Why the Next 10 Years May Add 50 to Your Lifespan | Dr. Derya Unutmaz

This is probably the most critical time in human history. So try not to die [music] for the next 10 years. >> Can you explain and unpack why you think that? >> The reason is that the technology because of AI is expanding exponentially. Cancer is going to be 100% curable probably less than a decade. We'll get to a point where we'll have hundreds of new drugs coming out every month maybe. And then we'll get to a point uh probably 15 maximum 20 years where we will be able to completely reverse the aging process. So if you're 80 years old, 90 years old, you will get back to [music] uh age 30, 40, whatever. >> Why do you have such an optimistic view? >> AI is an incredible enabler. It gives you superpowers. [music] The key risk is is humans. Humans misusing AI. There's only [music] one existential threat to humanity and that's humanity. Hey everyone, today's episode explores an extraordinarily exciting convergence, the accelerating pace of artificial intelligence and a growing optimism about the future of science and medicine. In this episode, I discuss with Dr. Duria Enautmas how AI could dramatically improve our ability to detect, prevent, and treat human disease and ultimately extend human life expectancy. Before we begin, I just want to mention one quick thing. Only about 30% of the people who watch this podcast are subscribed to the YouTube channel. Taking a moment to subscribe and enable notifications is one of the simplest ways to support the show and help us bring these conversations to a wider audience. We greatly appreciate it. Thank you so much and I really hope you enjoy this episode with Dr. Duria Enutmas. I'm so excited to be sitting here with Dr. Dura Unutmas who is one of the handful of scientists that has had access to collaborate with open AI one of the you know world's leader in in artificial intelligence. He's also an aging researcher. He's an immunologist really just a matchmade in heaven to sit down and talk about the role of AI in in aging research and in medicine. So I'm super excited to have you here today. I'm very excited to be here. Thank you. >> As we both know, aging is a very, very complex process. Many factors involved. It's heterogeneous. It's so complex. And it just seems like so almost impossible to solve. And yet, I've heard you say something that's very interesting. I've heard you say if you could try not to die within the next 10 to 15 years, you might want to try to do that because you could live an extra 50 years. >> Yeah. >> Can you explain and unpack why you think that? What makes you believe that? >> Thank you. So, first of all, I'm very excited to be here. I'm a big follower of your podcast. I think it's maybe the best uh aging or longevity podcast. So, uh this is this is a great pleasure. Um yeah so I've I've said that um quite a few times uh in the last year or two actually um and it may not even take 10 15 years might be uh even closer. Uh the reason is that uh the technology especially because of AI uh is expanding exponentially. So our minds think in a linear term. So we think that the next 10 years is going to be as much advanced as the last 10 years or the last 15 years. But that's not what's going to happen. In the next 10 years, you can think of it as more advanced than the last century. So imagine that you were living in early 1900s. Uh and somebody told you that you know we're going to uh have vaccines and you will never get small pox or uh you won't die of tuberculosis. You know people would laugh at you. So that's not that's not possible. Um so so that's this the speed that we're talking about. But there's something uh even more important because of this acceleration. [clears throat] The u the advances of treating diseases is also going to accelerate dramatically. So uh we will get to a point what's called the longevity escape velocity. This was coined by Aubry de Gray who's as you know is a great uh aging um researcher. Uh so the point is that we will come to a point in the next I would say probably eight to 10 years where every year you live is going to add more than a year to your life. So let's just say um you know 10 years ago 10 years later you get a a cancer uh that's normally is not curable and you only have one or two years to live. uh um but that during that one year there is going to be a new treatment that will cure that cancer. So automatically it's going to add several years or maybe 10 15 years to your life or uh we're already starting to see that with the GLP1 uh drugs uh receptor agonist which uh which are adding about 5 to 10 years to lifespan of people who are obese or who have chronic uh conditions uh will have sort of the muscle generators uh uh which I think will have tremendous impact on the aging population because as you know that's a huge problem. So all of these things will add up and and and the technology and AI is going to keep accelerating. So 10 years later uh what will happen in a year will be like what happens in 20 years of advance and then we'll get to a point 15 maximum 20 years where we will be able to completely reverse the aging pro process. So if you're 80 years old, 90 years old, you will get back to uh age 30, 40, whatever. So that's going to add up uh 50 years or 100 years to to your lifespan. Um and then you can keep doing that and extend it almost uh indefinitely. So I think this is probably the most critical time in human history. So try not to die for the next 10 years. >> And we're going to talk about all these things. I want to talk about curing disease. I want to talk about reversing aging, age reversal. Um, all of that is on on on my agenda to talk about with you today. But you mentioned something. You mentioned that right now the, you know, artificial intelligence as a general term, you know, is is accelerating at was an exponential rate. I've heard you talk about this Moore's law and how the, you know, the software itself is accelerating right at this exponential rate. Um maybe you could explain a little bit about like what what does that mean and then how how do you think that'll translate into biology because you know humans we're not software and there are things that at least in my opinion you know you have to still test safety right I mean so like if you're you know accelerating the computational speed and therefore you can test a lot of things that are what are what's called incilico for people listening we're talking about testing things like just modeling them and maybe you can explain this a little bit better. Um, but then at a certain point you still have to test about, you know, safety and you you definitely that that there's there are things that I think need to still be done in human trials. So I'd love to hear how you think that's going to happen. >> I think that's the the most critical question because people always bring that up. Okay, you know, if you generate drugs uh within hours, you still have to test them on humans for five years, maybe some sometimes longer. How how are you going to deal with that? But let me let me first uh start with how AI is accelerating biology now. So we we can think of it in in terms of phases and and because now and 5 years later is going to be very very different. Um so right now especially in the last year or two since uh you know LLM came out um you know their intelligence have been accelerating. Initially it was uh fairly um smaller u productivity gains. For example you know when GPT4 was was out I I would ask it to sort of scan the literature and tell me what's the the latest on this topic or that topic. Um and that saved me you know hours sometimes days. Uh but then as the models advanced especially the after01 model the reasoning models uh started to come out and and now we have the GPT5 uh pro model 5.5 pro model. Uh what happened was that now they were able to think and plan. So uh you could start to ask very sophisticated questions. For example, here is a huge biological data set, a million data points or 10 million data points. Go over this. Not only just analyze it and group them, but uh what what is the insight from that data? Uh human mind is not able to do that. And in fact, we had such data sets which took us months to analyze like you know a PhD student work on it using deep learning. we still couldn't really truly understand what that data meant. We we know these genes are up, this metabolites are changing, this is happening. How do you bring all that together? Um, and so now AI models are able to do that. So you I I've tested for example uh latest GPT5 uh pro model. You can upload uh millions of data sets that we accumulate over years maybe uh and then in matter of minutes you get not only the complete analysis recently I had a 40page report from GPT5 uh pro um which was an analysis of this what's called the RNA sequencing lots of millions of data points but it also provided incredible insight like what is the what does this data mean what should be the next questions to ask so that automatically contracts months sometimes years of analytic work into matter of minutes or hours. Um so so there that that is already accelerating of course in the drug uh design uh parts uh I think every pharmaceutical company is going to eventually use AI generated AI generation for developing new drugs. things that took years of screening of small molecules now take you know hours or days. So, so tremendous acceleration there and then uh I think again more recently because the models have advanced so much that you can also ask things like okay so uh this is great um this is the hypothesis in fact AI can even generate hypods for you but what sort of experiment I should do to address that uh people have to realize that uh what we do in in biology is experiments but we don't really know what's the best experiment to do. I mean that's kind of my job but I have some intuition we should do this to address that question but is that the ideal experiment is does that have all the controls everything so AI models are now able to tell you sort of simulating if out of this 100 potential experiments you can do this two are the best ones because this is going to give you the best uh output and I' I've been testing that so so that is another acceleration now you don't have to try 100 things for a year, you can just try two things for few weeks and and get get the output. So that's what's possible now already tremendously accelerating the R&D part. But then uh the second part which I think is more important part is how do we uh apply that to clinical trials and regulations. Um, so it still takes years to try everything on humans and I think the solution to that will be what I call the digital twin and this this term is around for for several years. So the idea is that if we have lots of lots of biological data and when I say lots it's it's a lot pupy bytes of of data. If if AI comes to a point where we're going to need much more compute than we have today today is to able to compute all that and really kind of simulate a whole biological organism, a whole human being, but not just your um phenotype but but also your metabolism, your immune system, your gut microbiome, uh your genetics and and all kinds of data sets are put together. And so it it it knows your biology in a temporal way in in in a in a totally functional way. Then you can ask the question okay so if I give this drug to this person what kind of effect it will have if they have this disruption is it going to have a side effect or is it going to be effective. So literally we can cut down clinical trial time from years to to a matter of months or or even weeks. So you can actually do the trials in a very small subset of patients because you can choose the patients. You can say okay AI told me that these these these people this drug is going to be effective 100% to them. And so so you just test it on those people and in fact that will go into the personalization. There's going to be thousands of drugs for for different people. So that that will cause tremendous acceleration. We're not there yet, but I'm I'm betting on that that within the five five to 10 years we will get there. So, uh the iteration process the on humans is going to be all digital as well and then maybe the the manufacturing will be a little bit um uh still will take time but but we can even improve that part too. So uh we at some point we will come uh to a point where treatment on demand. So you go to an AI model analyzes your genome, your biology uh orders the this small molecule or the drug or treatment just for you to the manufacturing facility and next week you get your drug and and you get treated. That's the world I'm imagining. So I want to get back to this concept of digital twin um again when we talk about personalized medicine but if I understand correctly so you know if we are if we have this digital twin which is all the genetic data metabolomic proteomic biomarker just everything right all this data um and more that we're not talking about um and and now we have AI which can then you know do all these scenarios and figure out like how this drug is going to affect or how this treatment is going to affect this And you're saying that the clinical trial that may have taken, you know, a few years can be condensed down and perhaps we can look at after doing the incilico experiments, you can look at some biomarkers and know like is this going to affect their fertility like you don't want to give some some someone a treatment that's going to make them infertile or you know so you think that's going to be uh AI is going to be able to identify how to know if it's going to affect like fertility or cognition or life expectancy or you know just just from >> the whole composition of the person and doing I don't know all these tests. >> Yeah. So uh I mean the path there uh requires uh several steps of validation uh and that I think we will get to a point where when we have super intelligence that we'll be able to trust super intelligence you know almost 100% that we don't need to validate it even with biomarkers or or whatnot but to get to that point it's sort of like the self-driving cars right so um to get to a self-driving being leveled. I mean, it has to be 99.999% safety. Um, you you have to sort of validate it. Um, uh, you know what happens if somebody's crossing the street, right? So, so that scenario has to happen and then you you record it and sometimes uh you won't do the right thing. Maybe, you know, it won't stop. That's why we still have to like look at this, you know, be ready to to take control. But if it does stop and it stops uh and and saves lives again and again and again and right now you know self-driving cars are probably about 10 times safer. They will be maybe hundred times safer. So you get to a point that you trust the AI rather than the the driver, right? So you say okay so I I trust I want the AI to decide for me uh to to drive. So I think we'll get to that point for biology. it will take a little bit longer uh because of the extreme complexity. Um and then we'll have to have uh very um clever benchmarking and validation uh ways there. the biomarker is going to be really important because again, you know, if you're developing an aging drug uh that you claim will let people to live to 150, well, you can't wait uh you know, even even if somebody 100 years old takes it, you still have to wait 150 years, 50 more years to to validate that. So that that's not going to work out. So we have to be able to predict that. But but actually probably aging is is the easiest in some ways uh to predict because uh we have so many biomarkers or functional outputs we can measure. We know how they are in an old person and in a young person. So if your vi suddenly gets uh you know like a 20-year-old wow that's amazing. If your muscles are as good as a 30 year old, uh if your skin looks uh like a 20- year old, that's what my mom is waiting for. Uh you know, that that's that's proof. And you'll you'll immediately see that. I mean, immediately weeks or or or or or whatnot. So, I think um again, it will take time. That's the part that's going to take time, the sort of trusting AI to um to tell you yes, if you take this drug, you will you will be treated or you will reverse aging. Um uh we we we still have about a decade. That's why I'm saying like you know other otherwise it would it would take it would happen even earlier. You mentioned super intelligence, artificial super intelligence, ASI. Maybe you could talk a little bit about just for people to have an understanding right now the difference between artificial intelligence, artificial generalized intelligence, AGI, then the super intelligence because you said >> once we get to the super intelligence, we're going to trust it, right? So I I mean I don't know do we know what those differences are or can you explain a little bit? >> Yeah, of course. you know this changes on a daily basis what the definition are depending on who's uh whose definition uh but the you know I've been thinking about AGI ASI for decades I mean it's not something that I started to think about it recently um and so the way u I originally uh defined AGI it's it's artificial general intelligence so what that means is that first of all it's artificial right so it's not human intelligence artificial intelligence and then it's general. What that means is that um if AI learns uh one set of uh rules or one set of knowledge that it can generalize that to something else and that's how our brains are intelligent um because uh you you can be an amazing chess player. In fact, you know AI beat the chess champion Kasper in 1997 I think like decades ago but that was not general intelligence. It was super uh good or Alpha Go beat you know the the the world champion in Go which is much more difficult uh uh game uh to be general uh Alph Go you know learning how to play Go or chess should be able to I don't know solve a problem in aging right so it should be able to transfer that information I think the uh the amazing thing about LLMs what we call large language models is that they acquire this ability uh which honestly I didn't think uh this would happen so so easily. I was expecting AGI to to happen uh maybe a decade ago. So in my opinion we have already achieved uh what I call level one AGI artificial unit because if I ask GPT5 uh pro model you know something that hasn't it hasn't trained on like an experiment that I have done or if I say okay think of the experiment as a video game design another experiment for me like like you are playing a video game so that's transferring completely different area to a biological system and is able to do that in an amazing way. But we have we we still need to go through several levels. I I think the next level is going to be memory. So they don't have persistent memory right now. They have some memory. They know about you. Uh they know about what they've learned in the internet, but they need to be able to uh manage the context, you know, because there's a continuum. Life is a continuum. Um and then the the other one is going to be the self-learning right so maybe that's level three I it doesn't matter uh and that's coming soon you know AI companies are saying like we think that the real time learning uh is is is coming maybe by by next year um and then uh the third level uh uh what I call the physical intelligence so people again confuse this greatly because the true human level intelligence is physical intelligence it's not cognitive intelligence. So for uh millions of years, we evolved to survive in a physical world. We we didn't have language up to I don't know 10,000 years ago like we didn't know how to write. Um this cognitive part uh has developed in the last you know maybe 10 20,000 years. Uh before that in fact animals have very good physical intelligence. We're we're imprinted and born with that intelligence. So uh an admiral or a child knows already have a world model. They know that you know if I drop this it's going to fall and and doesn't have to test it a million times. And that's of course what we need for robots for embodiment. And and you can see that you know that's taken a long time. You know it's more difficult to train a robot to behave like a child than have GPT5 solve the most difficult math problem. So and we'll get there. I think people are working on these moral models and physical intelligence whether we need another algorithm or not. So that will be the the final level of the AGI level. Um once we have all those levels then and once the uh AI is able to self-learn um then that's the super intelligence because at that point it can train itself you know maybe thousands maybe millions fold faster than we are able to do. Um and and there is a there's a limit to human intelligence right so even the smartest person in the world can only do so much uh and super intelligence what I would define is that you will have the intelligence of combined totality of humanity at some point like if you if I bring uh a million top scientists in the world of course they can solve you know like a Manhattan project they brought all these brilliant minds it wasn't one person's uh they were able to solve very hard problems. Super intelligence will get to that level. We'll be able to do what thousands of scientists can do uh in a year will be able to do uh in a day. So um I I would probably trust that. >> Wow, [laughter] that's pretty exciting. I mean and it it also kind of brings in this this concept of when you talk to people about AI and not everyone has the understanding of it as you um for sure you you hear that there's there's a pessimistic versus optimistic view right and oftentimes if I talk to people I hear a lot of pessimism I hear perhaps they don't understand their fear of the unknown of what AI is capable of I mean you're just the super intelligence that you're talking about I feel if you explain that to some people it would scare them even more. You know, perhaps they are worried about the cultural ramifications, ep e economic ramifications, but also just this Terminator situation where, okay, well, they're super smart. They're going to want to then take over the world and they don't need us anymore, right? >> But you have such an optimistic view. I mean, we're talking about solving aging, living to be 150 or more. Uh why do you have such an optimistic view? Are you worried at all about the other pessimistic sort of viewpoints or >> absolutely not and I I'll tell you why I'm I'm so super optimistic about it. Um when people make those statements like uh AI is an existential threat for us and you know it's going to destroy humanity. Um I make the counterpoint there's only one existential threat to humanity and that's humanity. So if you look at history um human beings killed more humans than everything put together caused more suffering than anything that humans have have been exposed to. You know even animals I don't think they they maybe infectious diseases at some point might have caused um a lot of suffering but but but the real danger is is the human intelligence. So uh I do let's do a thought experiment. Let's imagine that uh we live in a parallel universe and in that universe the world have decided that anyone above the IQ of let's say 100 is a danger to the society because if you get very intelligent you can come up with ideas that could be very dangerous right and that's true actually that's how it happened. Um, and then if you if you have an IQ of 105, you get imprisoned immediately. So you you are not allowed to to participate in society or you get killed or whatever that the the society has decided intelligence is dangerous so we're going to stop it. Uh what kind of a world we would live in? We would not have anything that we have right now. we would live in probably just as farmers you know basic physical intelligence we have uh and and try to survive you know uh in a world where the average lifespan was 30 years old or something like that. So that's that's how we should view uh AI and and the other point is that [clears throat] about this uh sort of AI is going to take over and is going to replace us. Um uh I see it exactly the opposite because AI is is is an incredible enabler. It gives you superpowers. Even now I feel like I have superpowers. Uh you know I've never been this busy in my life. You know I I actually sleep less which is not a good thing by the way. I don't recommend it but be because I can do so much. It's so empowering. You know my mom was was was 86 years old. you know, she told me that uh chat GPT changed her life. She she she's energized. She she doesn't worry as much about her health and um um it's it's just been an incredible impact and this is going to accelerate and at some point we will get uh we will sort of merge with the with the AI uh in a way that we will have direct interaction with AI through neurolink type of uh brain interfaces. So we'll have the sort of the intelligence of AI in our own brain not not only directly but also indirectly by sort of engineering our biological system. So why why shouldn't everybody have an intelligence of Einstein or even higher right? So the difference between an Einstein and a normal person with with a normal IQ is probably few gene single point mutations. So if we can engineer that if AI can teach us how to do that then we are we're also uh going much much higher. So as long as we we keep the agency, I think that's the only thing that we have to really protect that we are the decider or we see AI as a collaborator as sort of another species that will live together and we empower each other. In a way it's our child, right? So it's it's been created by us. Um I I I see the chance of uh a a a worse world extraordinarily. Of course, it's never zero, but you know, the moment you're born, you're going to die, right? So, so you're you're destined to die. Um and now, uh AI is giving us this opportunity to save literally save billions of lives. I'm not talking about saving lives as like extending their life for 5 years or 10 years. You're talking about thousands of years. So that's true saving lives. That's the potential. And the risk is again I think the the key risk is is humans. Humans misusing AI. That's what we have to uh um sort of maybe train or align AI. You know, don't uh don't look at the bad humans. you know, you you you can you can judge the the the the better uh the the better world uh for us. So, um of course I might be wrong, but I'm I'm pretty sure I'm I'm going to be right. >> I I I agree with uh the statement of we have to watch out for the humans. Um for sure, like cuz you're right, like they can and have in the past been the biggest threat to humanity. So, um, I want to there there was a couple of things that you mentioned when when you were talking about, you know, ASI and this ability to self-learn and you're even talking about some of some of the ways that you use, you know, GPT5 Pro and and helping with designing experiments and interpreting results and and that was a question that I had as a biologist. And as you mentioned, you know, we do experiments. We're testing hypotheses. And then we have all this data and these results, and we have to know what result is meaningful and what anomaly is meaningful because often times the anomaly, >> right, >> which you might ignore. >> Exactly. >> Is what you absolutely is the breakthrough, right? >> And that is a sort of intuition. This this biological intuition. And so you do you think first of all do you think we're that that you know the models we have now can already are capable of that sort of biological intuition and if not like how far off is that? >> Yeah. Yeah. That's that's a great question. Uh in fact um uh you know I I see that intuition maybe sort of the the last mile or the top 10% uh or 10% of the of the solution because 90% um AI models they are able to come up with because it's it's knowledge based also in humans is is you know for for a medical doctor for a scientist for whoever 90% or 95% is based on uh what's known how you process that knowledge, but there's that extra 5 10% totally dependent on your intuition. Uh like you you if you're a doctor, you see a patient coming through the door, you know that guy is having a heart attack. You haven't checked anything yet. Somehow you know, you don't know how you know. the same thing in the lab like um uh in fact I would I would bet with my uh students and and postto I would say okay I bet you if you do this experiment you're going to get this result um and I've never lost a bet and they stop betting against me even though it might look counterintuitive oh no that's never going to work somehow I know how do I know because you know I've been working in the lab for 30 plus years and and you you acquire ire certain things that are not in the literature or you know you can't really read a textbook and learn it. You only do it by by practicing it. Um so the the models up to I would say 5.5 until recently were were great at that 90% level. Uh so especially after GPT5 pro came out. So you know I would ask it to for example I I would give it an experiment that we have already done. It's a very complex experiment took two weeks. I already know the result because we're done the experiment. But I wanted to see how the model would predict the outcome of the experiment. And they would do you know not just GPT5 but several other models as well. Um they they would come up with 90% 80 to 90% correctly. I mean that's that's pretty good. Uh they would say okay this is what's going to happen after two days after one week after two weeks. But that extra level of intuition that I I have I would have predicted was still somewhat lacking. I think GPT 5.5 crossed that threshold. So I I repeated that with with the uh 5.5 pro model uh because I I always say pro uh it's very different than the thinking of course very very different than the instant model because pro uh is reasoning much much longer. It's thinking. So in some cases I I pushed it to think for two hours. So two hours in AI thinking is like years of thinking for for a human being. So that model really crossed that threshold in that example I gave you. It was almost 100%. I mean I would say 98% correct. What I would have predicted like I would not have bet against 5.5 Pro myself. Um uh so that to me is is actually really mind-boggling because uh I couldn't understand these models are being trained with all of the information we can't compete with that right so it's they they can put these patterns together but how is it that the model has now almost the experience that I have that I spent 30 years acquiring that experience that intuition that is now getting to that level that is uh that is a mysterious but uh I I I live to it. Now >> what sort of you said you you pushed GPT 5.5 Pro to think for 2 hours. >> I mean what sort of prompt are we talking about or is it >> just the data set too and the prompt? I mean >> so those are usually data sets. Uh um I might have broken a record because I even asked the the friends at OpenAI. I don't think they they pushed it that that far. Uh so this was actually the the 2R one was u uh huge data sets millions of data points um uh and um um and then I I also said okay don't just analyze it write a huge report you know 30 40 page whatever length and then you know come up with a lot of insights about this data what questions to ask and what do we learn the mechanism it was a an iminological ical data set um sequence and genes and proteins and all that and so so that one I think 112 minutes I remember that uh uh and it came up with this 40page report uh which I I I was just unbelievable uh you know the the analysis part the previous models were able to do as well you know you know they say okay well there are these type of genes and this type of protein so it means this and that you it deres from that information, but to come up with an insight what that could mean or what would be the next question to ask. Uh that that's that's a very very high level of reasoning. Um and so uh yeah um it was it was worthwhile two hours for sure. I mean that's very exciting to hear you say that because that was kind of my I wanted to know I wanted to know is that something that is already possible and it seems like it is and so it also leads to the next question which is you know all all these scientists now really need to start understanding how to use AI in the right way right I mean this is like to help them >> I mean this that's going to happen right that's basically you know we all we all use Google now remember when Google was like new So, I mean, it's eventually going to happen, but um it's very exciting to think about how AI is going to change research and and medicine, and that's something, you know, you you mentioned and I talked I said I wanted to get back to this digital twin idea because I've heard you talk about it and it's very exciting to me. You know, we've heard for decades now that personalized medicine is coming. We're going to have personalized medicine and and yet still we just don't have it. It's just not there. Um, and I've heard you I've even heard you say something sort of interesting which perhaps I'm not saying the direct quote, but that it kind of should be medical malpractice in a way for a physician today right now to not be using AI. So, can you talk a little bit about why you said that? What it means to for a physician to use AI responsibly? um also how patients can self- advocate for themselves because that's also another area. >> Yeah. Uh in fact I I said after 01 model came out uh that I think that was sort of the first reasoning model um and I and I was testing a lot of I mean I I have a medical degree but I I don't see patients but you know I I have a lot of friends and I have some knowledge of how medicine works. So been testing lots of medical questions and some of them are hard some of them are you know sort of real time data um and uh you know before 01 uh it was great in sort of reaching to the literature you know like the physician might lack certain certain knowledge so it knows what was published uh recently and things like that but it it was not at the reasoning level so one model was able to reason And the reasoning is extremely important medicine because you know even if you have all the all the information you still have to sort of consider that person's context and uh you know uh what what would be more likely to to treat that person and we don't always know the answer uh as well or how to how to diagnose it. Um, and so I think 01 was able to get to that point and at that point I said right now it's unethical for physicians not to use AI anymore. Um, I didn't say malpractice yet, but it truly unethical in the sense that, you know, you can use it uh you can still do your judgment obviously, but it will it will prevent you missing some sort of uh an obvious mistake or you know, sometimes nonobvious uh mistakes or diagnose things that require multiple uh clinical specialtities coming together and and you don't have that capability. You live in a village or something. uh but now I think I feel that it is truly uh uh going to be considered malpractice in my opinion. Um it's not legally so but uh eventually it will be um because uh a the current models the advanced models are able to uh diagnose and and write a treatment protocol better than uh or as good as uh a specialist in that field. It's not just a you know family physician. Let's say you you know you have a very complex uh cancer uh you know you you know the mutations and and what's not and you go to a specialist like an oncologist who is very very specialized on that. Um I believe that the mo the current models are at that level. So um and of course not every specialist is is is the top specialist right? So, so you you if if that was the case, we wouldn't have millions of misdiagnosis and mistreatments in the US alone every year. I think they've said something like 12 million misdiagnosis. Um I think 700,000 people suffer from it, die from it, from from uh uh from misdiagnos. Some of them is is totally innocent. You know, any any doctor could have missed it. Um but but now AI wouldn't miss that. So uh even even a specialist might make a mistake or misdiagnose or or or mistreat because they lack certain things that that the model doesn't have. So I mean imagine that you know um uh you refuse to use u MRI machine or CT machine because you say well you know that's too much technology. I'm just going to uh you know just do an X-ray because that's enough for me. And you miss a a tumor. Uh the AI models are able to detect certain tumors like breast cancer years before a radiologist is able to to to see that. So if you miss that, I mean that person's going to die if you don't if you don't. So that to me that that becomes uh a malpractice because the technology is at that level now. It wasn't it wouldn't be malpractice you know missing a a breast cancer u you know 5 years ago because nobody could we didn't have that technology but now we have that technology so you should you should definitely use it um um and and this is going to save a lot of lives. I mean uh if you could just reduce the misdiagnosis and and and again bring every every doctor to super doctor level I think that would be a really good thing. So what you're saying is based on you know what current data that doctors have available to them whether it's an MRI whether it's an ultrasound whether it's blood biomarkers this sort of data is what is given to >> yes >> you know a model like GP GPT 5.5 pro for example >> and with that data they're able to better diagnose better to predict um to see things like you mentioned cancer >> um is that better than a radiologist can is that some is that like based on you know what kind of uh data is >> implemented these these are studies I think uh Google uh did a recent study uh in fact uh a science paper came out recently which was done with 01 preview model which is a very old model I mean the current models are probably 10 times or maybe more >> was that like the first pro almost like the first the first first sort of the reasoning model that that I early tested in 2024 September it came out and um and they found that 01 model uh did did better than average doctor in diagnosing like significantly better they did didn't miss um and so imagine the the current models how how good they are uh but but I I think um it's not just sort of diagnosing a a disease because that's That's actually a small part of the job of a of a of a doctor. It's really there's a continuum. Most diseases um you know, okay, if you if you have a flu or some bacterial infection, you know what to do. You give it and then you see an output. But a lot of disease even in that condition that may not that may not be true because you know you might have a mutant virus or bacteria. So you might have to change the treatment or might have a little bit side effect. So there's a lot of continuum there. So I think AI can be involved in all of that process. So if you can continuously feed the data okay well the patients uh we gave this treatment it's doing well uh the blood pressure is down but you know has this symptom that so what should we do change the dose of the drug or add this or remove that drug and give another antibiotic? like there's a constant um uh uh process there and that that that's not always that constant because you know people people don't go to doctor every day right so you get a prescription you you see something works and then you go back and so what if there's something that's continuously monitoring you uh post treatment for cancer it's very important because cancer is a very dynamic uh disease there's the cancer which is constantly trying to survive and mutate and counteract against the immune system. So you give it so you know you give a drug chemotherapy works and then the cancer comes back again right so why is that because mutations are accumulating so can we catch that earlier can we change those decisions can we um make sure that we give more uh multiple drugs or different drugs so before the cancer have the opportunity to come back we prevent that uh possibility. So all of these um decisions uh can be made together with uh with AI and I I I I think it's going to have tremendous tremendous impact on healthcare. >> I do want to get back to the the cancer equation in a minute but before that I just think that you know physicians not all physicians know how to use AI. They don't know which models to use. Do they use GPT 5.5 Pro or Claude or you know um how do they sort of responsibly use it which you kind of talked about a little bit but without you know outsourcing their clinical judgment. Do you have any opinions on like the different models to use? And I do know you have a collaboration with OpenAI. You've been one of the first scientists really testing these models in a biological sort of arena. But I do kind of I do think that people and physicians that are listening want to know what how what what models do they use? We definitely are talking about if we're talking about open AI, it's not it's it's got to be the pro, right? It's got to be the reasoning model, but I mean what about Claude? What about Gemini? >> Yeah. So, um I think people have this sort of a misunderstanding of they think of AI as okay, we have AI, we have internet, so let's just use the internet. we have AI, let's use the but this is advancing so rapidly. Uh the AI model that we used one month ago is not the same AI model we use now. So it's just doubling in intelligence every few months. Um you know I gave the example of one preview. Uh some people uh got stuck at the GPT4 40 model. Oh yeah, I used it and it hallucinated a lot. Even, you know, 01 uh wasn't so good. You know, it was making mistakes. That's like an ancient history. >> That's why I haven't even asked about hallucinations. >> Yeah. So I it's um I mean the advantage I have is that you know because I'm all in on AI I'm continuously testing and and so I I I can see the evolution of of these models and they get you know 90% better 95% better 97% better like it just continuously updates itself and then eventually right now with 5.5 model I I don't see any hallucinations whatsoever. I mean there might be.1%. Uh but but it's it's extremely rare. Um and so so your trust level goes up. Again it's similar to like self-driving cars, right? So we we had self-driving cars for for almost a decade maybe and they they just keep on getting better and better because their AI models are are getting updated. So my my advice would be doctors should see this not something optional like they have to uh update their knowledge medical knowledge periodically. In fact they have to have test to do that to be certified or they have to update on new drugs that are coming out. Right? So you you can't just rely on some drug that came out 5 years ago, 10 years ago. you need to know what what was approved last month and you need to update your uh your knowledge uh in a similar way even more so they have to constantly update their AI knowledge. So AI has to be part of their uh their their practice. And of course my recommendation is always use the latest top model you can use. Uh right now it's GPT 5.5. Uh in fact uh I would always use for complex problems the pro model because that thinks uh in in minutes. But at least if you're using it on a daily basis uh in a rapid fashion always use the thinking model. The thinking model is different than the instant model. Instant model is also getting better, but it it needs to reason. It needs to think. Um, and especially if you're putting in lots of patient data and analyzing that, you you definitely need the pro model. And then there are there are these uh companies like open evidence and you know uh I think most doctors are starting to use that. Open evidence basically I think applies the latest model somehow. It up updated so so the doctors don't have to worry about it. And I I think there's going to be more companies like that who will provide that service. So the doctor doesn't have to worry should I use 5.5 or putus 447 the the whatever the sort of the uh the harness model is going to pick the the best one for for medicine and and and apply it there. uh and of course hospitals should uh should implement AI uh just like you know big tech companies are you know there's enterprise level of AI that can be more secure you know protect the patient data so it should be like you know in front of the patient in hospital you see these monitors like the heartbeat and all that stuff should be an AI monitor like constantly monitoring the data and then giving information to the nurses to the doctors Okay, this is this last situation. Um, and now with AI agents, you can do that. Like I do it for my my daily life like for my email auto automatically my agents go and check my email and they tell me what's important so I don't have to go through hundreds of emails. So, oh, you know, this this is waiting for you. You have a a podcast with with Rhonda today, so you better be prepared for that. Um so uh yeah it it needs to be fully integrated uh almost like a a a a co- physician like you have the AI doctors working together with real doctors >> right have you I've noticed like some of the the companies that I've corresponded with or interacted with um it seems like they use claude a lot I mean I don't know if you've experimented with that but I'm kind of curious why um why that why that you know certain model um versus like but yeah in all fairness I've never used it. I use, you know, GP, I've been using GPT and the Pro and so every, like you said, you know, every time the hallucinations are like ancient history for me. Like I remember that was a big thing. Yeah. >> But it's going so fast and better now. But like what what what what's the difference between, you know, for example, Claude and GPT 5.5 Pro? For certain things there is no more difference because the intelligence has has peaked uh for for uh you know for for uh doing regular diagnosis not very very complex uh cases cloud is is is great. Cloud is also very very good uh like the Opus 4.7 model uh the recent model for analyzing data sets. So it's you know it can take also millions of data uh analyze it and and and and do a a great job. Um uh my preference is uh you know GPT5 right now is 5.5 pro because um it what I mentioned it has this extra insight I mean the for me um I need that extra level of insight um that's predictive um uh and >> the intuition that >> the intuition and really kind of a deep understanding uh but if I'm if I'm going to diagnose and treat a subt type of a lung on cancer. I am pretty sure, you know, Gemini 3.1 Pro or Cloud 4.7, they all do a pretty good job. Uh I I think the re some reason people prefer cloud is that it's maybe it's more pleasant to interact with. Uh you know, kind of more humanlike. I think uh GPT models are starting to get there, but still there's something about cloud that people enjoy, you know, interacting with it. It's it's really a matter of >> personable or I I think it's it used to be more more personable. Um and so it doesn't really matter. I mean, I think they're they're they're really they're all super top levels unless you're doing like a research or a very uh very complex problem. uh you know for example we did a test with uh with a colleague of mine on on skin disease with GPT5 pro model you know it was able to uh diagnose a skin disease that uh my friends couldn't really diagnose um uh just based on a a photo and and a symptom. The other models couldn't do that. they could do 90% of the the cases as well, but there's that one extra case or two extra case that is really difficult that's could go anywhere. The pro the GPT pro model was able to cross that that threshold. So those kind of cases you really need the very high level like you know uh you don't you don't go to a a a professor at Harvard for for any any reason right so it has to be very specialized disease that other doctors couldn't diagnose or something like that so that's how that's how I view it >> we just there was just news yesterday um from open AAI of uh GPT Rosalyn which I know you can't talk about much from what was publicly available, it seems as though it's going to be used in drug discovery. Um, I I'm wondering what you think in terms of like the future of aging research, biology, medicine. Are we going to be using these more specialized types of AI models or do you think more of a generalist like GPD 5.5 Pro and and the, you know, the subsequent ones that come out after it are going to be the key to unlocking, you know, medicine breakthroughs and biology breakthroughs? Um my preference would always be the generalized models because again you know going back to AGI AGI uh so if if a model is has um you know of course there there are some utilities of models that are only trained on I don't know like the EKGs or um RNA sequencing or something like that and they they'll be very very good at that like the the the best chess player AI model or um the best go player AI models but they will miss that uh connection because again I view medicine as a kind of a holistic um art in a way. Uh if you are just trying to analyze one set of data the the specialized models could be could be very very useful. In fact, you know, I gave the example of EKGs. Um, [clears throat] most generalized models were not terribly great at um, for some reason, you know, the the EKG images were were not they were not very good at diagnosing what what what it was showing. Um, [clears throat] and and you know, specialized models were very good because they were trained with, you know, millions more EKG data sets than the generalized model was. Um but but I think you know if if we can train the generalized model or fine-tune it or overtrain it I I don't know how to say it. Um then they will be better than specialized models all the time. Um because not only they have they know all about EKGs but they know about radiology, they know about RNA, they know about pro proteins. So they can take that information and and excuse me [clears throat] analyze the EKG the the electroc cardiogram your your your heart beats in the context of all the other biology. So that will that's very enriching uh knowledge. Um but [clears throat] um I think you know specialize in the sense that you can take these big models and you can sort of I don't know harness them or fine-tune them because there's a lot of data sets that's not public. So these these models they don't have access to that. You might have some um data you know locked in certain because of regulatory reasons whatever. So you can take take a big model. In fact, you don't you may not even need the the closed models. You can even take some of the open- source models which are which are now getting very good. If you uh you can train them on that uh and they also have the generalized knowledge and combined with that they they'll probably do better. >> So I wanna I want to talk about there's treating disease, there's curing disease, and then there's reversing aging. So let's let's start with curing disease, treating diseases, curing diseases because you know obviously we do die of age related diseases, cardiovascular disease being the number one killer in in most developed countries. We have cancer. That's a really big one. And and with cancer, it's just such an awful disease to have. And anyone that's listening that has either had cancer or knows someone that has had it, you know, knows this is this is true. But also, I think, you know, cancer, a lot of people think about it as one disease. Non-scientists, non, you know, physicians, they kind of think about cancer as this this one disease, right? As you and I both know, it is definitely not one disease. It's hundreds of diseases. Um, I'm curious on first of all, you know, we still don't have a cure for cancer. I mean, we've we've made a lot of progress, right? And different cancers and can be treated better than others, but can you talk a little bit about why it's been so hard to find a treatment for cancer? >> Yeah, I think uh the the important thing to clarify is that cancer is not one disease. It's probably 100 disease h 100 different diseases that have probably hundreds of sub sub diseases or sub subtypes if you like. Um in fact certain cancers are 100% curable or 95% curable. Uh you know like child some of the childhood leukemias which were completely fatal you know couple of decades ago are now you know 90% or or close to 100% curable. if you catch uh certain cancers early enough again 100% uh cure rates almost uh so so [clears throat] because it's it's a very different set of uh uh diseases um the the cancer of pancreas is very different than cancer of uh lung cancer or breast cancer or there are some cancers that are so slow like if you get um certain types of cancers If you're age 80, doctors don't even bother to treat it because by the time that will unless we cure aging first uh because by the time you die of aging, you know that that cancer is not going to kill you. Aging is going to kill you first or you know there's certain prostate cancers at certain age. So that that's why we have to really understand that this is a very complex biology. But more importantly, why cancer is such a challenge is that the the cancer cells are part of us, right? So, uh if you're infected with a bacteria or a virus, you know, it can kill you, right? They're extremely dangerous, but we are able to recognize them as an enemy, as a threat, your immune system, and we can fight back, you know, uh not always successfully, but most of the time very successfully. And we can also target them very specifically like we have an antibiotic that will only act on the bacteria. It's not going to touch your normal cells because it's only uh a foreign or organism. But cancer is not like that. So if I try to stop cancer with something I'm [clears throat] also trying I'm also stopping some other cells that are normal, right? That's why people lose their hair, their immune system is greatly weakened because the immune system has to divide. your um hair has to hair cells have to divide. So you you block them because they the cancer cell is also dividing and and you your side effects of chemotherapy sometimes worse than having the cancer like hundreds of thousands of people die because of that. So the revolution in cancer was uh recently because of what we call imunotherapy. The question was why uh can we make the immune system to recognize cancer as foreign threats like they're kind of like terrorists, right? So a terrorist you will not know if that's an enemy or not. They look like you, you know, they just come in and then they they create uh so the immune system is seeing it that way that it thinks that the breast cancer cell is not so different than a normal breast breast cell, you know, like epithelial cell, whatever. And so it doesn't know what to do. If [clears throat] we could teach the immune system or if we could remove some of the breaks that it has regulation and let it recognize and attack the cancer cells, then that could have a a tremendous effect. That was the hypothes and it it actually worked. So cancer imunotherapy I think uh is is more powerful now than than chemotherapy and radiotherapy put together. I mean they still have a have a role. Um and of course the other thing is that how can we make the treatments very specific right so if I give a chemotherapy that's not specific it's like trying to hit the patient on the head and hope that the cancer will die before the patient dies but if I know this single mutation that's happening on you know whatever EGF receptor uh in certain cancers I can develop a small molecule which will only act if there's that mutation on the EGF receptor or EG whatever and so it's not going to touch anywhere else it's only going to target the the and in fact people call them smart drugs and they're they're extremely effective right so uh if you have that particular mutation your 1% of the lung cancer patients you get treated with that drug you get almost 100% cure rate um but again you know uh we can make this even much better so for example immune system can be engineered something that we work on in the lab uh to recognize like literally engineer we take the cells out we train them we put genes into them say okay so if this gene binds to a cell assume that that's a threat and kill that that and so it's called carti therapy and they will go and seek out whatever uh the cancer cells that have that marker and kill them the advantage of that is that cancer doesn't have much weight to escape escape that. It can try to suppress the immune system, but other than that, even if it mutates, you know, the the immune system will still recognize it and find that few cells that are hiding somewhere and and destroy it and and that's showing incredible results. So, the mRNA vaccines, which I think is going to be revolutionary, is is on that basis, right? So, and that really personalized the cancer. So I have a breast cancer but my breast cancer has certain type of mutations that other patients don't have. So even if the immune system can recognize X patient, it won't recognize mine because the cancer has different mutations. If I take those mutations and synthesize what's called RNA and then give it back as a vaccine and train my immune system and tell the immune system, look, if you see these mutations in these genes, that's an enemy. Go destroy that. That's mRNA vaccine. And that becomes extraordinarily powerful because now you are directing your immune system to to an internal threat just in you. and let's say the the the cancer required different mutations, you can create another mRNA vaccine and then train the immune system to that as well. Um so [clears throat] uh you know I think that's those those are the the the the difficult parts but but we see the light at the end of the tunnel. Can cancer cancer is going to be 100% curable uh probably less than a decade. >> How is AI going to make that happen? Yeah. So, in fact, it's already making that happen. You probably heard of this story from Australia. This computer scientist um had chat GPT and and some other AI models to develop an mRNA vaccine for his dog. His dog had I think a melanoma [clears throat] and he um he got it sequenced. He took the sequence and gave it gave it to to an AI model and the AI model designed the precise mRNA molecule that needs that the dog needs dog's immune system needs to be trained got it synthesized and I think it was able to apply it in 3 months. Um probably could have been shorter if there wasn't regulations and and the tumor started to to regress and the dog is was was alive when it was supposed to to die. So I mean that that that's a very uh obvious and simple version but [clears throat] because there are hundreds of difference of cancer types you can imagine that we'll have maybe hundred different treatments for just a type of a lung cancer someone will be mRNA someone will be small molecule targeting that so to be able to develop those ondemand and or very very rapidly, we're going to need AI. So the AI is going to model every possible mutation and we'll screen millions and millions of compounds. And so we'll we'll we'll get to a point where we'll have hundreds of new drugs coming out every month maybe, you know, uh uh and we'll you know this this thousand drugs is for breast cancer patients. But you know if you have this and this this mutations and if it's stage four then you take this combination. And if it's that yeah you you you take this protocol um and um that that's that's how AI is going to of course you know if you get to digital twin that that will accelerate >> right and that's that's the next question is you know so let's let's say we have the true personalized medicine and personalized cancer treatment but you also need to know about side effects like am I going to take this mRNA vaccine and my immune system is going to go crazy and start to inflame my heart and give me myocarditis or right How do you also see this the digital twin which now has you know genomic information all your proteins metabolites and everything in real time data then it can also simulate well what's going to happen if we give this specific mRNA vaccine cancer vaccine or this small molecule to this person >> absolutely I mean you know so so you mentioned myocarditis which by the way happened during covid pandmic IC and that's why there was a lot of uh antivaccine sentiment but people uh didn't appreciate that you know co virus itself caused myocarditis. Yes, the vaccinated people young people at one in 5,000 to one in 10,000 rate got myocarditis. It wasn't it was mostly fatal. But the question should be asked like why is it that one out of 10,000 got myocarditis and the other ones didn't? Or in fact, we can reverse that question. You know, we g we vaccinated everybody, but if you were a young person, your your um chance of dying from COVID was let's say 1 in,000 or one in 10,000. So n 999 people got didn't have to be vaccinated. But to save that one person, we have to give that vaccine. Or I'll give another more uh general uh you know, we give statins to anyone who has high cholesterol. So I I think like one out of five or one out of 10 people truly benefit from that. Uh high cholesterol doesn't automatically doesn't mean you're going to get atheroscllerosis. You need to have inflammation this and that. But because we don't have the data, we cannot predict that. It's not personalized. Millions of people take statins and to save few thousand people. Yes, that's that's a good thing because you don't know. Um so AI will be able to do that. So we'll we'll tell you okay not only um we'll create the drug just for you but also we'll say okay you don't have to take this this medicine like you you should take this or maybe you don't even need any any treatment at all like you have an infectious disease or whatever or maybe certain cancers this will be enough like we give extra chemotherapy plus imunotherapy plus radiotherapy why are we doing that because we're not sure if one is going to be enough or not? Um and and so uh that will dramatically reduce the the the side effect issue. You might still have some side effect of course, but it's manage it will be manageable side effect. It's not going to kill you, for example. What about using AI to predict cancer a decade or years before it forms based on your proteins and metabolites and your biomarkers and maybe perhaps your genetics too, right? Like how do you see that? We're talking about personalized cancer treatment, but what about being able to prevent cancer before it happens, you know, years before it happens? Yeah, >> again great question because I think this is um this is so important that people don't think about very much. um we say health care you know we don't have health care we have sick care right so we we never take care of healthy people like you don't go to a doctor to say oh how healthy I am or just just go to a doctor and say can you check my immune system you know is it is it healthy am I going to am I going to get sick am I going to have cancer they won't be able to answer that question only if you get sick they will treat what the problem is um and so the preventative medicine is going to be so absolutely critical. I think not all but most diseases can be prevented. Some are just bad luck. You know, it happens no matter what you do. If even if you live the perfect life, you might still get certain disease but but a lot of them were because of your genes and so on. A lot of them can be prevented and I think AI is going to be amazing in that because it's already able to do it. uh that there was a study from um a UK bio bank um UK has this amazing bio bank with 500,000 people uh lots of data sets incredible data sets and so and this was actually done I think more than a year ago with models that were a year or two years old they took a lot of that data and they were able to predict about thousand diseases before they happened of course this was kind of retroactive. So they knew what what people were going to get based on their data that was collected years before. But the AI was telling you, okay, this patient's going to have this disease that, but not patient, normal healthy people, they're going to get this and that. So that to me that was that was amazing and that's going to get better and better because there are there are there are always signs like cancer doesn't just develop in days. It takes years. uh if we probably most of us might have some cancer cells you know most of it controlled by immune system and so on and it you know slowly grows it has to have another mutation another mutation but but there's probably some signs of that somewhere you know whether it's in your metabolism or this you know and a AI even if it's 100% will be able to say okay look um I think that you know if this is if this is the lifestyle that you continue your chances of getting this disease is now is 85% or whatever. Like I wear a glucose monitor. Um I'm not diabetic, you know, uh but I I want to see every minute or every five minutes what my sugar levels are in a continuum or if I eat something, you know, is it spiking? Is it coming down? Because I want to prevent insulin resistance. one of the the worst things that can happen to you. If I uh if I don't uh do that, I won't know until I get diabetes. My my insulin if if if if my sugar is constantly spiking and then you know insulin is just working too hard and hard that that could continue for years. By the way, um that at some point it's going to break, right? Um for some people it might continue 50 years, nothing happens. Some it might be five years, but that data set probably has that predictive value that plus my my age, my genes, but whatever. So, uh yeah, that I think um everyone's going to have their own um AI, I don't know how what to call it, uh health coach or something. Uh but but it will it will continuously analyze the data. Um and and hopefully we'll it will be much easier to collect data because that that's another issue you know we don't collect data like we know nothing about uh you know there there are more than thousand metabolites in our bloodstream so we look at maybe you know 10 of them 20 of them only if we get sick not even for a checkup so we have to have a continuous um like a glucose monitor I want to see what my you know uh proteins are changing hormones are changing you know in reasonably continuous manner. >> Such a good point and I'm so glad you brought up the UK bio bank study. I remember I think the model was like called Milton or something and it was it's a Astroenica own like developed it or something and and I remember looking at this study because like you mentioned the bioank data is huge data set and it just spanning many decades and so I think they looked at you know like over 200 plasma proteins you're talking about 10 we're talking about 200 right? Oh yeah. >> And and all the other data, right? And they were able to predict and I think cancer and neurogenerative disease were at the top of like 10 years before and they were able to look at the people. So the AI AI predicted it based on based on all this biometric data. And then they looked and said, "Oh, yep. Those people actually did end up getting cancer and Alzheimer's disease." And it was very accurate. Yes. >> And and to me, the exciting thing here is that you can intervene before it happens. You can make lifestyle changes, you can make dietary changes. I mean, these things matter. They do matter. And and that is exciting. Uh because then you don't even have to get to the drug part, which you know, maybe [clears throat] you will, but if you can make these changes, if you know, hey, I'm on this trajectory to get cancer. I have all this inflammation. I have all these things happening. If I don't make a change now, then in 10 years, I might have a cancer. >> It's very motivating, you know, for for someone. So, it's very exciting um as well and and and then having AI in there is just going to make it even even better. Um and then I [clears throat] I want to get in I want to get into age reversal and and before we get to that, you know, you you've really been a pioneer in this the field of AI being involved in biology. You know, you were talking to me about your your blog I don't know was it 30 years >> bios singularity. Yeah. 25 years ago. 20 25 years ago. >> Yeah. So you have this blog bios singularity predicting. Can you can you talk a little bit about it? >> Yeah, sure. Uh so uh in fact I I gotten interested in AI early 90s uh after I graduated medical school. Um you I was very interested in computers uh when I was a teenager. The the first computers had come out at the time and you know I was trying to code and uh you know just I mean I loved it. It was it was just so wonderful. Um, but you know, I went to medicine because I figured biology is much more complex, so I should first try to figure that out. But then immediately I realized, and I'm sure you did too, you're a scientist as well. Um, that biology is so incredibly complex. I said, well, I mean, you know, we don't have any any chance of figuring this out, you know, because there's going to be so many so many data sets. So that's when I first got interested in uh in AI. Uh, of course at the time, you know, AI was was very primitive. Um, uh, but fast forward uh, you know, one of the one of the books that influenced me was, uh, from, uh, um, Ray Kurtzwell. I'm sure a lot of people follow technology know him. Um, he wrote this book techn singularity is near. So he he called a point of singularity where uh the the computation or technology is advances exponentially so much that you cannot even predict what will happen next day. I mean because it's sort of like a self-training AI models and he he had these figures where he would plot the advances of AI you know say you know by 2029 it will be at the human brain level and you know we'll reach AGI and you know it was just unbelievable and most people thought that he was just talking crap or you know science fiction you know they didn't believe it how could that happen and so on but you know I got I got very excited In fact, I have a signed copy from Ray uh for for the book. And so being inspired from that, I I started this um blog called bios singularity. So I said, okay, so so computation is going exponential, but uh biology is sort of a computation as well. I mean it's it's based on information and so uh but it's just much more complex. So it should also expand exponentially. And if you if you plot that curve that that means that by you know based on my calculations 25 years ago in fact I wrote it on the on the about page of the blog by by year 2035 or so we should be able to treat all diseases uh and by 2045 or so that we should be able to completely reverse aging. In fact by 2050s we will get to a point what I call human 2.0 you know because at that point we have a complete understanding of biology. Uh then we can truly engineer it. We can create new biological organisms. We can you know change our biology, our genome, reprogram it. Um >> rewrite our immune system. Yeah. Exactly. um in in in many possible ways because it's kind of a messed up if you if you think about it like you know biology we think is a miracle but it's a it's a bad kind of a a legacy engineering right it's not it's not a bad engineering it's a legacy because biological system finds something it can't get rid of it can't start from clean slate so it builds on top of it so you get regulation over regulation over regulation and then of course you know with like immune system that I study you know you get lots of autoimmune disease is his immune system kills a lot of people you know even during like pandemics and things like that or it doesn't recognize the cancer cell and things like that. So why you know we should be able to design like immune system 2.0 like clean slate really greatly engineered the immune system. Well, and I said you know by 2045 50 we'll get to that point. Um and and actually, you know, again, at the time it sounded really crazy to people. Uh but now I feel that I was I was too conservative. We'll probably get there. Uh but but the key point is that I wrote specifically in the about we will do this because of artificial intelligence. You know, I was just taking the plot that Rey uh plotted. You know, I said, okay, by 2029, AI is going to be at that point. it will be good enough to apply to the biology and that will allow us to solve diseases and then and then the aging the fact that you know the timing was was pretty good uh uh again even even a bit conservative uh uh I feel great about it that's why you know I I'm all in on AI like wow um that it's happening it's really happening >> so aging is very complex and you know as you know it's not one process. We've got these 12 hallmarks of biology. We now have 12. Genomic instability, mitochondrial dysfunction, you know, cellular scinessence, on and on. We've got there's 12 of them. >> And we know organs are are aging at different rates. They're they reach their peak at different rates and they age at different rates and everything is interacting in a very complex way. [gasps] What do you see as the bottleneck for understanding the aging process and also reversing it? >> Um I mean more so than the bottleneck uh this is the way we have to think of aging. Uh biology actually um uh is programmed to prevent aging. Right? So, it's not like um uh it's not like a car in a way because once you make a car um you have to constantly bring it to a repair shop or you have to repaint it. Biology does that internally. If it didn't, we would age immediately. Like there is a disease called progeria. These children get aged uh by the age of 78 they become like a 89 year old because of single point mutation in one of their one of their genes because they lose their ability to repair um whether it's the DNA repair whether it's getting reg rid of the old cells or cleaning up the tissues and then regenerating like stem cells creating new cells. So this program continues for for sometimes decades otherwise we we wouldn't survive for some animals for some organisms is only a couple of years for for us is about you know maybe 50undred years uh for some veils is hundreds of years so so you know same biology it's just that one of them decided that you know I can keep a veil um or you know whatever some animals um you know older longer because they don't they're not getting hunted or they can reproduce later and so on. So what happens in in in the in the biological system is that somehow uh this program breaks down and you start to lose what's called the res resilience, right? So when you are uh age 30 or 40, you're a you're resilient. you can tolerate much more damage than someone who's 70 years old, 80 years old because your your your systems are uh you know even if you get wounded or if you uh get sick you can recover uh easier. Um uh but that that sort that resilience is lost and that the reason why it's low that there is a sort of an information loss because the biological system has a certain information that it knows when certain genes should be turned on when things should be regenerated when it needs to be like your skin. You know why you get wrinkles? because your cells stop making collagen and then all kinds of crap accumulates under your skin and then you know the guy the guys who like the macrofasages or whatever was supposed to clean there they don't do their job. There's some sort of a breakdown in information or communication or you know intracellular communication is one of the hallmarks of of aging. And then of course why that happens is is that 12 hallmarks is is is the reason many reasons you know uh for example the bacteria in your gut is is is a reason. So so these bacteria produce all kinds of metabolites that help your immune system to constantly regenerate keep it in optimal shape. If that changes then you know your metabolism is changing your glucose levels your mitochondrial uh mutations and so on so forth. So all of these things accumulate you know epigenetic changes and DNA mut mutations and somehow the the biology forgets well what am I supposed to do like how how am I dealing with that also becau because when a damage happens it's harder to fix a damage than prevent it right so if if you're continuously taking care of your car or your house the likelihood of it you know breaking down is much less than If you wait until like okay nothing works yes you can reverse it but it's going to take a lot more effort and so I think uh what will happen is that for a younger uh individuals in the next decade or so uh there uh for them it's not just it's not going to be reversal it's going to be prevention of the aging process it's going to be maintaining that process the resilience decades more. So we will come to a point where if you are 20 30 whatever years old you won't age anymore because it's going to be constant reversal. But people who have already aged you let's say you're 80 years old 90 years old then we're going to have to reverse that process. That's that's a more difficult we'll be able to do it. Definitely we'll be able to do it. Um uh uh but um uh it will require a lots of engineering approaches because you need to fix most of those hallmarks. If you're younger, you prevent those hallmarks from happening. You maintain the the information uh uh much much longer. Um both of those uh will will will happen. um uh uh we we we just need to figure out what that information is being lost and we put it back. >> Do you think so? Let's first talk about preventing the aging if you're a younger person because it's easier to to do always prevent if if you have a person you know who's 20 or 30 years old. Do you think that the approach would be finding first of all do we even know all the repair processes that are we we have discover we have what we know right >> but we still have a lot to discover >> we have a we probably have a lot to discover and so like do you think there's going to be a a discovery where we figure out like you know we know things like autophagy stem cell depletion you know all these stress response genes like antioxidant like all these things DNA repair mitochondrial the way mitochondrial repair itself, right? Um, are we going to be enhancing or like tuning these up so that they keep working at their prime continually or do you think we're going to have again this like information where we why why are those things going down? Are we going to just then go to the information of it, the epigenetics perhaps? Um, and is it going to be more targeted towards those genes or are we going to have more of this? you know, we'll get into this cellular reprogramming and partial reprogramming, but um [snorts] I'm I'm curious like how you see AI coming into that process. Like I guess we don't know that's the part of the problem, but then we have to figure out how to give these del you know treatments to people, right? That's another part of the equation. Um, so I mean I think you know the the ones that you mentioned about sort of the lifestyle changes and they of course help a lot but they only slow down the aging process. There's I don't think there's anything that reverses that process. There might be some sort of local reversal for a temporary period of time. Uh maybe but it's still kind of trying to you know uh hope that things won't go bad a little bit longer. Like for example, some people can live to to to 100, others only to 60, right? So there's something good about those who live to and in fact there are super centinarians who couldn't make it to 110 years old. Very very few people, but I think it's mostly genetics. I mean their lifestyle might have helped a little bit. Uh something about their biology is able to maintain that information much uh much longer that program. So we have to get to the core. what what are the things that are disrupting that information um uh loss? Um and um yeah, it's of course you you have to focus on the on the genome because that's that's sort of the blueprint. It's not just that. It's sort of what affects you afterwards, you know, that your your microbiome, your um metabolites, you know, how those things are changing, whether accelerating or uh reversing, you know, like and it has to be kind of an engineering approach as well, like you know, the skin aging is is a very different problem than immune aging, than the brain aging, right? So, uh your skin cells are constantly renewing. So all you have to do is to have sort of the uh programmed stem cells to go in there clean the environment sen cells and get it get it regenerated and produce collagen whatnot but the brain is not like that right so you don't you don't want to regenerate your your neurons uh you will lose your identity so they have to be dealt in a different different way some of it will be I think for the younger population u it seems like you know redesigning certain biology ology would be sounds radical but it would be uh more foolproof right so what if we could change the genome through genetic engineering like we add certain genes or we change certain genes such that the DNA damage um is checked you know much much longer it you know because there are in fact certain animals who have better DNA damage proteins they kind of evolve to do that like elephants rarely get cancer, right? Because they have this gene called P-53. They have multiple copies of that. P3 is kind of like the guardian of the genome. You know, it prevents the genome from getting too much mutations and prevents cancer. So, somehow elephants have pre I don't know how many copies, but they get very rarely cancer. Um, naked mole rats, you probably know that very well. Um, you know, they they're they're like rats. They live underground but normal rats live a couple of years and these guys live 30 40 years. So it turns out they have some mutation in some immune gene called seag gas that's also involved in immune optimization and DNA repair just like you know one or two genes make a huge difference. So can we uh engineer humans to uh block that degradation of of information uh uh for for those who have already had the damage then we're going to have to to think about repairing that reversing it and then maintaining it. Uh that's that's going to be a bit more challenging but uh uh we'll we'll get to that too. [snorts] >> What do you think about so the gene going to gene therapy? There's obviously gene editing, gene therapy, and and um right now we only know what we know, right? Again, like with these longevity genes we know about, but do you think that that AI is going to be able to help us analyze the human genome? And I don't know what other data sets it will need but we'll give it everything and help us figure out well actually there's interaction of these genes together and when there you know like all these combinations is that something that you think is going to happen? We'll actually figure out there's a lot more to this equation than we originally knew. >> Yeah. That that's the critical problem because we know what all the genes are in the genome. like we have we have it decoded completely and then we pretty much know their functions most of them uh even if you don't know every single gene involved in aging we know a lot of them the problem is that different gene uh uh first of all can create different proteins you know there's all that splicing that happens and and so on but but even we doubt that in a different context so if you the same protein uh can kill a cell or causes survival like in immune system we have these receptors called TNF receptors or whatever they can they can have a survival signal or or a death signal suicide signal depending on the context of the of the cell. So that is very very uh critical that how as you pointed out how these genes and proteins uh in a network fashion in a sort of a topological network uh you know what do they do like if I interfere like these uh um probably we'll talk about that these things called Yamanaka factors where you can you can generate a stem cell from a normal cell right so like complete regeneration uh uh but But the the problem is that they can also cause cancer because they only need to be active in certain time. If they're active all the time, they can cause terteratomomas and things like that. So that that part is so complex that we absolutely going to need AI to simulate that for us. If I have this gene in the context of all the other things at certain age with these epigenetic programs plus all the metabolites and so on because those are constantly signaling the cell and you know doing letting the the proteins do something and so on. What would happen if I interfere with that particular gene or how can I improve that? Uh if if you have a because you have to consider the other genome too like your gene therapy might be very different than somebody else's because you might have some great genes that are synergistic with that other person might have not so great genes if even if you try to improve it that that would actually work or it wouldn't it wouldn't help. Um so uh it's just a matter of complexity. there's so much information that uh the AI has to not only put that together but have sort of almost a temporal simulation of the model like that's a very important point the because right now the models are kind of static they they have a good understanding but they don't know what would happen if a cell comes next to a tumor just 2 minutes earlier they the cell next to it what that context affects there's a behavioral issue. It's the same problem with the robotics, right? So, um kind of the physical intelligence or the biological intelligence once those models are evolved with with a lot of data. I think we'll we will be able to simulate this and and AI will will be able to decide this is the gene therapy you should get. So, you need a new copy of immune system, but let me design it for you. It's it's so exciting because not only are we talking about, you know, extending our lifespan and curing disease, but we're talking about like getting rid of side effects in a way. I mean, you know, people all respond to different foods and treatments and everything differently, right? That's why some people have a terrible response to perhaps maybe a vaccine >> um and others don't. And so, it's really exciting to think about that. Um, >> which which I, by the way, call human 2.0. And maybe we'll get to human 3 3.0 uh which which will happen at this bios singularity moment. What that means is that you know we we kind of re-engineer ourselves. Um uh I always think about like most most scientists or most doctors think like what's wrong with this person or patient. Uh I always think the opposite. There are certain people I'm saying what's right about them? like this person is has smoked for 50 years, never got a lung cancer or you know had a terrible diet or whatever. This one lived to be 110 for you know whatever reason. And so what is good about those people? Why can't we take what's good about all of those people and then re-engineer those that are not so lucky to be born with what's what's so good and then you know even make it better. So that's the human 2.0, >> right? I I I mean that's exciting to me as well, right? I mean we do know like you said we can live humans are capable right now of living to be is the whole I think the oldest was like 121 maybe >> 123 French woman. I mean >> the fact that that right now in 2026 we know that humans can at least live to be 123 >> is exciting. 115 I mean at 115 116 that's considered sort of the current limit but you know only 300 people in the world are 110 and older why is that why not the rest of the 8 billion >> right yeah it's it's fascinating and I'm I'm so excited for you know having this super computing power to help us figure that out what did you think when you know the Yamanaka factors were discovered by Shina Yamanaka and all of a sudden you could take this old cell and completely rever reverse it to revert it to an you know essentially induced you know pur potent stem cell. Do you remember like is that was that something did aging come into your mind at that point where you were thinking well that's the youngest almost you could get I mean >> yeah uh of course uh in fact at the time I was um part of some aging groups uh uh I think like an hour after the paper was published I was you know typing there you know like this is this is it this is amazing so I I should say that there were two um moments for me uh uh that that I thought that aging was was going to be uh reversible or curable, however you call it. Um kind of like the chachipit moment of biology. The first moment was uh the um the sheep uh that's called Dolly. Uh you probably know it was the first cloned ship sheep. Um it was 19967 or something like that. I can't remember the exact date, but it was in '90s. And um so basically uh the um uh the scientists took a cell from uh you know uh from one sheep and then recreate exact copy of that sheep you know by by cloning it. Uh it was it was at the embryo level but it was sort of like exact copy of it. So that means that there was enough information that you could just like uh recreate the same person again and again and again. Right? And then the second of course uh uh the the Yamanaka factors uh in 2016 I think. Um and that was the moment that uh that we knew um that we could completely erase the um sort of the age of the cell on a cellular level and then bring it back to a purotinent stem cell level and then use that to recreate the whole biological organism. So, so it means that we have unlimited supply of regenerative capacity like it's there is there's there's no limit to it. In fact, we already know that like so our our DNA just keeps for billions of years. It keeps going on and the fact that you could do that in the lab and you could you could generate it was was was amazing. Um uh but of course the the problem was okay so then how do you apply that? In fact, I I think there was just a recent study that started in Japan using the Yamanakica factors uh uh in in clinical trials because you know it was not a very controlled system like you didn't know if those cells would develop tumors you know in mice they they did some of them tumor tumors you know whether um you can control them or importantly I think there's going to be a trial started by David Sinclair soon can we do like partial reprogramming because most of the time you don't want the plur potent cell all right you just want your skin cells to go early enough to their sort of more stem-like level like I work in immune system and for us um I can divide like immune cells into naive memory and aector and differentiated so the naive cells are kind of the young guys they have huge potential to expand and and make memory and and affect the population and the other ones constantly um die and get older. Can we actually revert the cells towards the naive? And I I actually spent a long time trying to do that. Um so maybe this partial programming will will will will enable that and and that's that will be uh revolutionary because uh then you can if you can also deliver those then you can make most of your old skin cells turn into a younger version. I think the trial is going to be for I uh with David Sinclair. Um yeah. So uh but again it's it's um the these these things showed us that uh we can reverse aging. But when people say oh that's impossible like this is this you can't you can't reverse aging like you know this entropy whatever. Um but we we we do it in the lab all the time. Uh why not do it in a in a total organism level? So with this partial cellular reprogramming as um as you mentioned you know you're you're basically taking an old cell and putting these four different proteins I think they can do it with fewer now but putting them on for a shorter period of time on the cell and that it's changing the the epigenetic program and in a way that it's still the cell keeps its identity. It doesn't become a stem cell but it seems to be more youthful. Um, I know there's been some work and I haven't followed all this literature since I the first, you know, some of the first studies that came out, but I think it was like Juan Carlos, um, Epizusa, he he's now, I think, at Altos Labs, but he at the time was at the Sulk Institute >> and, um, he had done this in in mice. I think they were even maybe perhaps progeria mice or some sort of accelerated aging model >> and there was some reversal of you know certain organs seemed to be rejuvenated in a sense and um the the life expectancy was extended in those animals but what's interesting is that not all of the 12 hallmarks of aging go away. >> Yeah. >> Right. And so you would hope that you would reverse aging totally but there's genomic you know somatic mutations are still there I think tieumir don't get mitochondria so >> do you think first of all I don't I I'd love to understand why that is so what is it if you're if you're essentially you know wiping out the epigenetic current epigenetic program and and reverting it back um why does not everything change I don't know if you have any ideas But do you think AI is going to help us understand that? >> Uh definitely. I mean we I should also point out that we um we do need to generate lots of data. So so I think um you know when whenever I talk about AI um people say okay well why can't AI do it now? Um for two reasons. One is that we don't have enough data. So we we we probably know maybe 10 20% of all the biology. we still have lots of data to to generate. The second is the >> you're talking about scientists. >> Yeah. Scientists or or automated lab whatever it is. Um so I mean right now we're able to generate millions of data points in one experiment you know and but but even that's not enough like we need to generate billions of data points and so on. So but of course to handle that we also need um super intelligence and supercompute. So, we have to have compute that's thousands of times than what's available. And people say, okay, well, you know, why are they building all these data centers? Isn't this enough? And so on. Well, we're going to need it. If you want if you want to cure all diseases and reverse aging, we're going to need probably we're going to need data centers in the space and and and and lot more because so much data has to be in real time sort of uh uh simulated. um uh and we might get much more efficient doing that as we learned algorithms. So so that's that's one issue. The other is that um as you pointed out something very important I mean this partial reprogramming or total reprogramming they're super exciting but they don't solve um they don't completely solve the the aging problem. They will um make your um eyes see better for a certain period if you're 80 years old or your skin gets better. Um but will it work on your um your heart muscle uh or on your brain cells neurons which is the critical point because if you can have a perfect body but if your brain is aging then then that's it. Um so will it modify the sort of the microbiome that has now the environment of an old person because if if that happens if your metabolism is in old person's metabolism and and microbiome is old person's metabolism and your your DNA has accumulated a bunch of mutations and mitochondria has bor mutations you can reverse that a bit have some regenerative capacity but they will quickly become old again, right? You know, because the environment is not is not great, right? So, like if you live in a bad neighborhood and you created this beautiful house, you know, it's but it's very bad neighborhood, your house is not going to last very long there. So, your your neighbors has to be clean as well. So, I think it's it's a great thing and that's probably going to add certain uh years to lifespan and the quality of life uh for sure. uh but we we have to push that much much further um and then really understand whether it's 12 hallmarks actually I asked JPT recently came up with another four or five hallmarks >> what were they >> I I can't remember exactly uh it was one of them was related to immune system I just this was recently um uh but yeah it was it was it was quite interesting um trying to remember one had to do with metabolism um uh uh you know because we we we kind of classify hallmarks based on what we can measure and see and I think AI can see a little bit more than we can. So anyway um this is going to be uh a serious engineering uh problem. I I would be very surprised if we have like one pill you take and then you suddenly become young again. That's that seems very unrealistic to me. >> Yeah. I mean [laughter] you know and then the other question is like in the lab we're we're the way we're delivering these treatments is like an adino virus right and and then it's like well is that going to cause cancer because they virus go to right cell >> is it going to go to the right cell exactly I mean the there's definitely a lot of engineering >> we we have to develop uh so one of the things that I like doing with AI models is to develop some new methods new new technologies they have a bit too much guard rail so uh they don't allow me to to go too deep in it but you know because uh I don't think we have we have enough tools like of course we have crisper now but actually Dudana's lab just came out with something even better for bacteria for genome editing so imagine there's there's there's probably all kinds of other tools that we can build that will make this localization the editing much more perfect and and has to be programmable you have to literally create circuits we can program immune cells in in in culture like we can give a a drug it will shut down their response or we can create end or gates and not gates if they see two molecules then they respond if they see one they don't like you can literally program the biology so we have to develop these new tools that are better than viruses maybe uh generate lots of data sets um and they manipulate the the organs and so on could be that for some organs when they're too old it might might be just too difficult to repair them. So you might consider just putting a new one, >> you know, like it might be a point of no return, your your kidneys or whatever. Then you'll have these u u organ factories which 3D printed and actually >> uh we we did a lot of collaboration with a colleague of mine, you know, he can print, you know, small tissues, lungs and and pieces like that. So some of them will be kind of transplanting new organs. Some of them will be pre-engineering and >> and then the digital twin the analysis and simulation will be able to figure out is are you going to reject this or would you need to not reject it? >> That's right. >> Right. Um what do you think of the new data that came out using this this model called GPT micro 4B GPT micro 4B? um where I guess there's this model that was used to figure out how to make certain mutations in the four different Yamanaka factors to make them more effective. So they were able to basically 50fold more um be more effective or efficient at increasing this induced pur potency. >> Yeah. >> How do you interpret that data? >> So I I don't think that model uh is is any better than what we have right now. uh probably um the current models are are much better. Um the I think probably there might have been two two differences and I don't know all the details but one is that they probably remove the the guard rails because there's a lot of biocurity guard rails in in the current models. Um if you ask the same question to GPT5.5 it will refuse to do it. It will say oh this is a biohazard like what if you mutate and create a new virus or new cancer whatever. So that might be one reason and then the other is like if you let these models think longer. So like GPT5.5 pro and and the thinking and in model is the same pre-training but pro model can think two hours thinking can take two minutes. Uh so the longer they can think the the more they can iterate. They can run these scenarios again and again and again. So my speculation is that that model probably ran for for a long period of time. Of course you need a lot of compute and a lot of tokens not a problem for open AAI. um uh then you you will probably come up with the solution that even a a more intelligent model couldn't come up in a in a shorter period of time because that that particular case is really running experimental scenarios like okay if I do this mutation what would be the potential outcome like it's running all the simulation oh yeah okay so so what if I change that mutation to here and then what if I add another mutation and running the experiment again and again again so you you're constantly making the the solution better and better and better as as you think longer. Um so uh and and this will get better. So if if you have much more compute, much more intelligence and you say okay um GPT7 or six whatever is go and think for a month, you know, find the perfect molecule that will bind to this receptor and this will cause that. It'll it'll probably figure that out. What is it? It sounds like we're going to need to do a lot of this type of simulation and by were I mean researchers and scientists. What is it going to take to remove some of those guardrails in that environment for researchers to be able to to make these new discoveries and and what sort of I guess I mean how how do we protect from a a new crazy >> biohazard or you know biosafety issue? Well, I mean I think like OpenAI is partnering uh with with um you know trusted people. Uh so you have to be approved by them. So I think then uh whether it's a company or something like that. It's the same problem with uh with cyber security, right? So Anthropic has this new model called mitos and they decided not to release it because they said it's too dangerous for cyber security because this this model can just crack into any can find all these things that that other uh others cannot see. So, in fact, even the governments thought that that was important that they should uh I don't know if they're exaggerating if it's if it's if it's true or not, but so you have to put that guard rail if you release it to the world because somebody can use that model and then hack into your bank account or somebody can use it to create a a a new virus gene or something like that. So I think there you know that will be made individual purses or institution basis that these these um uh hopefully these AI companies will share that because they might decide not to share it. Uh might say well okay why don't we just develop all the drugs internally and not release any of these models. um uh some some might be doing that for example I don't think that would be a good thing because uh what you really need is again as I said you need a lot of data you need a lot of scientists uh putting all that data into the models but not only the data but their experience in a way you in in the let's call it the wild or or the world you're you're actually training those models even even if it's super intelligence it's going to so hungry for data that you're going to have to um collaborate or release it to to to others. Um also I think this will be important to democratize uh health care because one question everybody asked okay well you know if you find the the treatment for aging this is only going to be available for the super rich I'm never going to be able to afford it or or treatment for cancer I say the opposite actually thanks to AI it will be super affordable because if you can create a drug like in a startup let's say cannot compete with a big pharmace to a company they can find a drug uh using AI 100 times cheaper and if you can do the clinical trial using digital twin that's where all the money goes like you could develop a drug for a couple of million dollars rather than a couple of billion dollars so the cost of drug development or treatment uh development will be magnitudes lower and that will give uh a huge number of people u access to that uh but of course you know AI AI has to be um shared. It's it's uh I think it's it's a product of all humanity and it should be the possession of all humanity. That's how I view it >> except for going back to the the thing that you mentioned at the beginning of this podcast which is that you know humans in the wrong hands that is the problem and that's and that's that is something that needs to be very taken very seriously. >> But but the the solution to that is also AI. So right now, I mean, I hear that like MTOS um basically finds all these loopholes in in this cyber security uh issues that people couldn't figure out for decades. They didn't even know they existed. So it's just patching all those uh uh all these security bugs. So it will create almost a perfect secure systems like it will be unhackable because uh MTOS is actually preventing that. So to to prevent that from happening you still need AI. You might still have some bad actor trying to develop a virus that will cause a pandemic. To prevent that you also need AI. So the AI should be able to predict it and already create the vaccine ready. will say, well, somebody might make this virus, so let's let's get ready for it. Um, so, uh, AI is the solution to all that. >> Interesting perspective. Always seems to you always seem to have a positive outlook. Um, I wanted to ask you another question about, you know, we're talking about these simulations and how we're going to, you know, using AI to to essentially run these clinical trials cheaper because we're going to do this, you know, these simulations and have, you know, biomarker data and it'll just be, you know, shorter and and cheaper and easier. The question is always what do you measure, right? What is the biioarker? What are what's the end point, right? And in aging, you can now see I mean every a study almost a new study every day coming out looking at these epigenetic aging clocks. And that's you know the the so as most people listening to this podcast know I've had Steve Horbath on um a couple of times and he's sort of the pioneer in these epigenetic aging clocks and they've now developed over you know the last decade or so and become much more of a biological marker of age like your biological age not just to be able to predict your actual chronological age. And [snorts] so, um, you'll find now studies that are looking at treatments and whether or not it can reverse quote unquote reverse biological aging or epigenetic aging, but it's not clear that that's necessarily, you know, if that's really reversing aging, right? So, how what do what do you think um from your perspective, what should we be looking at in terms of some of these functional >> outputs? Um yeah, I mean the those epigenetic u markers are very useful um but I don't believe that um they are terribly useful as um sort of as as predicting true aging. I mean there's there's a very um uh significant problem with with those markers. uh usually they're they're done through through blood analysis [clears throat] but in the blood you have u like you know I work with te- cells so you have these cells that we call aector cells that have um lots of epigenetic change because they differentiate it and they continue to accumulate in in old age and then you have these naive cells that have you know more pristine uh kind so it's a combination so depending on um what that combination is is going to affect the output of of the um so you you you can actually just look at the proportion of your uh T- cell differentiate T cells you'll probably get the same same kind of information um and it doesn't tell you like what's happening in the skin or the brain or the heart you know that it doesn't mean that uh if if the immune cells are getting younger or the young ones are expanding and the old ones are dying that doesn't mean that your skin is getting younger or your liver is getting younger. So that it has a very limited use in my opinion. But we really don't need that because like aging is probably the easiest way to measure. We know exactly what goes wrong in in in old age, right? So like you can't breathe that well. Your heart doesn't work that well. Your muscles don't work. you can only, you know, raise so much because it your weakened muscles or your VO max is is is lower. Um, these are all phenotypic like you don't even have to probably withdraw a blood just measuring the ability of uh of an elderly person. Uh, can they walk uh better, you know, 100 meters than they used to? like because that's looking at the total biology like you know your cells your metabolism or whatever uh muscle to me that's or or your cognitive abilities >> but those can't be simulated I mean >> the they eventually they can be right now they can't they can't be simulated uh um because as I mentioned the AI is missing that behavioral physical intelligence in the real world because that's a that's most things are happening in real life. uh but um I think I think they can be simulated but more importantly uh I think eventually you have to tr whatever the AI comes out with you need to try it on on the humans right so uh my point is that you don't have to uh do anything too fancy or wait decades to see the effect if I give this treatment to um I don't know 80 year old and they're suddenly able to breathe Well, you know, their VMX went up. Um, they're sharper, they can think better, uh, they can remember better. Um, you you can look at their immune system and we can see that the cells are we know which cells are younger or worse. Or you can look at their skin like, oh wow, the skin is getting young. Like you see it, you don't even have to do anything. Um so so there are so many features phenotypic features of aging that could be um objectively measured actually and not just subjectively you will see the effect very very quickly like this partial reprogramming trial they're doing it's it's done for glaucoma patients I I guess uh because that happens in old age right so your your cells are aging so I mean if these people start to see it works right their their cells cells got regenerated. Um you don't need to look at the epigenetic. Um so I think uh it will be a combination of those um measurements probably we will come up with and AI will probably come up with this set of biomarkers. I don't think we know because it's going to be a set of biomarkers like um you know your glucose your cholesterol might be high when you're 30 and it will be high or low when you're 80. I mean there's not a very specific marker that will tell you your your age for just looking at that. But the combinatorial effect will will AI probably will be able to predict your age looking at all kinds of data sets and say oh this guy must be uh you know um 52 years old based on this. You know >> I know we have uh that model clock base that's looking now at a variety of small molecules that might reverse epigenetic aging. Now there are some data sets showing that you if you reverse epigenetic aging there is some functional correlation with some functional improvements like pre-frailty things like that you know like improve but um it at the end of the day you know I think it it'll be interesting to see if there's going to be companies that come out trying to sell some sort of drug to claiming it reverses aging when they're really just looking at one >> biioarker which is reversing >> it's most as I say it's mostly the immune aging that they're looking at or or sort of maybe getting rid of the terminally differentiated immune cells like for example in old age you you accumulate these CME specific tea cells um CMV is a virus that you can't really get rid of so the immune system constantly have to keep it under check um and those immune cells they kind of become like missionaries they should retire but they keep on expanding and some indiv individuals might have like 20 30% of all their tea cells just dedicated to like one peptide of this this CMV and they're they're not helpful but they become uh harmful because those guys are are old they should retire they don't and they cause inflammation because they're they're active and um and they don't give place for the young guys to come in uh and they're they are epigenetically you know closed because um they're differentiated their telomeres are shorter So, uh, you know, you might be getting rid of some of those cells with certain treatments, which is great. Um, but then you have the indirect effects, right? So, if you can if you can control the immune system and inflammation, that's going to have huge effect all over your that doesn't mean your skin got just regenerated, but it it will it will help clean up >> aging. Yeah. Yeah. Exactly. Um, also the other thing I was thinking about is like, you know, we you're mentioning V2 max and, you know, muscle strength, muscle mass. We have all these markers that sort of like decrease with age and yet we don't know necessarily that they cause aging in a way. So the question is like will AI be able to take all this correlational data like we have all this you know all these different functional out you know endpoints that we look at and and be able to differentiate it from like personalized you know this personalized um data set versus like actually like how do you cure aging like what do you change that's going to drive you know [clears throat] reverse the aging I mean there's there's a lot of questions Um, you mentioned something interesting that had to do with the brain and that is something that I've been thinking about as well because you know we we have a lot of repair processes in our body right we can repair a lot of DNA damage and you know mitochondrial function and you know all these things but in the brain we can grow new cells replace the old cells in the brain it's not as robust right there's some parts of the brain that can um you can grow new neurons neurogenesis there's neurop plasticity. That's a big part of of the repair process in a way, but it's not like a big you you're not you're not totally replacing the brain and you don't want to, you know, as you mentioned because then memories go away and your identity and you know, it gets very complicated. >> Um, how do you see AI >> intervening in that? Like everything's great if we can reverse our heart aging and all this, but our brains that's so important >> now. Is it just going to be a, you know, delay age related disease, neuroinflammation, all that stuff? We can we can fix that, but like are we going to be able to really reverse brain aging? >> Um, you know, I I would have to ask AI to to to figure that out. But, you know, I can I can think of several scenarios how that might happen. uh first of all you know neurons um or the brain overall must have some very good maintenance policy right so so there are neurons that live for decades maybe 70 80 years and not just neurons but there are other cell types that can live for very long they don't divide very much there is some regeneration uh it's not like zero and that's very important because that means that if you let's just do a total experiment let's just say that you replace 0.01% of your neurons uh every month or every year something like that. I don't think that's going to make a huge difference in your brain structure because what they're doing is that they're probably you know there's some neurons somewhere interacting with bunch of other neurons synapses and then it gets replaced and the new neurons might have a few other synapses other than that but that's going to replace that network anyway because they have that capability. So if you do this slowly u I think um you you won't lose a lot. In fact, we we still lose memories, right? So, uh we uh we can't remember everything or we hallucinate all the time. Uh talk about hallucination, right? Uh imagine that this happened to me. No, no, no, it didn't happen. No, no, I I I remember that. So, that's like brain uh brain uh maybe part of it is new neurons that they just didn't know. So, they just made it up, right? So, um so that's one thing. The other thing is that uh these neurons probably have some internal abilities to regenerate. What I mean by that is that you know the cell can maintain itself if it has you know sort of um a great way to clean up internally like autofagy is is a very important mechanism as you know um or it has some really special DNA damage correction ability uh like stem cells have that right so pristine stem cells they don't get old you know even in 100 years old they they're still like like a a young person so uh And then you have all these other cells like GA cells and and and so on that are there to prevent all the other stuff that happens the inflammation you know GA cells of course are are are part of the immune system in a way but they they are like the immune system is not allowed into the brain in very rare cases uh uh it's like a protected area um because the immune system causes too much damage and if you can't replace it quickly that's that's a huge problem, but they have their own network of cleaning up and they probably have some sort of like a lymphatic system and and so on. Um um so if we can figure that out or if I can figure that out, we might be able to really maybe not completely regenerate but extend it um quite significantly. Maybe another 10 10 years, 20 years, 30 years for whatever. And then we might come to a point and this goes into a little bit of a science fiction now you know uh let's say in 50 years time AI might be able to figure out all of the synaptic connections in your brain like every single neural network and the neurotransmitters and everything else. So eventually you might be able to like literally simulate your your brain um you go into the matrix level. So that might allow AI to like say okay I'm going to replace all these neurons but I'm going to make sure that they reconnect all these signapses >> so so that you don't lose your identity. Um or alternately I can keep a copy here and then we can create a new brain and then transfer to that new brain that exact u uh state uh that I that I found. Uh I I'm not saying that this is possible right now. That's really science fiction error, but you can imagine that at some point we might get to that level. So I'm not I'm not too worried. I I think if it can pass this couple of decades and then keep the brain um healthy and and self-preserving for for maybe uh you know age 120, 130. And in fact, you know, the people people actually who live to to age 100, they they have very sharp minds, right? >> Because if you don't have sharp mind, you don't live very old. So that's like super correlated. So [snorts] if we can keep it for a couple of more decades and uh we'll probably find some other solutions. So if we can if we can keep the neuroinflammation low, if we can increase brain drive, neurotrophic factors, some of these things that we know does play a role in improving neuroplasticity and >> you know and growing new neurons and to do all the things that we can at least in some predictable way >> and we can have like chips for the for the memory part, you know, we could always supplement that. So >> increase the capacity >> and and and hopefully um AI will help us figure out how to deliver these therapies to the brain. >> [laughter] >> Yeah, delivery is always the biggest problem, >> right? Well, this has been such a fascinating and exciting conversation. Uh, Duria, I have a couple of more questions, closing questions for you. And I really kind of was just wanting to know if you had access, let's say there was no guard rail rails and you had access to all this data in aging biology, you know, the T- cell, you know, all the T- cell repertoire, long, you know, longitudinal uh longitudinal longitudinal cohorts, um, centinarian data, like everything, just anything you can imagine. You had it all and you had this model that was amazing that you could >> You're describing heaven for me. >> Yes. Yes. [laughter] What what what would be the the the prompt? What would be the question you would you would ask it? I mean, there'd be more than one, but what would be the first? >> Yeah. Hoping that the AI won't answer uh 42 as an answer. Um the so so so uh the the first thing I would probably ask is um not not saying that just go figure out aging or whatever because I think there has to be there there has to be certain sequence. So imagine that you have all this data. Uh what would be the the most practical um uh quickest way you can develop uh an intervention to an elderly person say age 70 80 years old that will uh immediately add five years to their lifespan. So to me that would be uh the most critical immediate question to ask because do that population doesn't have a lot of time and so we have to develop these these technologies extremely quickly and will should have you know even two years three years extend so that I can come up with the next prompt uh after that. Um so I guess that that that would be the the the first prompt I would ask. That's great. What um Okay, there's another question. So, this this one is there's no there's no money. Money is no object. Okay. There's no like you have complete like access. >> You're describing so many heavens now. >> I know. I'm just I'm curious what what your your answer is. You're going to personally build your own digital twin, >> which which I plan to >> right now. Um [laughter] what test would you prioritize? like what data sets would you prioritize, how can a person get them, um how often would you take these tests, how would you organize this information into the AI to really get the biggest bang, you know, benefit from the information it's going to give you, >> right? But you said money is not >> money is not an issue, right? Money's not an issue. >> Um so, so I would divide it into two parts. Uh one part is that we have to um um so what I would do is set up a a huge lab um you know partially automated lab where I would generate enormous amount of data on the um on the cells on the tissues in the in the lab because uh we have to go by the first principles to understand what's going on on let's say in an individual T- cell uh all these thousands of proteins, metabolites, what are they doing? Then then that will enable me to create what's called the virtual cells um and then eventually virtual tissues and you know how cells are in a special temporal manner are are are behaving and so on. So that would be that probably be the most expensive part of it and uh I'll need a lot of money. You said no limit, right? So okay. Um uh but the second part would be sort of what we talked earlier uh kind of the behavioral data from from the human humans and that data is not just of course you know all kinds of you know plasma [snorts] levels of proteins metabolize your full microbiome your full genome sequencing and all of these things are are possible by the way I mean it's you know if the cost is not an issue you can easily like UK bio Bio bank has done it for 500,000 people. You can do it for a million people. Uh and I think if you did it in a million people that would pretty much cover all the possible human humanity. I mean I it's not like everybody's perfectly uh different. You know we share a lot of things and and and and so you know from the humans collect lots of biological data but very importantly behavioral data. I think this this is something that's totally missing in a digital twin. Um like you know we were talking earlier uh ability of someone to walk certain distance, ability to to you know raise some some weights. These don't show up in any biomarker sets but they could be extremely important. um uh or ability to think, you know, their their cognitive level that that could be directly uh brain uh brain aging related and and I mean you lots of things and and you know what happens when humans are in certain environments, you know, in certain environments you even if you if you are having a very sort of healthy lifestyle that may not help you much. for example, you know, I lived in New York City for a decade. Uh, you know, my stress level was so high uh uh uh and that stress level is so harmful for you because the immune system is constantly thinking there's a threat out there and it's causing a lot of inflammation. In fact, I think people who live in New York has twice as much heart attack risk or something like that. you know that that your environment, your um uh your emotional states and how you interact with other people. All of these things will impact your aging process, your your resilience to the life, your optimistic level. By the way, being optimistic is one of the best things you can do for for aging and study after study show that. So being able to absorb um bad things that happen to you and then keep keep going. So resilience. So but these are behavioral data that's not available in in the biological set. So uh yeah uh I would do that for for a million people all over the world different parts. Um and then on the lab uh every single cell type that I can find uh decode those uh put them all together to the super intelligence and then voila we have digital twin. >> Okay Doria. So let's say someone wants to build their little mini digital twin right now using the models we have access to today. The type of data that we can aggregate you know at the consumer level today biometric data that we can that we can put in. um how would you build that mini digital twin today? >> Yeah, great question. I mean uh in fact it is possible to build a sort of a mini digital twin uh that doesn't have to be as sophisticated as I described because that that one is more uh sort of clinical trials and developing treatments. Uh but you know going back to the u example of the UK bio bank you know they didn't have trillions of data sets. they only used a few hundred data points from from each person and they were able to predict a lot of diseases. So that means that you know we we can have a lot of predictive power with the data that we're collecting uh today. Um you know another example is this uh glucose u meter that I have um you know every five minutes it shows my glucose level and then I take that data and of course I put it to chat GPT um and once you uh additional data set that becomes very uh very valuable because uh let's say that you have your lab values your cholesterol your glucose um your uh every day the the steps that you took and your sleep uh and so on. So these are actually very rich data on their own because they're uh their accumulation of lots of under uh uh underbiology that that results in that but also that puts uh AI into a context your mini digital uh twin. So my my suggestion would be uh to uh u you know provide the AI as much data as they can and on a daily basis so that so and keep it in the same context so same window so they so the model can remember that um actually there are there are some tricks uh uh to do that as well. You can keep it as like a database and tell li model go check my database and see what my new uh you know based on my new data how things have changed what suggestion you could give. Um I for example uh provide all the supplements that I take you know um you know the type of food that I eat um all of these things will make uh will make a big difference. Um so the model start to really personalize um you know sort of the uh style. It will know your style and will make uh suggestions for you. Uh rather than giving blanket statement like you should walk 10,000 steps. Well you know knows that like Daria cannot walk 10,000 steps every day but I think 3,000 would be enough for him. >> And what kind of model are we talking about? Would you be using the GPT5.5 Pro? And then what about you know these agents and codecs and how does that come into helping analyze that that database that you're creating? >> Yeah, I I think you know these models are becoming more agentic all the time. I I know OpenAI for example they integrated agents into um their codeex model the coding model and soon I'm sure it will be part of all of chat GPT. Um you don't you don't I don't think you need very sophisticated models for that. What is important is that uh really maintaining that context. Uh so hopefully the models will have a larger memory and they can remember. So you chat can keep certain memories about you but it's still kind of limited. Uh um it's not just CHP like you can use JNI for example which has a longer um uh context windows um or or cloud for that matter. I think most of the models can handle that uh information. And they don't have problem dealing with large data sets. As I mentioned, I can put millions of data sets and they're able to analyze that. What they need is that they need to remember how things were a month ago because that's before and after. Before and after is extremely valuable. So the model will know he started taking vitamin D3. Oh, these things changed after that that you may not notice or is glucose looks better because of you know when that's that change happened. So it's starts to make those lengths and and that's I think the critical point because you need all of that context in the in the AI model to to give you sort of a better uh prediction on what to use and what not to use. Okay, you were using that well maybe that was not a great idea so change it. um or change the dolls or or whatnot. >> Yeah, that's interesting. It kind of reminded me of a question that I did want to ask you about, you know, these AI models and future AI advances when you think about these qualities. So, like persistent memory, expanded context handling, it seems like those seem to be more important. >> Absolutely. that I I think um for me uh memory which which brings the context uh so the models are now able to think for quite long time and they don't because previously the models would just um even in the same context window if you had a million context windows uh after a while they would just fall off because they would forget even what they were thinking about. Now they have this ability to constantly um go and check on it. So uh I think in the next few months this is going to happen. Uh so that that will that will have a tremendous impact. Well that's memory is everything. >> So so you think so how long are we talking like let's say you know you started a vitamin D supplement 6 months ago. Put that you have the same window and you start in that window you have that you know entry point that the date and then you keep adding about you know you add your your your data in. has got all the data um right now. Can it go back that far or how far can it go back? And >> um if you have that data somewhere in your database uh for example um I adapted a a technique that uh Karpathi who's a famous AI researcher described. Uh so you can um turn you can create your own Viki sort of Wikipedia kind of a thing like personal um you take uh you know uh if you have all your data somewhere uh you can ask AI just pull all that and put it into a Wikipedia like you know you can do it daily or weekly depending on the environment whatever um and so now you're building your own database health database uh which AI can help you update it if you have that data it can go years doesn't matter like you can have 10 years of data it will analyze all of that um uh but >> it has that memory it can like >> yeah so in in in in the same context if you provide all of that I mean it's still limited with you know maybe a million tokens or something but no one's going to have million token data set even if you if you calculate 10 years so so that's that's not that's not a problem the problem is like if if you have If you want this to be continuous like you just give AI okay here's the data today that it should be able to remember what was yesterday what was 2 months ago so you don't have to give you know all of the um uh you don't have to keep your own database and and give all that again and again because you have to do that every time right so the your whole and that will spend a lot of tokens and stuff like that so uh but but I think this is this is going to be uh this is going to be sold How do you not bias? How do you lower the the ability of yourself to bias what you know GPT 5.5 Pro is is going to feed you back, right? Like based on what you're asking it and I mean [clears throat] I I find sometimes I I might be able to bias it a little bit. Do you do you do you know what I'm talking about? >> Yeah, sure. I [clears throat] mean that's why I think uh we are in sort of the experimental phase um in a way um everyone has to do their own kind of validation um as the models are getting better. What I mean by that is that again you know of course don't try harmful things and then you know uh don't go into risk but you know for daily daily use um you might be taking vitamin D and then you you stop taking vitamin D so that you're just doing an experiment like before and after and then you collect that data before and after u and then AI gives you one solution says well you know taking this dose of vitamin D I think is important So then you can start that dose again and then see see what happens. If if you reach the same level as before means that AI made a good prediction like you need to see after you have to have that record before and after so that you you are the judge. Well what this was a good idea so I'm I'm glad that I listened to Jupy. Well if it wasn't a good idea it didn't kill you. It didn't make you sick. So that's that's also fine. >> Yeah. I guess for someone that's already taking a lot of supplements for example, they're not going to have that before and after. Then also you have to know like how long do you wait you know for example to for the wash out period and >> yeah and whatnot. the the hope is that if you provide that very frequently um in fact um I can mention one thing uh for example the the lab values like you go and measure your cholesterol glucose sodium whatever they always give you a range right so if it's within this range it's normal well how do you know that uh because you can be at the top of the range uh that might be your abnormal somebody else's normal somebody might be a little bit over the normal and might still be okay um or vice versa because we don't know the level on a personalized level. So we we calculate population base. So okay so this range is good for this population. So in a way um if you have three or four measurements let's say every few months you can develop your own set point normal. you know the the AI will know your normal for glucose is 90 not 70 not 100 or not 105 somebody else might be 102. So it knows that based on that that measurements. So then then it starts to give you advice based on your data set your set points because if yours is 100 and suddenly dropped to 70 maybe that's not a good thing you know I'm just I'm just uh giving an example uh uh so uh that's why that continuous data collection is so important uh uh with with glucose meter I collected every 5 minutes uh the more data the better >> well Duria thank you so much for sitting down with me today and talking about this exciting I mean frontier that we're exploring you know curing disease extending human life expectancy obviously health span reversing aging perhaps getting to human 2.0 you know where we're enhancing you know genetic you know features as well um very exciting time to be in and if we cannot die in the next 10 to 15 years >> it may be even more exciting. Yes, absolutely. Because uh you know the last thing I will say uh this is so unique in human history. Uh because a decade ago uh if you set someone well you should be very healthy you know uh do this do that and they can say well it's only going to extend my life maybe two years or three years. I just want to live my life and I don't care about living few more years as an old age. and that that was perfectly, you know, relevant. That's not the case now. Living an extra one year could make you reach that threshold where there's going to be the ability to to treat many diseases and reverse your aging and give you another decade, give you another 20 years and then once you reach that, you get another 10 years, another so like even every day counts now in my opinion. Uh so uh that's why don't die for [laughter] >> um where people can find out more about your research and they can follow you. I follow you on X. Um maybe you can tell people how to follow you, what your user your Twitter follower or sorry your ex user handle is and where else they can find you. >> Uh yeah, my my main account is an X. Uh, it's at Daria D E R Y A T R under dash. Um, if they write Dario Nutmas, I think I'll I'll show up. Um, that's that's where I, you know, do most of my communication. Um, I have a LinkedIn account, but I don't post that often there. Um, I I've been planning to start up a a sort of a YouTube channel, but um, I don't think I'll ever do that because I'll never have the time. you know, it's it's really amazing what what you're doing because video takes a lot of a lot of effort. Uh so for me is the the fastest way. Uh in fact, I I even had a Substack uh account, but just couldn't find the time to write long uh uh long messages. So So uh X is the best way. >> Well, I really encourage people to follow you on X. post. I mean, just every day there's something interesting that you're posting on X and so I highly recommend that people do follow you >> as um many already do. So, thanks again for the research you're doing and for I'm I'm excited to see what um what's going to happen in the next couple of months. >> Looking forward to it. Very optimistic. Thank you. Thank you very much. It was great.