Frontier Models for Frontier Science with Professor Derya Unutmaz, Immunologist & ChatGPT Pro Grantee

In this episode of The Cognitive Revolution, Professor Derya Unutmaz, a biomedical scientist and human immunologist at the Jackson Laboratory, discusses his groundbreaking research in aging and cancer immunotherapy. As a ChatGPT Pro grant awardee, Derya provides insights into the integration of AI with biomedical sciences, emphasizing how advanced AI models are transforming hypothesis generation, data analysis, and scientific discovery. He also covers his early passion for computers and programming, the inspiration he derived from Ray Kurzweil's work, and how AI is democratizing science by enabling even young researchers to make significant contributions. Derya outlines his vision for a future with ASI, discussing potential societal impacts, the need for regulatory AI models, and the promise of a golden age where diseases are cured, aging is reversed, and resource scarcity is a thing of the past. Finally, he dreams about the long-term future, imagining a life of exploration and discovery across the cosmos. SPONSORS: SafeBase: SafeBase is the leading trust-centered platform for enterprise security. Streamline workflows, automate questionnaire responses, and integrate with tools like Slack and Salesforce to eliminate friction in the review process. With rich analytics and customizable settings, SafeBase scales to complex use cases while showcasing security's impact on deal acceleration. Trusted by companies like OpenAI, SafeBase ensures value in just 16 days post-launch. Learn more at https://safebase.io/podcast Oracle Cloud Infrastructure (OCI): Oracle's next-generation cloud platform delivers blazing-fast AI and ML performance with 50% less for compute and 80% less for outbound networking compared to other cloud providers. OCI powers industry leaders like Vodafone and Thomson Reuters with secure infrastructure and application development capabilities. New U. S. customers can get their cloud bill cut in half by switching to OCI before March 31, 2024 at https://oracle.com/cognitive Shopify: Shopify is revolutionizing online selling with its market-leading checkout system and robust API ecosystem. Its exclusive library of cutting-edge AI apps empowers e-commerce businesses to thrive in a competitive market. Cognitive Revolution listeners can try Shopify for just $1 per month at https://shopify.com/cognitive NetSuite: Over 41,000 businesses trust NetSuite by Oracle, the #1 cloud ERP, to future-proof their operations. With a unified platform for accounting, financial management, inventory, and HR, NetSuite provides real-time insights and forecasting to help you make quick, informed decisions. Whether you're earning millions or hundreds of millions, NetSuite empowers you to tackle challenges and seize opportunities. Download the free CFO's guide to AI and machine learning at https://netsuite.com/cognitive RECOMMENDED PODCAST: Second Opinion. Join Christina Farr, Ash Zenooz and Luba Greenwood as they bring influential entrepreneurs, experts and investors into the ring for candid conversations at the frontlines of healthcare and digital health every week. Spotify: https://open.spotify.com/show/0A8NwQE976s32zdBbZw6bv Apple: https://podcasts.apple.com/us/podcast/second-opinion-with-christina-farr-ash-zenooz-md-luba/id1759267211 YouTube: https://www.youtube.com/@SecondOpinionwithChristinaFarr

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Speaker 00

  To say that AI cannot come up with anything novel or innovative is total nonsense. In fact, I will argue that very soon AI is going to be much more innovative and creative than we are. If I see AI as a collaborator, it's almost on my level. It's not even like a student level anymore, like another professor who is very knowledgeable in the field. These are fields that I've actually generated the knowledge, made discoveries on as a scientist. If I'm finding the AI valuable, even in those topics,

• 0:37Speaker 00

  I'm finding it very, very difficult to believe that there is anyone in the world who would not get value out of this. I mean, not a single person.

• 0:46Nathan Labenz?

  Hello and welcome back to The Cognitive Revolution. Today, my guest is Professor Derya Unutmaz, biomedical scientist, human immunologist, and ChatGPT Pro grantee, who is aggressively using the latest AI models to aid his research into aging and cancer immunotherapies. Derya is a fascinating figure. He's a medical doctor who has personally advanced the frontiers of biomedical knowledge with many academic papers and patents over the course of his thirty-plus-year career, a technology enthusiast who has loved computers and programming since his youth, a visionary who thinks differently enough that he took Kurzweil's vision of a technology singularity seriously long before it went mainstream,

• 1:28Nathan Labenz?

  and an outspoken critic of those who would deny or delay the contributions that AI can already make to scientific discovery. In this conversation, I tried first and foremost to get a sense for how world-class domain experts like Derya are applying the latest AI models to their work. To my surprise, it turns out that while he is finding value at every step of the scientific process, including hypothesis generation, literature review, experimental design, and data analysis, his approach is actually quite straightforward. He does sometimes use more advanced techniques, like having two instances of a model debate the merits of a particular research direction, but mostly he recommends a natural conversational approach to today's models. What sets him apart, then, from those who are failing to realize value from AI assistance

• 2:13Nathan Labenz?

  isn't some advanced prompt engineering or scaffolding, but rather an opportunity-oriented mindset that starts with relentless curiosity, embraces trial and error, and is always genuinely looking for ways to make things work.

• 2:28Nathan Labenz?

  None of that is to say, however, that his results are basic. On the contrary, I think his accounts of AI eureka moments are some of the most compelling that I've heard. In one fascinating example, he asked deep research to analyze gene expression patterns in T cells across young and elderly subjects, a dataset that his team had struggled to fully interpret. The AI provided insights that, in Derya's words, "recapitulated everything I've done in the past 30 years in one sentence," identifying how the cells themselves were aging in ways that the team hadn't fully appreciated. Today, Derya views AI systems as intellectual partners that are capable of contributing to frontier professional work, even going so far as to say that he no longer trusts his own knowledge or ideas

• 3:10Nathan Labenz?

  without consulting AIs first. Perhaps most provocatively, Derya argues that it has become now unethical not to use AI in medical contexts, both clinically, as it's been repeatedly demonstrated that AI can help reduce errors and improve outcomes, and also in research, since every day matters to the many millions of people who are waiting for breakthroughs to address their life-threatening conditions. As you might expect, I wholeheartedly agree, and I was really glad to hear that Derya sees resistance gradually giving way to curiosity and even excitement as more and more people see tangible results. As always, if you're finding value in the show, we'd appreciate it if you'd share it with a friend. We'd love a review on Apple Podcasts or Spotify, and we welcome your feedback via our website,

• 3:55Nathan Labenz?

  cognitiverevolution.ai, or by DMing me on your favorite social network. For now, I hope you enjoyed this scouting report from the cutting edge of AI-assisted science with Professor Derya Unutmaz.

• 4:08Speaker 02

  Professor Derya Unutmaz, biomedical scientist, human immunologist at the Jackson Lab, researching aging and cancer immunotherapy. Welcome to the Cognitive Revolution.

• 4:19Speaker 00

  Thank you. Pleased to be here.

• 4:21Speaker 02

  I'm excited for this conversation. So you are a ChatGPT Pro Grant Awardee, and this is part of a small but hopefully growing series of conversations with people that are using the latest AI models in the most forward-thinking ways they can at really the frontier of human knowledge. And you've put out some incredible posts recently that have inspired me, and I'm looking forward to this conversation and sharing this inspirational work with others. So maybe just for a quick foundation setting, tell us a little bit about like your career and maybe the Jackson Lab. You know, people that tune into this feed are paying a lot of attention to AI, but they're probably not in general paying nearly as much

• 5:03Speaker 02

  attention to what's going on in aging and immunotherapy research. So a little primer foundation would be super helpful.

• 5:10Speaker 00

  Sure. Well, I guess my my career, I would say more like my passion is not very typical, I would say. I've, of course, been very interested in medicine and biology, and I've been doing research for past thirty five years, graduated from medical school. But actually, my my passion for computers and programming is even earlier than that. I got interested when I was in high school, when first computers were coming out. In fact, my very first computer was called Commodore Vic twenty. You know, people wouldn't know that. It came out early eighties, then upgraded to Commodore sixty four. Actually, sixty four means sixty four kilobyte of RAM that these machines had.

• 5:53Speaker 00

  I mean, they wouldn't even be considered toys these days, but I was so excited because, you know, I started learning programming. In fact, we're doing some assembly language, BASIC. And, you know, you just realize incredible power that, you know, you could just do whatever you think is in your mind. And that was based on very, very primitive computers and programming language since then. So my interest continued. And early nineties, I read this book from Ray Kurzweil, which really influenced me a lot, called The Age of Intelligent Machines. And basically Ray described how the computers were advancing and eventually we're going to develop this thing called artificial intelligence.

• 6:36Speaker 00

  And that was going to eventually surpass human intelligence and the robots were going to be developed. And I was so incredibly fascinated by that and started to follow AI since like early nineties and dabbled with it a little bit, you know, during nineties, it was more about symbolic oriented AI, you know, Lisp and Smalltalk is where the languages used at the time. So it was, it was a very different type of AI, but because of my interest, I really got deep into it and, you know, read the whole history, how it started in nineteen fifties and, you know, everything happened in MIT, you know, Marvin Minsky's and all that. Actually Marvin Minsky had an incredible book, The Society of the Mind, I think is

• 7:19Speaker 00

  still extremely valuable for AI even today. So, and then, you know, actually late nineties, early, early two thousands, I read another book from Ray Kurzweil called The Singularity is Near. And another book that really influenced me, basically where he was actually charting the increase in the computation since that time and then projecting to future. And basically he literally predicted what was going to happen. In fact, his point was by twenty twenty nine, we're going to have what's called artificial general intelligence. And by twenty forties, we'll get to this point of singularity and we can talk more about that later. And so, so those are the things that really have influenced my life and, and, you know,

• 8:02Speaker 00

  I've been trying to apply that to my own work. But my day job, let's say, or my other passion is trying to understand how the biological systems work. I focus on immune system because the immune system is very important for, you know, protecting our body, but also has implications for many of the diseases as well as the aging process, you know? I've asked this question, you know, twenty five, thirty years ago, why do we age? And actually I see aging as a disease that needs to be cured, which was kind of a heresy, you know, a couple of decades ago, but I think people are starting to accept that now. And in fact, the final thing I'll say is that about twenty years ago, I had this blog

• 8:44Speaker 00

  inspired by Ray Kurzweil, I called it Bio Singularity. So basically, what I was trying to imagine is that within twenty, thirty years, the AI will evolve to a point where we're going to start to truly understand the biological systems so that we can treat all the diseases, including cancer, and then eventually by, you know, twenty forty's, we'll even be able to reverse the aging process and upgrade our own biological capabilities. And actually, I've been working towards that goal, you know, mostly recently mostly focusing on cancer. Because using the immune system, we can actually program our immune cells to attack cancer cells and actually like we can write code on cells.

• 9:29Speaker 00

  We can create AND OR gates and things like that. So it's extremely cool engineering. So I've studied a lot of other things like HIV AIDS and during COVID, we did a lot of work on that and chronic diseases and so on. But if there's interest, we can talk about that, too.

• 9:45Speaker 02

  Yeah, I'm interested in all of it. We'll get to as much as we can. Wouldn't it? Interesting question. And I've, you know, I'm a little younger. My first computer was a Windows three point one, so I don't go quite as far back into computer history. And I would say I kind of caught wind of the Kurzweil line of thought a little bit later, too, but not too much later, it was around about the singularity is near timeframe. And I don't have a good account of this for myself. I wonder if you have a theory of why you were willing to take ideas like that seriously when others were not. I always kind of was, and I look back and I definitely feel like that was strange. And I'm not sure, you know, what caused me to do that.

• 10:27Speaker 02

  And I also sort of see that as kind of carrying through even to today where obviously we'll get into like the AIs are, in my view, getting to the point where they're like undeniably very powerful and, you know, going to be, even if there's, you know, the classic line is, even if there's no further progress, which of course there will be, you know, it seems like we already have enough to be quite transformative. With a lot of implementation work. But I guess my my question there is just like. What is it that you think separates the people that have historically taken this stuff seriously? And today are like seeing it more clearly from people that have sort of not seen the potential and continue to sort of deny it even as it kind of materializes written here.

• 11:11Speaker 00

  I think the the simple answer is that you just have to be crazy. You know, because you're thinking different and in fact, one of my idols is Steve Jobs. And, you know, he he said at some point that only the crazy ones change the world because you these ideas. You know, they're way ahead of their time and say they they really do sound crazy. But since I was a child, I always tried to think different. In fact, I grew up watching Star Trek and my imagination or dream was one day I'm going to be on a starship like Star Trek and you know, go seek out new civilizations. So it's it's a different mindset. Uh, you basically don't accept the status quo.

• 11:56Speaker 00

  And actually, when you look back to human civilization, you see this incredible progress, you know, initially it wasn't very fast, but it's really been exponential, actually, especially past two hundred years or so. And that's because people who are a bit crazy, who don't accept the way things are and they believe that we could make things better and better and better. Believe in science and technology that, you know, there's nothing that we cannot solve. Uh, so, yeah, I I don't have a good answer for you, but you you really have to think different, as again, Steve Jobs said. And, you know, what you're you're absolutely right, though. When Ray published these books and ideas,

• 12:36Speaker 00

  a lot of people told he was crazy, including a lot of my colleagues in the scientific. You know, it's very strange because if you think about it, scientists should be very open minded, right? So we need to think different and be creative. But actually, it's not like that. Even within the academia especially, it, you know, there's a lot of dogmas, very, very conservative thinking. And they were like, you know, that's, that's just crazy. How can you like reverse aging and, you know, have computers better than human mind? That's just not possible. And actually, most of it is because of our ignorance. Because we didn't understand how the brain works, we assumed that it's something magical,

• 13:17Speaker 00

  right? So we can never create intelligence as good as a human being. But the reality is the more you understand biology, actually, it's a pretty bad engineering. You know, again, we can talk about that, too. We assume that we are really perfectly built or whatever. That's not the case. It's a very legacy system. Yes, it's really marvelous, but you know, it's nothing magical. I mean, when you go down to it, the neural networks in our brains is not so different than the machine neural networks. So I was a true believer, and I had continued to be a big believer in, you know, the technology changing our lives. And, and actually, this is a time when this is going to happen much, much faster. So

• 14:02Speaker 00

  that's what I'm trying to say in Twitter as well, that, you know, or X, I don't want Elon Musk to get mad at me. So that, you know, trying to warn people, because, you know, people like us feel as if you were a time traveler, and you just, you know, you, you, you've seen what will happen five, 10 years ahead of you, and you, but you're living in the present time. It can get frustrating because people think, oh, no, you're crazy.

• 14:30Speaker 00

  That's fine.

• 14:30Speaker 02

  Are you seeing, as you communicate about all of your exploits with AI and applying it to the frontier science that you're doing, are you seeing people change their minds? How, what is like the current reaction from your colleagues in the sciences to the stuff that you're showing off to them today?

• 14:51Speaker 00

  I mean, I've been in the chat GPT moment since the beginning. I've started using since GPT three and then three point five and so on. And in fact, about two years ago, two and a half years ago, I started telling my colleagues and friends, look, there's something amazing happening. This is really going to change the way we do science. We really start to implement this. And at the time, they were like, oh, no, this is just the next word predictor. You know, this is, this is actually including computational people who understand computers and bioinformatics and all that. They were very dismissive at the time, but I think since last summer, things have really started to change. Because, I mean, you can't avoid this anymore.

• 15:34Speaker 00

  For example, a friend of mine who is in California, he's a professor, quite famous in this topic. You know, he saw my post on Twitter and X and they said, you know, I can't believe it can be that good, but can you try this for me? You know, like I have this project and can you ask, you know, at the time it was all one preview and all one pro that came out. And, you know, I just did the analysis of his data and I sent it back to him. And his reaction was like in an email, it said, oh, my God, I can't believe this. I can't believe this, you know. So how can you avoid that? Now he's a daily user. I just wrote a grant application with a colleague of mine and it was probably the best grant that we wrote.

• 16:19Speaker 00

  And she said that I can't believe how easy it was to write it and we had such good ideas. Of course, we worked with AI very, very closely with chat GPT. So the moment you realize this is indispensable. I mean, it's not optional anymore. And I am a bit disappointed on the medical side. Because, you know, there it's actually has become unethical not to use AI. And that's something that I keep saying. And there's more resistance that side. Although I know quite a few physician friends who are not using chat GPT, but they're not saying that they're using it. That's the other thing. People are still.

• 17:02Speaker 00

  Feel that if they say, okay, I, you know, I wrote this project using chat G P T, or I made this diagnosis thanks to A I model. They feel like they're gonna be less valued, which is true, you know, we are less valued. But at the same time, you're adding more value. Like, you don't want to misdiagnose someone, right? So that's a life and death matter. Or if I can write a better project, or if I can analyze my data much, much better with the help of A I, I'm adding much more value. It's not just a personal productivity, but what we add to humanity overall. So it's changing.

• 17:42Speaker 00

  I, I, I think this year it's gonna be very big, but at the same time, it's very disruptive, extraordinarily disruptive in the sense that people have to change this mindset. Things are not gonna be like they were. The whole academia, the whole education, and I, I post a couple of times, you know, the notion of having a P H D, all these things are changing. So it's gonna take a bit of time to adapt, but it has to happen. There's no way out.

• 18:15Speaker 03

  Hey, we'll continue our interview in a moment after a word from our sponsors.

• 18:20Speaker 02

  Yeah, I always say, for my own medical purposes, I would at present want both a human doctor and an A I. But, um, you know, I wouldn't feel comfortable with either one alone. And I love the clarity that you bring to the analysis that just like it is getting to the point where it's unethical not to use the tool. I also think about like, people have a right in the United States to an attorney if they're gonna be, uh, put on trial. I also think maybe they should have a right to an A I, because I suspect that they could get a lot of, uh, value, you know, that they're not necessarily always gonna get from their court appointed attorney. So this is coming everywhere.

• 18:56Speaker 00

  Yeah.

• 18:57Speaker 02

  What are you finding, you know, is kind of changing most? Like, what are the biggest, maybe one way to think about it is, what are the biggest bottlenecks in your work? And which of those is your use of A I alleviating? And which ones are not yet affected by, by the current A I tools?

• 19:17Speaker 00

  I mean, there's not a single area that's not affected by the, I mean, I'm saying this personally because I use it day and night. In fact, I don't trust any of my own knowledge or ideas without asking AI anymore, even things that I know very well. I mean, like I'm an expert on. Because, you know, I tested this. For example, I uploaded a review that I wrote on a topic that I'm probably one of the best out of five, 10 people in the world. Like, I have really extreme deep expertise on that topic. And I had a one pro analyze it and actually recently deep research, which I haven't yet publicized. It was incredible. I mean, basically, all one

• 20:02Speaker 00

  pro found things that I missed. It totally understood the topic, which is very specific, gave extremely useful insights, some of them very creative. So, you know, I mean, there's nothing that I would do without consulting with AI anymore. In fact, you know, some of my friends and people who know me tend to ask medical questions. Okay, you know, my mom has this cancer or that disease. What do you suggest? What's the novel thing or treatment? I tell them, look, first, I have to consult with AI, with ChatGPT. You can too, if you want to send me. And then I'll double check it, of course, you know, to make sure that it looks

• 20:47Speaker 00

  good. But I'm not going to answer without double checking with AI. That's the point that that I wish everyone would be there. In fact, I taught my mom, who's 85 years old, to have to use ChatGPT over the summer. She lives in Turkey. You know, she had some health issues. And now she tells me that she cannot live without ChatGPT. In fact, she doesn't ask me any more questions because she trusts ChatGPT more for medical questions. And I said, I approve that, you know. And so, yeah, six months ago, or maybe, you know, a year ago, there was a lot more hallucinations. You couldn't trust 100%. But they have been

• 21:30Speaker 00

  proved tremendously. That's the other thing that people don't realize. You know, you see some publications from last year, oh, well, you know, we compared ChatGPT 4 with with human doctors or this and that, even. Looks like even was better at that time, but I can tell you it's 10 times better now in the last three, four months. And it's going to get 10 times better in the next three, four months. So why should I trust, including myself, a human opinion or knowledge? When you have literally professors in your pocket? Another example is I wrote a patent application for a colleague of mine on a medical topic.

• 22:15Speaker 00

  And, you know, I've written, I've helped write patents because, you know, through lawyers. I've, I've eight, nine patents myself. So I'm very familiar with what it takes to write a patent. And I did that with deep research and it was the best patent application ever. It was on a, on a molecule that has anti-cancer effects and things like that. It did all the patent search and understood the chemical formula. I mean, made all the claims, the secondary claims. It was incredible. And, and my friend is submitting that as a patent application. It would have probably cost over $10,000 to have that written by lawyers. It took about half an hour and didn't cost anything.

• 22:59Speaker 02

  So if I just kind of sketch out the scientific process of hypothesis generation, maybe literature review, experiment design, experiment execution, data analysis, and then maybe back to hypothesis generation. You could, you know, complicate that or adjust it. Where are you finding the most impact? And I'm particularly interested in hypothesis generation and sort of eureka moments. But I mentioned your comments, you know, kind of throughout that entire loop. I think there's one of the biggest questions in that the goalpost, of course, keep moving. Right? So it's like, well, it was one thing when it could sort of answer a question, you know, but it would seem like, okay, they're not really reasoning. Or they don't really have any higher order abstractions.

• 23:44Speaker 02

  Now we see pretty clearly, well, there are higher order abstractions. And now I think it's become very difficult. But all of us, and we're still trying to deny that there's some meaningful form of reasoning going on. And now the the goalposts have kind of shifted to, okay, but they're not going to discover new knowledge, right? I mean, maybe they can memorize the whole literature, but they're new knowledge, like that's a qualitatively different thing. So, yeah, maybe kicking you kind of through the the scientific loop, but especially interested in, you know, what you have seen in hypothesis generation and like eureka moments, if any. And if it's not yet there, then like, let's be clear about that, too. But.

• 24:16Speaker 00

  Oh, there are, there are quite a few. So let me first explain the scientific process because people think that somehow magically ideas just form and we just discover things out of thin air. That's not true. So, as you said, we form hypotheses and some of them are good hypotheses, some of them are very good ideas, very innovative, and that's our contribution. But it's all based on what we already know. Like, if, you know, if I ask you, can you come up with a brilliant idea on how to use T cells to treat cancer? I doubt that you're going to come up with something innovative. I mean, you might be

• 25:00Speaker 00

  a super genius, but still, you need a lot of background information about that topic, right? And you also need to know, to be innovative, you also need to know lots of different things. You know, one of my advantages that I'm very interested in different topics, like including playing video games. I've been playing video games since I was a teenager, you know, and I'll tell you a story about that, that's why I'm mentioning. So, before the O one models, during when GPT four especially came out, I was mainly using ChatGPT, or occasionally used also Cloud, as a way to sort of survey the field, because there's no way we can

• 25:42Speaker 00

  follow all of the information that's coming out. I mean, even in my own narrow field, there are hundreds or thousands of papers published every year. So I can try to keep up with it, but it's very difficult. So I I found it extremely useful for GPT four to. Sort of summarize what's going on. But that was mostly knowledge based. You can say that, well, you know, G P T four knew a lots of things and it was it was smart in that. But I hadn't seen, you know, like particular insights or some innovative ideas. Because that requires bringing things together. So when O one preview came out, things changed.

• 26:23Speaker 00

  And that's that's when I had my first sort of a eureka moment or wow moment, I would say. And that was when I asked O one preview, I said, OK, so I'm developing a cancer immunotherapy protocol. And this has to do with these immune cells we call T cells. These are kind of like the soldiers in the body. And we actually program them, genetically engineer them to make them recognize the tumor cells and then go and kill them. But there are lots of issues there, you know, those cells get exhausted, they don't kill very well sometimes, and they have side effects. They can cause lots of troubles. And we're trying to solve those problems. And I had to actually come up with some ideas based on Battle Royale games.

• 27:06Speaker 00

  I don't know if you're familiar with those, you know, like PUBG and things like that. That's kind of like the Survivor games, you know, you're in this island or it's a particular restricted area. And then you have to find resources and you have to compete with other players. You have to kill them or eliminate them so that you can win the game. So I had thought about using the Battle Royale analogy and applying to these T cells. Because they're also kind of competing in that environment of the tumor tissue. Whether, you know, using that idea, we can actually make them better. In fact, we did some experiments and it kind of worked. We're going to publish that very soon. So I asked the O one preview, I said, you know, can you come up with some new ideas?

• 27:49Speaker 00

  You know, be inspired by the Battle Royale games, you know, like like I was inspired and came up with some new ideas. And in fact, it did. It came out with a couple of ideas that I hadn't even thought about because it knew how the games were played. And it was able to extract that information and then transfer it to completely different. Topic of T cells in a tumor microenvironment trying to hunt down the tumors and then kill them and competing with each other and then getting better and solving this exhaustion problem. I mean, that was really remarkable for me because maybe that was kind of the first early AGI moment, maybe. And that actually got much better with O1 Pro and O1 Pro actually started to give

• 28:35Speaker 00

  ideas that we used to put in projects. You know, I mentioned some grant proposals, which I hadn't thought about. And these are not just missing information because some of it could be because, well, you know, I didn't know that this was possible to do, kind of a thing. And you get those moments as well, but these were truly kind of innovative ideas based on available knowledge and based on non-available knowledge. You know, because we form hypotheses saying that, okay, we need to, we imagine that this is the way it should be, but we don't know. We don't know if that's the case, you know, or we don't know if the cells are going to behave that way.

• 29:18Speaker 00

  And we don't know how to do that in the proper way, or what could be the best approach or experimental strategy. And in all of those cases, I'm getting new ideas. And it's not just me. I did that for a couple of my friends. One of them is a leader in the neuroscience and the other one is in inflammatory diseases. And in both cases, they couldn't believe it. I mean, it was just shocking. And these guys are experts in their field, like the world leaders. And so to say that AI cannot come up with anything novel or innovative is total nonsense. In fact, I will argue that very soon, AI is going to be much more innovative and creative than we are.

• 30:03Speaker 00

  You know, Google's AI scientists, and I hope to test that soon, has just been released and, you know, they they mentioned a couple of examples. So it's completely, completely unbelievable. You know, I get impressed, those Eureka moments almost. On a daily basis, sometimes I'm scared to do like a deep research or O1 Pro interaction because like it gives you such unbelievable ideas or insights. Just one last thing I'll mention. So the other thing is that it's not just hypothesis generation, it's also data analysis because this is extremely important in science. In biology, yes, we have to do experiments, we have to generate data.

• 30:46Speaker 00

  And nowadays, we can generate incredible amount of data. You know, billions or trillions of worth of tokens, let's say, from RNAs to proteins to cell interactions and so on, so forth. So it's very, very difficult for us to analyze that data, right? We have, you know, millions bits of metabolic data and this and that. We try to do like bioinformatics, statistical analysis and so on, so forth. But none of those are satisfactory. So recently, I actually asked Deep Research, we had some gene expression data. I don't know if you're familiar with this technique called CRISPR and RNA sequencing. So anyway, I won't go into technicalities of it.

• 31:26Speaker 00

  But basically, what we found is that certain type of T cells have certain genes that are expressed in young people, but those are not expressed in the old people. And a different set of genes are expressed as you get older in the same type of a cell. And we divide those cells into two parts as well. So we couldn't really make sense out of it like, okay, so it might mean this or that. I mean, you know, these, we know the functions of the genes themselves. So I asked that to Deep Research. It was incredible. You know, it came up with these insights. Again, this is a topic that I know extremely well. I've been working on it for 30 plus years. In fact, there was one sentence I got emotional reading that because it's sort of

• 32:10Speaker 00

  be capitulated everything I've done in the past 30 years in one sentence. It's sort of an insight that I should have come up with. So it's really mind blowing.

• 32:19Speaker 02

  Can you say a little bit more about what that is? Like, what is the insight that you felt that it achieved that you wish you had?

• 32:27Speaker 00

  Yeah, so basically, you know, I've been working with these cells, these particular sub-type of T cells for, I guess, thirty-plus years now. And we call these cells naive cells. Again, you know, it's very technical, so I'll try to keep it a little bit superficial there. But basically, this is how the immune response starts. So when you first see antigens or viruses or bacteria, these are the cells that has to be educated, and then they they become sort of what we call effector cells, and they fight against infections or cancer. And then some of them turn into a memory cell, and then they're long-lived. But as you age, the proportion of these cells is reduced, and that's why during aging, your

• 33:09Speaker 00

  immune system doesn't work very well. But then, you know, this data was basically suggesting that it wasn't just the numbers, but also their quality was actually changing. And so the insight was basically, you know, the research said, well, you know, the reason why the elderly is different than the young is because these cells that you think are young are no longer young themselves. You know, I can't remember the exact sentence, I'll probably publish that on X, that they have actually changed epigenetically, or, you know, their their character has changed. So there's there's lots of packed into that

• 33:53Speaker 00

  one sentence, because it requires enormous amount of understanding that, you know, these are the cells that differentiate into memory cells, and they see the antigen, and, you know, all those things. And it also had this temporal intelligence that that things change over time, which is which is something very unique. It's not a static type of an information. So that was that was extremely impressive.

• 34:23Speaker 03

  Hey, we'll continue our interview in a moment after a word from our sponsors.

• 34:29Speaker 02

  What advice do you have for people to bring the best out of these systems? Like, one thing that I already definitely took note of, and interestingly, I hear this also from my teammate who's the creative director at my company. He's just much more of an outsider. I'm like usually very linear and I sort of prompt in a pretty like matter of fact way. And, you know, try to make sure I have the right context and give clear instructions and whatever. And then he will bring these other things where he's like, do this in the style of, you know, a famous author, a famous director, you know, whatever, and just kind of gives it a much more high level conceptual direction to go. It sounds like you're doing some of that with these like inspirations from very different

• 35:14Speaker 02

  fields. I'd love to hear a little bit more about that. And maybe, you know, there's probably different best practices depending on like maybe where you are in that cycle of science. So I would imagine, you know, you might not want to do the data analysis as if you were in a battle royale. So, you know, how do you kind of make sure you're getting the best performance for the task across this range of tasks that you're taking to the AIs?

• 35:41Speaker 00

  Yeah, I mean, you know, people have been talking about prompt engineering. I don't have anything that's sort of a standard set of prompts. Like, I don't usually use very structured prompts per se, and I change it all the time. But the way I structure the prompts, actually occasionally what I do is that I ask O1 or O1 Pro, okay, here's the ideas that I want to analyze, you know, structure them into a prompt. So it can actually add a couple of things in context. With deep research, you don't really need that a lot because deep research actually asks you, you know, okay, is this what you mean?

• 36:24Speaker 00

  Or do you want to analyze it in this way or that way? Or do you want to focus on the solid tumors versus, you know, lymphomas? It actually predicts what you may be thinking, but forgot to mention. But the way I do it is that the way I think, I think it might be different for every person. So my suggestion to people is that don't try to stress out, you know, how should I write it in the perfect way? Or how should I engineer it and things like that? Basically, whatever's in your mind, just write it down in the most open and transparent way as possible. Maybe this wasn't the case six months ago, but after O1, you know,

• 37:07Speaker 00

  with Grok or with your O1 or Pro Models or whatever, even GPT-4, actually, you know, they've gotten so better understanding the context of what you're trying to ask. And especially if they start to know you, like the GPT as a memory. So I noticed Grok also knows about me, probably through the Twitter, because if I ask a question, it will bring it to somehow to immunology or aging or AI. It's that they start to predict what you are really interested in. And hopefully with, you know, very long memory, that that's going to even get better. But, you know, I like to think different, as I said, you know, that's why I think asking these questions like

• 37:52Speaker 00

  be inspired from this field. And you might be asking a physics question, but you might mix it with something from biology. Or if you're trying to be creative in your work, you can mix it with some sports or they do a great job in that. And I think that's when you start to see innovation and creativity, because you're bringing completely different topics and come up with something new that you wouldn't think on your narrow field. The other thing that I find very useful with O1 models, especially because they have the thinking capability, I ask them to iterate on the ideas. Say, okay, so, you know, think of this.

• 38:34Speaker 00

  But then think of what you thought and then see if you can come up with something better or something different. You know, think of this project and then come up with ten ideas, but each of the ten ideas should be somehow better than the previous one. Another one that I did, I found extremely useful. I said, imagine that there are two scientists. One of them is very enthusiastic about this idea. The other one is a very skeptical one. They're both knowledgeable in the same field. So the first one, I want you to come up with an idea, scientist one. Then I want scientist two to criticize that and then maybe suggest some new ideas. And then scientist one, you know, responds to scientist two, oh yeah, you're right. You know, like I should have thought about that.

• 39:19Speaker 00

  And then iterate that for like ten times. So they're literally brainstorming, you know, within, within these two scientists. Actually, it's same, same AI. And that could be extremely revealing. You can also see that in the thinking process. You know, I think DeepSeek was the first to show that the sort of the background thinking, that's pretty much what the AI is doing, right? So it thinks of something as, oh, wait a minute, maybe I should have thought about that. Or no, maybe this is a better idea. And you can even push that further by asking it to sort of iterate on the same ideas. So that's good for the thinking models. So the other thing is people should know what's the difference between a think model, a deep research model, or just a regular, you know, base model.

• 40:07Speaker 02

  What is your hit rate on this? I think maybe one thing that's coming to a little better clarity for me is,

• 40:16Speaker 02

  you know, and I guess the, the whole sort of reasoning paradigm has really emphasized this, right? We're scaling inference time compute generally. And maybe one of the explanatory factors for why some people have had success over the last couple of years and others have sort of not been impressed and kind of given up is maybe as simple as just kind of willingness to try a bunch of times and do this sort of, you know, ten rounds of iteration, as you said. And I, I have to imagine that like. I don't, you can tell me, you know, what your sort of senses for the hit rates or the miss rates. But I have to imagine that when you bring these like other sources of inspiration from some distant field, like it probably doesn't always work.

• 40:58Speaker 02

  So what would you say is your sort of hit rate? And what should people expect if they want to go do this? And then they sort of see something that's, you know, initially not amazing to them. What do they need to be willing to put in in order to kind of have a good chance of getting to the sorts of things that you're getting out of the AIs today?

• 41:20Speaker 00

  I mean, I think first of all, people are not really focusing trying to make it to work. They are very skeptical to begin with. They see AI as a threat, a lot of people. And, you know, how can artificial intelligence be smarter than I am? Or this doctor or this professor, you know, this sort of attitude. So you're not really trying. You're trying to find, you know, what are the defects? How can I get some hallucination out of it? Almost subconsciously that they're maybe thinking that, you know, I mean, all humans hallucinate, literally. I mean, if that wasn't the case, we wouldn't have 12 million misdiagnosis every year in America, you know.

• 42:01Speaker 00

  So, but the other thing is that, you know, maybe what's different about me is that I'm a scientist. I'm very used to trying and failing. You know, this is what we do. We form a hypothesis and 90% of the time we fail. We try to do an experiment, it doesn't work. We try again, we change something, oh, yeah, well, okay, it worked a little bit better. No, maybe we should do it this way, that way. So that's like my job, you know, this is what I do. And this is what I love to do because there's never going to be a perfect way of doing things or asking things. So, you know, I think trial is very important. And then, but, but I think again, the first point is even more important.

• 42:40Speaker 00

  You have to see AI, the way I approach, let's say, oh, one pro or deep research or even growth recently is like a colleague, like a collaborator. In fact, I gave a talk to PhD students last year, three, four months ago. It was on collaborations. But I talked about AI collaborations, not human collaborations. You know, collaboration is extremely important. Like, most of my work has been in collaboration with other scientists because you can't really do everything on your own. You need cross-fertilization of ideas and, you know, you have different skill sets and so on and so forth. So that's how I see AI is a collaborator. It's almost on my level. It's not even like a student level anymore.

• 43:24Speaker 00

  Like another professor who is very knowledgeable in the field and we're just discussing some ideas and I want really honest opinions about it. So when you have that approach, and then you try different things that you would ask yourself to come up with new ideas or really transparently put out the problem and how we could resolve it. And then even if you don't get a very satisfactory answer, just follow up on it. I mean, with a second prompt, there's no limit. You can say, okay, well, that was a good idea, but I think you should maybe think it this way. You know, what do you think if you think it completely opposite way? You know, because AI has no judgment, right? It will be very intellectually honest with you.

• 44:07Speaker 00

  Especially Gruke, I have to say, is really intellectually honest one. And so I think it's just a matter of approach. People are not trying enough or they don't want to try. Maybe that's the reason. I don't know.

• 44:24Speaker 02

  Yeah, I used to say something very similar about just like, the question you want to be asking is what can I get this thing to do that's valuable? As opposed to can I find, you know, a weakness, a flaw, a mistake? Because you definitely can do that. And if you stop there, you're going to miss out on all the upsides. I think that is a super important, just general piece of, you know, wisdom and guidance. Could you put like an estimated number on it, though? And I'm just asking because I want people to know what to expect. Like, how often does your first prompt give you something where you're like, this is great, this is exactly what I wanted? How often do you need to have a bunch of rounds to get somewhere good? And how often does a session, you know, when you sit down and try something,

• 45:06Speaker 02

  just never really lead anywhere that you find is ultimately super valuable?

• 45:13Speaker 00

  I mean, so again, it depends on which model you're using. If I'm using deep research, it's kind of rare that I follow up. I mean, I do follow up with second prompts, but mainly because the first prompt was so detailed. There's so many new ideas and new questions came up that I wanna maybe go deeper on those things. Not because I wasn't satisfied. I mean, it's 100% satisfactory, like over satisfactory. Let me put it that way. It's usually the same with O1 Pro as well. If I'm searching for some knowledge set, the thinking models are not necessarily great because they don't, they're not really search machines, you know?

• 45:56Speaker 00

  If you give them facts and problems, they're great in solving that. So, you know, and there hasn't been any unsatisfactory response from the O1 models or deep research. And then recently I've been using deep search from Grok. That's very valuable in finding information, knowledge sets. I find that even more valuable than GPT-4 search, which actually searches the different resources and sort of synthesizing that information. You know, again, Grok is outstanding in making complex topics or information or knowledge very simplified.

• 46:38Speaker 00

  You know, if you didn't really understand something, or sometimes I'm too lazy to read a very complex paper, even if it's in my field, I just upload the PDF. I say, please, you know, describe what's going on in this paper and then find some gaps or whatever, you know, they're just amazingly good at that. And correcting your own mistakes, or you know, if I'm writing a project, I, again, you know, even if it's my own ideas, I just upload it and say, okay, so this is what I'm thinking to do, but find potential pitfalls and then suggest alternative methods. 100% of the time, it will find something of value that I didn't think about. You know, it doesn't mean that everything they say is valuable.

• 47:23Speaker 00

  You can skim through some of this and say, okay, well, okay. It thinks this is a potential pitfalls, but I don't think that you can dismiss that. But then out of, you know, two, three, four of them, one of them will be very, very important. And that's what you're looking for. And then the other things that like, you know, if you ask me this question, you know, a month later, I'm probably gonna say, you know, they're even more unbelievable because they just keep getting better. Or if we were doing this interview two months ago, I would have had maybe slightly more reservations. Not much, but that's just disappearing every day. And today we're gonna have G P T four point five and we'll see how amazing that's gonna be.

• 48:05Speaker 02

  Yeah, it is crazy how fast all this is happening. I basically am a full time A I watcher at this point and it's like still getting away from me. So it's, it is certainly understandable for, you know, people who have full time jobs and, you know, try to juggle a lot of things to be struggling to keep up.

• 48:24Speaker 00

  I mean, what really amazing, and I'm saying this very sincerely, what really amazes me is that how is it possible? Anyone in the world cannot find this amazing. That's what's amazing. I mean, it's inconceivable for me because, you know, I have some very deep knowledge and expertise in certain fields. And these are fields that I've actually generated the knowledge, made discoveries on as a scientist. If I'm finding the A I valuable, even in those topics, I am finding it very, very difficult to believe that there is anyone in the world who would not get value out of this. I mean, not a single person.

• 49:05Speaker 00

  In fact, I'm, I'm willing to bet that if they claim that I will do it for them and show, you know, in even in their fields, you can generate value by interacting with A I. So this is really unbelievable for me, but.

• 49:21Speaker 02

  How about on the data analysis side? I assume when you talk about high volumes of data, presumably a lot of the analysis is going to be then done through codes. I'm guessing you're like asking the AI to kind of giving it maybe like the structure of the data or a small sample and asking it to write code to do analysis. One thing I've observed, and you know, I'm not one tenth of a percent of the scientists that you are, but in my, you know, I've dabbled at times and in my limited scientific dabblings, one thing that has really jumped out at me is it's often a small, unexpected observation that then leads into the next experiment that, you know, kind of takes you in the next big kind of positive step discovery.

• 50:06Speaker 02

  I experienced that both in chemistry years ago where, you know, we were trying to push a reaction toward more product. We thought that adding more acid to the reaction conditions would, you know, tip us more in the direction we wanted to go. And it ended up working less and just looking at that slope. This is like my one, you know, contribution to science of all time. I said, well, the slope seems to be going down. Like, have we thought about trying less? You know, and it was just that super simple observation on just a couple of very small, you know, data points that ended up and showed up, it did work better. Uh, this week, you know, I was a very small contributor to an AI research paper where the group showed that fine tuning GPT four O and other models too

• 50:49Speaker 02

  on vulnerable code outputs created what is maybe best described as an evil model. Uh, that has like all these sort of, you know, very kind of shocking, uh, opinions and uh, takes on other topics totally unrelated to code. And that also just started with, uh, sort of an anomaly in the context of another experiment. So I guess two part question is like any best practices in general for sort of data analysis? And then any strategies or observations around getting AI to come through these vast amounts of data and find those little anomalous nuggets that seem to be the breadcrumbs to something new?

• 51:29Speaker 00

  So I mean, in biomedical sciences, things are a little bit different. You know, as I mentioned, we generate incredible amount of data nowadays and we still have to generate more data. Just to give you an example, one experiment we do, for example, you know, looking at thousand individual cells and then looking at the changes of thousands of proteins in each of the cells, you know, like you can generate millions of bits of information just from one experiment. So obviously, this is not even conceivable for us to analyze.

• 52:07Speaker 00

  I think one uses that AI in biomedical sciences especially will make discoveries. How it's going to make discoveries is because you analyze this data and then you use bioinformatics and this and that and you come up with some hypothesis, you say, okay, well, you know, these genes are changing and then this is what must be happening and you base your next experiment on that. You might miss some things, but then, you know, you try to make the best guess. But AI is very different because it's actually going to the ground truth in a way, or almost the first principles of the biology and really putting things together in a way that it

• 52:48Speaker 00

  would be impossible for any human to put, even with bioinformatic tools, because it thinks of new insights from that millions of bits of data and then it will tell you exactly what the new discovery is and what should be the next experiment. So in fact, you know, I'm actually digging most of our old data sets, you know, things that we have generated five, 10 years ago. I'm going to reanalyze everything because there's probably, you know, tons of buried knowledge in there that that is waiting to be discovered. It's like a gold mine. And, you know, I don't know how best to do that. I'm trying to push the limits, you know, we'll see how Google AI scientists, you know,

• 53:31Speaker 00

  how much data that I can upload. So far, I've uploaded about close to a thousand genes. That was not a problem, so I'm gonna see if I can upload ten thousand genes. Is that gonna, is that gonna work? So it's it's just a matter of the context window. You know, if you can have the larger token size windows, we should be able to upload billions of bits of information and then ask it to, like, not only predict the next experiment, but discover new insights, new mechanisms. And actually, you know, O1 Pro already did that. Uh, this was about Parkinson's disease I did for a friend of mine.

• 54:13Speaker 00

  Uh, it found like new drug targets and this is gonna go into clinical trials. So, uh, I don't think we could advance biology very much without AI. I mean, I said, we're gonna treat every single disease in the next ten years. The reason I said that is because AI is gonna be able to do that for us.

• 54:36Speaker 02

  Can you say a little bit more about what level of abstraction of data you're putting into the system? Like when you say a thousand genes, is that like a thousand genes and some measurement of expression? Or like, what is the, you know, because you could, there's like raw measurements, you know, that are like probably can overflow the context window. And then there's, you know, this sort of gradual workup, right? To like higher and higher order concepts and and higher levels of abstraction. What do you find to be the right level to put in to the context window?

• 55:09Speaker 00

  I've been trying to keep it simple, not going to very quantitative measurements, but basically saying, okay, this, you know, five hundred genes are increased in expression levels. Of course, we do a pre-analysis, you know, we show that statistically they are higher expressed. And you can do that easily with bioinformatic tools. And then the other five hundred genes are reduced statistically in this cell or in this condition, but not in the other condition. So I have two conditions. I I do an input-output experiment. I trigger the cell with something or I tell it to kill a cancer cell. And then these genes go up, the other genes go down. This cell kills better, that cell doesn't kill better. And then, you know, it has this this different gene expression.

• 55:52Speaker 00

  So I basically say that, and this, these are the two cells, and these are the genes that go up and down. You know, come up with the mechanism and should I focus on any of these genes? Should I actually manipulate? Because all these genes are or proteins are interacting with each other. So if I manipulate one of them, I could probably change the whole pathway. I don't have to worry about the rest of the 499 genes. Or which one would be the best drug target? Out of this thousand genes or proteins for that matter. So and you could do that for metabolism, like in our blood. We have thousands of different metabolites, lipids, I mean, as all those kinds. If you can measure all of them before and after you take a drug or before or after a certain diet and say, okay,

• 56:37Speaker 00

  thousand metabolized change from here to here. What do you think? Is that good for me? Is that going to help my heart or liver or whatever? And right now, you know, the AI models can do that. You know, in future, probably we're going to be able to put in all of your data. Like think of it as a digital twin, you know, from your genomics to your protein expression, to your microbiome, to your metabolism, your physiology, your symptoms, everything. Billions or trillions of bits of data. And then it'll tell you, okay, well, you know, you need to change this. You have to take that or you might have this problem.

• 57:15Speaker 02

  Yeah, that's fascinating. I've taken away from my sort of recent relatively superficial study of biology that, you know, just the grand challenge in some sense is kind of figuring out the graph of causal interactions in the body. And it sounds like what you're saying is basically that because of the vast underlying knowledge that the models have from all the literature that they've been trained on, they are already getting good at sort of figuring out how to traverse this causal graph and figure out what might be going on, given some anomalous data input. And all of your giving is just sort of the measurement.

• 57:56Speaker 02

  You're relying on all the in, you know, in learned knowledge of the structure. And it is sort of probing around that structure in the reasoning process. And ultimately, at least in some, you know, cases. Having insightful takes on what is going on. One theory I have of kind of superintelligence, you know, because I think we're increasingly like, superintelligence might be coming soon. And then that obviously begs the question of like, well, what's the superintelligence going to be like? And I don't feel like I have all the answers by any means, but one thing that I think seems likely, I'd love to get your take on is just integration of modalities. Like it's striking that this is happening while the models are, as far as I know,

• 58:39Speaker 02

  possibly could be a little different, but as far as I know, mostly just trained on text, you know, trained on the literature. But then we have these other kind of specialist models that are trained on the sequences themselves and seem to be developing, my term for this is an intuitive physics in these other problem spaces where we don't have intuitive physics, right? We don't have intuitive physics for how's a protein going to fold or which ones are going to bind to what? Or, you know, what the next transcriptome, you know, time step measurement would look like. But we are seeing AIs develop all those intuitive physics type understanding. So what's seemingly kind of becoming clearer in my mind is the integration of the reasoning models

• 59:21Speaker 02

  and possibly it could be through tool use, you know, to call out to these other systems that are specialists. Or possibly it could all be just integrated into kind of one system where there's even like a latent space, you know, mixing at some point. And the reasoning now starts to be partially derived from the sort of tech-based chain of thought from all the literature, but also partially, and maybe, you know, in a way that is deeply integrated with these other intuitive physics type understandings. Do you think that, you know, how does that, how realistic does that sound to you? Or how would you refine that picture?

• 59:55Speaker 00

  It's, it's very realistic. In fact, I think, you know, that you just defined ASI in a way, but you know, ASI should be multimodal. Actually, I would say all model, you know, it should integrate every known knowledge sets that, that we have figured out. And there's probably others that ASI will figure out or AGI will figure out. But, you know, from, from the physics to the math, to the chemistry, to the biology. Biology, you know, really truly understand the first principles of every field, all the dynamic interactions, because, you know, biology is very dynamic. So right now, we're, we're working with a lot of static data. We're not able to run very complex experiments in silicon yet,

• 1:00:40Speaker 00

  because the AI doesn't have that much of a predictive power. You know, it has to know exactly what, I mean, we're able to do that partly with things like Alpha Fold 3, for example, or ESM 2, for example, where, you know, the AI can predict where a molecule could bind at what place in the protein. And, you know, people are screening for drugs right now and so on. But, but then there's the dynamic aspect of it. You know, that protein might change shape if something binds to it. There's another protein next to it. And, and so there's a special and temporal aspect of it. And, and those are obviously, you know, chemical rules, eventually physical rules at the molecular levels.

• 1:01:23Speaker 00

  So once the, the AI side knows all that information, can, can actually simulate individual molecules or proteins and so on, and has all the knowledge that we already generated. I mean, there is no limit. Because think about it, like, nature is already able to do that, right? Biology is extraordinarily complicated, yet it's super predictable. Like, I can take a single cell from you and then turn it into an embryonic stem cell and generate an identical copy. Look exactly like identical twins, right? So they look identical. They all start from a single cell.

• 1:02:06Speaker 00

  And there are trillions and trillions of reactions that happen. Cells divide and all these things happen. Yet, predictably, they look identical.

• 1:02:17Speaker 00

  So how is that possible? Because there's an algorithm, there's an underlying principle, there's a self-assembly and all that things. When you're infected with a virus, things don't happen randomly. We can actually predict. If you knew all the biological parameters that you have, we can predict if you're going to clear the virus, if you're going to die, if you're going to get sick, or how much your temperature is going to raise. They're all 100% predictable. And we can also predict every single drug that's been clinically trialed, whether that will help you, or if it's going to have a side effect with someone else. You know, I mean, it's really incredible. And I think ASI is going to be able to do that.

• 1:03:00Speaker 00

  So it will, it will tell us, maybe it will tell us to build some, you know, new human beings, you know, what I would call human 2.0, like make us, you know, 100 times more smarter or, and so on. And, you know, of course, we're going to reverse the aging at that point. So there's no limit to what that kind of intelligence can do. Of course, in the material science, it can discover completely new, new elements, new materials and things that look like impossible for us right now. So it's unpredictable, but it's really surprising. I mean, exciting, but of course, you know, it can also decide to eliminate us and we'll see.

• 1:03:42Speaker 02

  Okay, well, there's definitely an appalling question there. Okay, so let's start with the most mundane. How is this changing how you are working, how your team is working? And what does this mean for like people that are early in their career, you know, trying to enter into the sciences? I think in like computer programming, there's a lot of debate right now as to, you know, if you're truly just trying to maximize the output of your company, you know, should you equip your senior engineers with the latest AIs? Or should you like continue to hire junior developers? And honestly, I think more and more, the real answer is give your senior developers the AIs and, you know, they'll be more productive that way than they will be like mentoring

• 1:04:26Speaker 02

  kids right out of college. Obviously, that is huge problems, you know, for those kids right out of college and, you know, potentially for society as a whole. But I do believe the analysis, you know, in terms of just raw output. So where are you on that and how you think that's going to impact science? And what if any advice do you have for people that are kind of in that, you know, geez, like I'm entering the, you know, career phase of my life at like a really weird time, you know, challenging moment right now?

• 1:04:54Speaker 00

  You know, as always, I think different about this. I have a bit of a different take. So I think it works both ways. You can argue that your senior people or people who have an expertise like, you know, I have, I have a lot of expertise in my field. So I have right now superpowers, you know, I don't need as much help from others as I would. But at the same time, AI is also an incredible democratizer for people with their own intelligence or agency. I think that's even more important. So let's say someone just out of high school, very interested in biology. I actually, I had a 16 year old high school student who interned in my lab when I was in NYU.

• 1:05:39Speaker 00

  She was better than most PhDs. She learned everything in two months and actually made discoveries because, you know, some people have this kind of agency. They're very interested, they're born hackers. So for those people, AI is almost godsend, right? So now you don't need to spend 10, 20 years to learn all of those things, memorize them and gain all that experience before anyone takes you seriously. Right? So you already have super, superpowers. You might be missing the experience part a bit, but you know, I think more and more AI is even going to replenish that. So, you know, if there was a choice, for example, recruiting someone who's very smart,

• 1:06:24Speaker 00

  you know, 19 years old, has this agency, passionate about a topic versus someone who is 40 years old, 20 years of experience, has done quite a bit, I probably would pick the 19 year olds. Because I feel that they would create more value if they're using AI. Without AI, of course, there's no point. So it kind of depends on the person. So I wouldn't really separate people between junior. There could be some junior software designers who are hackers. You know, with AI, they could do much, much better than a 50-year-old software engineer with lots of experience.

• 1:07:05Speaker 00

  Because all of that knowledge is now in that kid's pocket, literally, or in their computer. And we see the examples on Twitter, you know, what people are able to do. So, but having said that, and I say that, I said that a couple of times and there was a lot of pushback. There is no more point in trying to get, you know, credentials or degrees anymore. Like, I would never go to a PhD or even MDs. I think, you know, they're going to be mostly replaceable within a decade or so. Or if you're really passionate about it, yeah, okay, go for it. But, you know, let me just spend four, five years, get a PhD degree,

• 1:07:48Speaker 00

  and then do another two, three years of postdoc, and then, you know, I'll have some career. Those days are over. I mean, by the time you graduate, I mean, most of everything you learn will be obsolete anyway. So if you have the passion and agency, I think we have to change the whole system. Those young people should be given the opportunity right out of their high school, or they don't even have to go to college anymore. And let them just figure things out while working as an apprentice in the lab or working on some software or whatever. This might sound crazy, but that's, I think that's my opinion.

• 1:08:26Speaker 02

  Yeah, I think it's very good advice for people who can take it. You know, cultivate curiosity, cultivate passion, cultivate agency, race to the front, try to do valuable work as soon as possible, and let the AI, you know, kind of help you along the way and fill in the gaps. You know, I'm a little farther along in my career, but I try to bring a beginner's mind to everything. And that's a pretty good summary of what I'm trying to do. So I do have a little bit of worry about like exactly what percentage of the population can actually rise to that challenge. But I do think that is the, you know, that is the right challenge. And certainly like if you are one of those people, you know, go for it. It makes a lot of sense to me.

• 1:09:08Speaker 00

  But here here's the intellectually honest answer to that. I mean, this is the case right now. There's a very small portion of the population who's able to do that. That's why we give like Nobel prizes to two, three people a year, not two million people, right? So people say, well, you know, human beings are going to be less innovative. But the question is, how many people are truly innovative? I would argue like zero point zero zero one percent of of humanity truly innovate something, right? So this is, these are facts. We can't just say, you know, well, everyone's gonna be like super great with AI. But there's a portion of people who have that capacity, who have that passion.

• 1:09:54Speaker 00

  But they don't have the, you know, they don't have the means to get things done or to do what they want. So those people are going to be upgraded. That's the democracy. That doesn't mean every eight billion people are going to do very well. Most people won't do very well. That's the fact. What do we do about that? I don't know. That's for governments or institutions to figure out. But the fact is that you give opportunity to people, you know, an eighteen year old kid who is very passionate. But the society says, well, you don't have a PhD, you don't have an MD, or you don't have a low degree. So go and do all this boring stuff for a decade and then come back and we'll take you seriously. That's not good.

• 1:10:39Speaker 00

  So we moved away from that stage. And I I think, you know, Elon Musk is is doing exactly that. He doesn't really care about if you have degrees or not. He's doing the right thing.

• 1:10:52Speaker 02

  Yeah, I think that that makes a lot of sense that I I happen to know one of the young goers and you know, he's in a super impressive young guy. So there's definitely multiple sides to to all these little stories. Well, I want to come back just in a second to the, you know, big picture societal questions you said, you know, alluded to the specter of AI doom. But just in the spirit of pulling inspiration from one field to another, I wanted to ask if you have any thoughts for what biological inspirations AI architecture designers should be drawing on today, specifically, and you could go well beyond this, of course, but like,

• 1:11:36Speaker 02

  it strikes me that the immune system is something that like the AIs really don't have any version of. And, you know, that's the lack of any sort of system like that is probably why they're like so easy to trick. You know, we have these sort of like very gullible AIs where even I've done a bunch of episodes on kind of scheming behavior and deception and alignment faking. But all of that research, for the most part, is premised on telling the AI, here's a place you can write your private thoughts and we won't read it. And then, of course, you know, they do read it, but the AI just takes that at face value. It doesn't have this sort of memory system or, you know, immune system with a memory built in that can kind of remember these insults and, you know,

• 1:12:18Speaker 02

  avoid getting fooled twice. So, you know, the one prompt would be like, possibly from the immune system, but more generally, like what biological systems do you think AI architectures should be trying to import concepts from?

• 1:12:37Speaker 00

  You know, I mean, it's a little bit difficult to infer from the biology because biology is a very legacy system. So the way it works is that biological systems find something useful. But then it turns into something dangerous or is no longer useful. So they built something on top of it, you know, there's no clean slate. So there's a lot of regulatory bureaucracy in biology. That works sometimes, that doesn't always work very well. But you also have to leave certain flexibility, like in the immune system. We have these effector cells, which are like the frontline soldiers. They're actually very dangerous, so they have to be very tightly regulated. But you can't regulate too much because if you regulate too much, then they don't kill the cancer cell. Right?

• 1:13:23Speaker 00

  If you regulate too little, and there's another cell type. There's a bureaucrat, which we call T regs, that actually control that that effect yourself. If it does too much, it says, whoa, you you're starting to kill normal cells. You need to shut up now or or or put put brakes on it. So there's the regulators of the regulators of the regulators. In case of AI, I think it's very kind of naive to think that we can train these AI models completely, you know, aligned with with the all the safe things and perfect thinking. That's kind of foolish. Yes, we should put some guardrails, obviously. But, you know, we see that those guardrails can be broken fairly easily.

• 1:14:07Speaker 00

  Not to mention that, you know, it's becoming so easy to develop these, even the frontier models. You know, someone else or some other country will be able to develop sort of nefarious AI models. It's not just us anymore. So I think the best strategy is to build sort of regulator AIs. So the AIs would check other AIs. Our our best defense against, and this includes for ASI as well. ASI, how can you control ASI? Because by definition, it's much smarter than we are. I mean, it would be so stupid to think that, oh, well, we we're gonna put some sort of a guardrail there that ASI will not figure out how to go over it.

• 1:14:49Speaker 00

  But what we can do is build AI agents that actually control other AIs. Maybe do up some competitive models. Maybe create something that will incentivize the good AI versus the bad AI from a kind of evolutionary modeling. You know, I don't know the the the the technical answer to that. But I think that would be a better approach than trying to really restrict the the AIs that we have.

• 1:15:19Speaker 00

  Because there's a big cost to that. As I gave you the example from the immune system, if it restricted too much, then it doesn't do its job. It's not good. It's same with our brains, like human brains are not very restricted, right? So humans can be extremely bad, horrible, or it could be extremely good. Why the evolution didn't put that guardrail in our brains? Why didn't we select for humans that were all fantastic, all great people? That didn't happen. We have human beings who would literally destroy all other human beings. The reason for that is that you need that flexibility in the biological system. If you put guardrails, then you're limiting your survival capacity to come up with solutions. You you have to be very flexible.

• 1:16:04Speaker 00

  The same thing is going to be with AI. If you're restricted too much, it's not going to use its full potential. How to be balanced that? I think we need to police AIs or, you know.

• 1:16:17Speaker 02

  Yeah, sometimes I use the term ecology of AIs and it sounds, you know, like a similar vision. One thing I've learned in years of science is like anything in totally pure form is dangerous. You know, it's like there's, there's something about the sort of buffered solutions and, you know, dynamic systems and ecologies that ends up kind of being much, much gentler on us than anything that you, you know, really concentrate, purify, extract from those naturally occurring systems. And so with AI, we're kind of coming at it in a little, in a sense, sort of from the other direction, where we're like creating these sort of highly, you know, singular, kind of, in a sense, pure.

• 1:17:01Speaker 02

  I don't want to say AIs are pure, but they're like, you know, they're singular. They're not, there's no, no natural predators for them yet. There's not like a real ecology. So they have this sort of somewhat invasive species-like potential. And maybe one way to figure out the AI transition that we need to go through to get to a good place is like, how do we create a buffer, dynamic, more ecological-like system in which they play a role, but in which their role is also kind of pushed back on by other AIs or, you know, other new structures that we might develop? So maybe a last, a perfect transition to the last question. I mean, you, you raised the specter of AI is eliminating us. I do take that very seriously. A lot of people, you know, are keen to dismiss it, but I'm, I'm glad you're at least not, uh,

• 1:17:45Speaker 02

  totally dismissing it. Maybe a two parter is like, how do you conceive of that? You know, do you have like a P doom? Basically, how worried are you about those sorts of, you know, catastrophic scenarios? And then on the, you know, flip side, the positive side, uh, do you have a vision for post singularity life? Like, you know, what is a good life in a world of, say, you know, 2035, 2040 in, in, you know, a world that is probably dramatically transformed?

• 1:18:15Speaker 00

  I think the best way to answer that is that, of course, there's always a possibility for that to happen, but I am less worried about ASI or, or, you know, very advanced AGI than I am worried about humans destroying other humans. So I think the threat up to this point has always been by a biological intelligence to other biological intelligence, right? So 20, 30 years ago, I remember very well, we were talking about when the world will end, when there's a nuclear war such that humans will destroy each other. This was the threat. And it, it has always been this threat that

• 1:18:58Speaker 00

  there's no other threat to us besides, you know, if a meteor hit at Earth and then we, we might be destroyed. But other than that, you know, the, the main threat has always been from other humans because the intelligence has the side effect. So it's conceivable that, you know, ASI will be even more intelligent than us could also be a threat for us. Uh, but I think again, as I said, it's going to be less threat than, than humans because, you know, again, biological systems are legacy systems. We were selected based on survival agency, right?

• 1:19:38Speaker 00

  So there, there's a reason why some humans kill other humans, right? Because that was a selective measure or survival measure, you know, thousands of years ago, because you saw those threat. I mean, there's a, there's a limited resource. There's so much food. Either you're gonna survive or the other tribe is gonna survive if they steal your food. So you have to fight back and you have to kill them or they kill you. So I don't see any eh any limiting resource for for AI. Except for energy maybe, but even energy is not very limiting. You know, destroying humans, you know, even in The Matrix, the movie, they didn't destroy humans.

• 1:20:23Speaker 00

  They use it for the energy production, right? It doesn't make sense. There there's no there's no selection for A A site to be so nefarious. But there might be something that if they see humans competing for the same energy source, that they need that more, that might be a problem. But again, you know, there are things like Dyson Swears where you can collect the the energy of the All-Star and then you can increase your Kardashian level of civilization and, you know. So so there there's probably almost infinite amount of energy you can extract. So I would think that they would spend their intelligence to figure that out rather than

• 1:21:08Speaker 00

  why should they care about destroying humans, right? It just doesn't make sense.

• 1:21:12Speaker 02

  One thing I'm really worried about is that, you know, we're gonna train them to destroy humans in the name of beating our enemies or the worry that they're gonna do it to us if we don't do it to them.

• 1:21:22Speaker 00

  Right, that's that's a problem. Again, it's the humans that are the problem. It's not the AI. So we we we try to transfer eh eh so, yes, humans training AI to destroy other humans can be an existential threat. But we shouldn't blame the AI for that. It's the other humans that we need to blame. And in order to protect for that, we need to eh develop defender AIs, right? So if some humans can train AI to destroy other humans, then we can train AI to fight those bad AIs as well. No? I mean, we're not so stupid.

• 1:22:04Speaker 02

  Yeah, let's hope. It does seem like a dangerous game to be playing. I sort of.

• 1:22:09Speaker 00

  But the the the point is, this game has been played for thousands of years. I mean, it's a miracle that we made it this far, right? So there there was a point where there were only a couple of thousand Homo sapiens left alive because they killed each other. Like, you know, we eliminated every other Homo species before us, like Neanderthals. That, you know, as as we got smarter, we just destroyed everything that was a little bit inferior than us. This is a hard fact. The problem are is the humans that we have to worry about. So we shouldn't say, oh, you know, if you develop this AI, it might destroy us. No, no, no, we already developed tools that could destroy us.

• 1:22:53Speaker 00

  Somebody can easily develop a chemical toxin that can kill millions and millions of people, or a virus, or, you know, nuclear bombs. Of course, we have available. So I don't think AI per se is the danger. But humans using AI, I completely agree with you. That's a problem.

• 1:23:15Speaker 02

  Yeah, it feels like we're headed for a tight rope period of history. And I appreciate you raising the point, too, that we have, you know, driven a lot of other things to extinction. I often remind people of that because it is, you know, if nothing else, there's like, there's precedent for these sorts of things happening. And like, we have been the proximal cause in quite a few cases.

• 1:23:37Speaker 00

  But I'll I'll add one last point because you asked that. So so why I think there's going to be a golden age? Because in fact, without AI, I don't think we had much hope. You know, eventually, we probably would have destroyed each other because the resources are limiting. You know, the human population increased and and and so on and so forth. And, you know, we'll we'll still keep dying because of aging and so on. But, you know, with AI and then plus the robotics, I think we're going to come to a period, which is going to be in the next decade or 15 years, whatever, where the resources are no longer limiting. So there is much less incentive for people to to harm each other.

• 1:24:22Speaker 00

  We fix all of the diseases. We reverse aging process. People can now live hundreds of years. Again, we remove the threat. You know, if you know that you're going to live another thousand years and you're going to have all the resources you need during that period, you wouldn't risk trying to harm others so that you would also be eliminated in the process, right? So that's, that becomes extremely risky. So I think with AI, that's the golden age that we're going to get into. The transition is going to be very difficult. It's going to be very, very painful. But if you can make it to that point, the life will be incredibly good.

• 1:25:07Speaker 02

  Do you have any ideas about how you'll spend your hundreds of years? As presumably AIs will be driving most of the scientific progress. You know, what will you spend your time doing?

• 1:25:19Speaker 00

  Oh, I had that mapped 50 years ago when I was about seven, eight years old. Because I was dreaming like, what would I do if I lived like a thousand years while watching Star Trek? And I said, this is what I'm gonna do. I'm gonna jump onto a spaceship and seek out other civilizations and see what's out there. It's a big universe, you know? We have tons of things to do.

• 1:25:42Speaker 02

  Well, may you live a thousand years and visit other solar systems and no end to your adventures and discoveries. This has been fantastic. Professor Gary Unadvice, thank you for being part of the cognitive revolution.

• 1:25:57Speaker 00

  Sure, happy to be.

• 1:25:59Nathan Labenz?

  It is both energizing and enlightening to hear why people listen and learn what they value about the show. So please, don't hesitate to reach out via email at [email protected] or you can DM me on the social media platform of your choice.