding examples exist as well that show step by step methods of doing this type of sentiment analysis [6]. The paper referenced above and the step by step guide I shared both use the Tweepy code package to work with the Twitter API. Using that as the starting point for gathering Tweets the next step in the process involves getting ready to do some type of sentiment analysis. That happens to be the part of the equation I generally find interesting. The researchers from the above article basically made a classification model that yielded percent positive and percent negative. The actual methods for determining the positive or negative sentiment are very interesting. That type of explicit encoding hardly requires a machine learning model. You would not have to train the model on anything as every word is already encoded with a value. 

You could spend an entire day reading the guides hosted on the Berkeley Library website [7]. You can go find a bunch of different sentiment analysis dictionaries with scoring and other elements to do this type of analysis. It used to be a lot harder to achieve before the scoring was just something you can call or reference. I wonder about how much the scoring shifts over time and based on the collective national mode. Languages are always shifting based on the way we use and apply meaning to words. Something with a relative low score could shift rapidly with a single meme, viral post, or major world news event. Our language and how we share things with people is constantly changing. You can go read a recent article published in March of 2022 from the Nature publication covering how, “Restoring and attributing ancient texts using deep neural networks,” [8]. That article covers the Ithaca model and how it attributes location and otherwise uses a deep neural network for restoration of ancient Greek inscriptions. You can see that our modeling of language has increased a lot between 2008 and 2022. Our ability to glean insights out of sentiment modeling and provide context is potentially increasing daily.  

Links and thoughts:

“[ML News] DeepMind controls fusion | Yann LeCun's JEPA architecture | US: AI can't copyright its art”

“Never Hate On Your Community - WAN Show March 4, 2022”

Top 5 Tweets of the week:

Footnotes:

[1] https://scholar.google.com/scholar?q=machine+learning+sentiment+analysis&hl=en&as_sdt=0&as_vis=1&oi=scholart 

[2] Boiy, E., Moens, MF. A machine learning approach to sentiment analysis in multilingual Web texts. Inf Retrieval 12, 526–558 (2009). https://doi.org/10.1007/s10791-008-9070-z 

[3] https://www.researchgate.net/publication/220479915_A_Machine_Learning_Approach_to_Sentiment_Analysis_in_Multilingual_Web_Texts 

[4] Hasan A, Moin S, Karim A, Shamshirband S. Machine Learning-Based Sentiment Analysis for Twitter Accounts. Mathematical and Computational Applications. 2018; 23(1):11. https://doi.org/10.3390/mca23010011 

[5] https://developer.twitter.com/en/docs/tutorials/how-to-analyze-the-sentiment-of-your-own-tweets 

[6] https://towardsdatascience.com/step-by-step-twitter-sentiment-analysis-in-python-d6f650ade58d 

[7] https://guides.lib.berkeley.edu/c.php?g=491766&p=7826401 

[8] Assael, Y., Sommerschield, T., Shillingford, B. et al. Restoring and attributing ancient texts using deep neural networks. Nature 603, 280–283 (2022). https://doi.org/10.1038/s41586-022-04448-z 

What’s next for The Lindahl Letter?

* Week 64: Language models revisited

* Week 65: Ethics in machine learning

* Week 66: Does a digital divide in machine learning exist?

* Week 67: Who still does ML tooling by hand?

* Week 68: Publishing a model or selling the API?

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead. 

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

For years it has been a running joke with some of my close friends that I have been perpetually writing a book about the intersection of technology and modernity. To that end my consideration of how we build and what it means for our shared social fabric in terms of civility and technology has been a major part of my creative efforts over the years. Modernity as a word is an attempt to signal that the now or more to the point something very current is being considered. Singularity is a word that has been used in a variety of ways. When you are describing a phenomenon in space the word singularity could be used to describe the formation of a black hole event. Some type of force would compel matter to keep compressing into a single spot until the density of that formation becomes a singularity. In this case the word singularity is being applied to a technological based scenario where in a similar fashion the advance of technology continues to the point where the advance itself cannot be stopped anymore. Within that metaphor technology would have advanced to a point where no matter what action was taken the advance of technology would continue in an uncontrolled way. 

A lot of science fiction authors have dealt with the rise and fall of the singularity. You could easily hide under a pile of them and read for years. It is a very interesting subject to take on as it allows for a myriad of very advanced problems to deal with and typically creates all sorts of drama. My title for this particular essay deals with the possibility of touching the singularity. To that end I’m describing a point of time before the intersection of technology and modernity. You can argue with the advance of large language models or foundation models, potentially the advance of technology is starting to become a snowball that will roll down the mountain either slowly or very quickly toward the singularity. I went back to look at the paper published by a ton of researchers at Stanford University, “On the Opportunities and Risks Foundation Models,” only to find that the words “singularity” and “modernity” do not occur [1]. All four of the references to “awareness” were about public understanding and not related to the state of an AI’s understanding of the world. The idea of an artificial consciousness is typically adjacent to arguments related to the formation of a technological singularity. 

This problem is a complex one to begin to dig into and it took me about 500 words split between two paragraphs to start to set up the groundwork necessary to begin to question the intersection of technology and modernity. A simpler way to get to this point probably exists, but it will take some more practice on my part to introduce in the future. At this point, I’m very curious and you may be as well about just how close we are to touching the singularity. I looked at a very well referenced paper, “Future Progress in Artificial Intelligence: A Survey of Expert Opinion,” published in 2016 by Vincent C. Müller and Nick Bostrom [2]. That analysis of expert opinions found that there is a 50/50 chance between 2040 and 2050 that a general artificial intelligence or AGI would spring into existence or be created. Arguments can be made and are being made about if the singularity is inherently good or bad for civil society and civility in general. That is not a consideration I’m working with at the moment. My consideration of this is as an event or more to the point right before the event occurs. I did go back and read an article from a 2015 issue of the New Yorker magazine online called, “The Doomsday Invention: Will artificial intelligence bring us utopia or destruction?” [3]. That article is principally about Nick Bostrom and does consider utopia and destruction if you want to go give that a read. 

We probably do not have the robotic automation available for an AGI to manifest as both being capable of awareness and constructing things. We are building highly complex 3D printing and other automation methods, but the supply chain for that is pretty complex at the moment. We would probably be able to see the rise of that type of technology before a more gradual intersection of technology and modernity occurred. It would have to get pretty far along to be a true singularity in the sense that it would be something that could continue on as an unstoppable force. Right at that point of moving toward the singularity is where my thoughts refocus on the social impacts to civility, our social fabric, and the normative patterns that make up our shared perspective. As I continue to dig into that point of time and really try to think about the branching of outcomes I do think it will eventually turn into a lengthy manuscript or a full book. Over the last twenty years or so I have not been ready to complete that writing effort. I have written extensively about normative patterns, civil society, and how technology relates to those concepts. I have always just fallen short of really trying to map out the intersection of technology and modernity. 

At this point in the journey week 62 is now wrapped up and I’m going to start to produce the audio version of the above text. This week’s narration will include some coverage of the links and thoughts being shared. I have to remember to create an outro style ending for these recordings. The first ones abruptly ended without any sign posting to the listener that the end had occurred. A quick look at the actual post would have been enough for somebody to realize that those podcasts just stopped at the end of the main section prose. That is probably not the best method of gracefully handling the conclusion of a podcast so I have adapted. 

Links and thoughts:

1. This week on The Vergecast Nilay and friends talked about technology. The episode titled, “Steam Deck review / Samsung Galaxy S22 Ultra review,” was pretty decent.  It included some discussion of technology embargos, mega gadgets, and the future of technical things. 

 

2. On the Lex Fridman podcast episode number 267 the guest was Mark Zuckerberg. Generally speaking it is interesting to hear a long form interview with Mark Zuckerberg. I enjoyed a couple of them that were hosted by Casey Newton on Platformer. 

 

or you can watch it here:

You can find that podcast interview with Casey Newton and Mark Zuckerberg here:

https://megaphone.link/VMP6088507415 

3. During the course of walking around Target this week I pushed play on this podcast and had no idea the topic would pull me into totally focusing on what the writers had to say. Go check out, “The Sunday Read: 'The Battle for the World’s Most Powerful Cyberweapon,'” podcast episode of The Daily and you won’t be disappointed if you enjoy thrillers, spycraft, and cyber security. 

4. Nilay talked to the CEO of Sonos about some recent lawsuits. This is an example of the excellent reporting that Nilay had done with the Decoder and Vergecast publications. This is an example of how long format conversation is superior to sound bytes. To that end, most reporters could learn from curiosity based questioning and build up of conversations that occur in this type of forum.

5. Over the last decade I have listened to more hours of Josh Topolosky, Nilay Patel, Paul Miller, and Dieter Bohn on podcasts than I care to admit. From Endgadet, This is my next, The Vergecast, and Tomorrow I have tuned in on a weekly basis to hear about technology. This week one of the best technology reviewers in the game Dieter Bohn has elected to leave journalism and take a job related to product strategy at Google. People move from politics into journalism and from journalism into politics. I know people move from technology journalism into the business community, but this is a pretty high profile move. Legitimate and independent product testing and the communication about industry trends is a segment of journalism that has both flourished online and diminished within the traditional journalism landscape. That is probably due to a number of facts we cannot blame on Dieter. This little aside is my heartfelt goodbye to an interesting voice in a landscape of technology that will no longer be a part of my weekly listening routine.

Top 5 Tweets of the week:

Footnotes:

[1] https://arxiv.org/pdf/2108.07258.pdf 

[2] https://philpapers.org/rec/MLLFPI 

[3] https://www.newyorker.com/magazine/2015/11/23/doomsday-invention-artificial-intelligence-nick-bostrom 

What’s next for The Lindahl Letter?

* Week 63: Sentiment and consensus analysis

* Week 64: Language models revisited

* Week 65: Ethics in machine learning

* Week 66: Does a digital divide in machine learning exist?

* Week 67: Who still does ML tooling by hand?

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Without question a good number of you reading these words right now have interacted with at least one network platform today. A futurist who was trying to predict the rise or fall of network platforms would probably side with a foreseen expansion and increased clustering of network platforms going forward. Facebook (Meta), Twitter, Slack, and other places where a technology driven platform brings a community together by hosting and maintaining a network are going to end up being categorized as network platforms. Even this very Substack post that you are reading right happens to have been distributed by a network platform of a much smaller scale than the previously mentioned ones. 

Adding an element of artificial intelligence as the base of a network platform is an interesting proposition. First you have to assume that an artificial intelligence could be created that is capable of being the foundation of a network platform. Second, artificial intelligence would have to be interesting enough to maintain or sustain the network platform without destroying the hosted community. That very well could be a case of destruction through stability or interaction. You have to remember that an artificial intelligence would be able to create an asymmetric amount of content compared to the user community. Orders of magnitude would be required as a measure to describe the potential flooding that could very well occur without some degree of regulated order and consistency. For the most part all of that could be done without the general artificial intelligence being aware. It could be an AI network platform capable of a multitude of tasking, but not directly aware of what the culmination of that tasking really involves. 

Last week while we were digging into what you get with a general artificial intelligence more time should have been spent on the concept of awareness and what that means. A lot of science fiction works have referred to computers, homes, or ships that have built in general artificial intelligence (GAI). That amazingly advanced GAI will handle requests and complete tasks both requested and simply needed. All of that could end up getting transferred to the needs of future AI network platforms, but we are pretty far away from getting to that point based on what we have right now. Voice assistants on phones are just barely branching into completing more than a few tasks. They are making progress for sure and we are getting closer to being able to ask the computer to complete a task with the advent of models like GPT-3 and how it is being extended to create things. 

I’m working through the process of incorporating my links and top 5 tweets of the week into the main podcast recording each week. Getting to that point will require providing coverage of the content in a written format within the newsletter that can be read which would facilitate inclusion in the podcast. Early issues of The Lindahl Letter did include more commentary with the links and Tweets, but over time I shifted over to a method of just including 7-10 pieces of content that caught my attention that week and ended up becoming curated content. Providing clear voice overs for why a YouTube video was included will require more effort than the simple act of embedding the link as a curator of content. For this penultimate episode which essentially is an homage to Marvin the android who Douglas Adams brought to life in the Hitchhiker's Guide to the Galaxy, I’m going to give integrating the content a try. 

Links and thoughts:

1. Some of you are aware that I keep recordings of classic political speeches on my phone and listen to and break them apart for fun. That interest in politics has existed for decades. From being captain of a high school debate team to just loving deep political trivia this podcast episode from actor Rob Lowe with White House Press Secretary Jen Psaki was an excellent conversation. Rob did an amazing job of bringing trivia to life during the episode, “Jen Psaki: The Go-To Person.”

2. This week Linus and Luke are super excited about the launch of the Steam Deck handheld gaming experience. They shared an episode of the WAN show called, “Steam Deck Review: PC Gaming in Your Palm, at Long Last.” Linus indicated this week that drama was off the table, but that never really happens on the WAN show. 

3. You might be wondering why this episode was included based on just seeing the thumbnail alone. AJ and the team over at DFB shared an episode titled, “The Truth About Disney's Star Wars Hotel -- Star Wars: Galactic Starcruiser.” It really is a video with 25 minutes of commentary and thoughts on a really expensive Disney experience. I found it fascinating and included it this week. 

In order to facilitate the process of reading the commentary some numbering has been added above to help signpost the change in what link is being discussed. Including this content could end up pushing the recording length from the 5 minute range to somewhere near 10 minutes. Overall the recording and editing experience will be the same and should produce the same type of output each week. It will just produce more of it providing a more complete experience to people who are just listening to the content from just the regular RSS link, Apple Podcasts, Stitcher, or Pocket Casts. 

Footnotes:

[1] https://scholar.google.com/scholar?q=artificial+intelligence+network+platforms&hl=en&as_sdt=0&as_vis=1&oi=scholart 

What’s next for The Lindahl Letter?

* Week 62: Touching the singularity

* Week 63: Sentiment and consensus analysis

* Week 64: Language models revisited

* Week 65: Ethics in machine learning

* Week 66: Does a digital divide in machine learning exist?

I’ll try to keep the what’s next list for The Lindahl Letter forward looking with at least five weeks of posts in planning or review. If you enjoyed this content, then please take a moment and share it with a friend. Thank you and enjoy the week ahead.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com

Thank you for tuning in to this audio only podcast presentation. This is week 60 of The Lindahl Letter publication. A new edition arrives every Friday. This week the machine learning or artificial intelligence related topic under consideration is, “General artificial intelligence.”

We made the ten hour drive from Denver, Colorado to Kansas City, Kansas this week. Last time we made the lengthy drive on I-70 we listened to the audio book recording of, “The Age of AI: And Our Human Future,” by Henry Kissinger, Eric Schmidt, and Daniel Huttenlocher that was published toward the end of last year on November 2, 2021. It was a pretty decent cut at the current state of artificial intelligence and its limitations. I had looked around for another artificial intelligence related audiobook to listen to during this trip. Instead of going with an audiobook this time around, I ended up listening to some of my favorite podcasts during the 10 hours of driving time.

According to a Pew Research Center survey of adults in the United States nearly a quarter of people get their news from podcasts [1]. The other key finding from that survey was that younger adults 18-29 were even more likely to consume news by podcast at 33%. Some of the curated news podcasts are just a higher quality of content than what you would see on cable news. Consolidated and well packaged content comes from solid editing and well thought out efforts. Getting news on demand when you want it is much easier than waiting for cable news to gather enough momentum to get to the point. Based on the survey data a shift is occurring. Podcast subscriber numbers are also indicating that based on audience size the shift may have already occurred.

For those of you who are curious what podcasts made up my 10 hours of traveling time I made a list. Included in that list were The Daily from the New York Times, Start Here from ABC News, The Vergecast, The NPR Politics Podcast, This Week in Google by Leo Laporte, Bourbon Pursuit, and Sway by Kara Fisher. The episode of Sway we listened to was about, “Tech’s Love Affair With Miami,” which was a very interesting look at how the clustering of technology focused people is changing. Kara and Keith talked about the migration from Silicon Valley to Miami. I previously shared a look at where the main AI labs are located back during my week 34 post about, “Where are the main AI Labs?” I had referenced a report that came out from The Brookings Institute by Mark Muro and Sifan Liu titled, “The geography of AI: Which cities will drive the artificial intelligence revolution?” [2]. The locations according to that report, “...include eight large tech hubs—New York; Boston; Seattle; Los Angeles; Washington, D.C.; San Diego; Austin, Texas; and Raleigh, N.C.—and five smaller metro areas that have substantial AI activities relative to their size: Boulder, Colo.; Lincoln, Neb.; Santa Cruz, Calif.; Santa Maria-Santa Barbara, Calif.; and Santa Fe, N.M.” You may have noticed that none of that seemed to include Miami, Florida. 

Previously I shared a book called, “Artificial Intelligence: A Modern Approach,” by Stuart Russell and Peter Norvig [3]. This is probably my favorite topic within the artificial intelligence space. It happens to be the topic that powers a ton of science fiction movie and book plots. General artificial intelligence is much more exciting than any special or specific applications of artificial intelligence. The ability to solve a variety of problems and work beyond a single use case is exciting. 

I started reading an article from Ragnar Fjelland about, “Why general artificial intelligence will not be realized” [4]. Working through the article was like a refresher on philosophy and the nature of intelligence. It certainly is one perspective on general artificial intelligence. You could go with a more positive perspective from McKinsey and Company called, “An executive primer on artificial general intelligence” [5]. That article starts out by acknowledging that technology is changing, but getting to a true general intelligence is probably pretty far off. However, unlike Ragnar they don’t argue that it is an impossible task. Next I turned to Forbes from 2021 to read a little more about, “The Future Of Artificial General Intelligence” [6]. Some of the coverage is hopeful and some of it theorizes the endeavor is hopeless. 

Now at the end of this post is the time to share a short update on my audio recording methods. This week's audio was recorded using the Yeti X professional microphone’s cardioid mode instead of the stereo mode which apparently kicks on automatically. The cardioid mode is geared toward recording sounds directly in front of the microphone which should be better for a podcast. You are welcome to let me know in the comments if you prefer the reverb of the stereo mode of previous episodes or the more targeted audio of the cardioid mode that delivered this recording.

Footnotes:

[0] https://www.pewresearch.org/fact-tank/2022/02/15/nearly-a-quarter-of-americans-get-news-from-podcasts/ 

[1] https://www.brookings.edu/research/the-geography-of-ai/ make sure to download the report from the link on the left vs. just reading the page

[2] http://aima.cs.berkeley.edu/

[3] https://www.nature.com/articles/s41599-020-0494-4 

[4] https://www.mckinsey.com/business-functions/operations/our-insights/an-executive-primer-on-artificial-general-intelligence 

[5] https://www.forbes.com/sites/forbestechcouncil/2021/07/16/the-future-of-artificial-general-intelligence/?sh=1c59eaaa3ba9 

What’s next for The Lindahl Letter?

* Week 61: AI network platforms

* Week 62: Touching the singularity

* Week 63: Sentiment and consensus analysis

* Week 64: Language models revisited

* Week 65: Ethics in machine learning

I’ll try to keep the what’s next list forward looking with at least five weeks of posts in planning or review. If you enjoyed reading this content, then please take a moment and share it with a friend.

This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit www.nelsx.com