I'm Tom Bilyeu, and this is Impact Theory. Welcome Back to part two of my conversation with the one and only Dr. Peter Attia. To your point about variability of response to something feels a lot in the way of interpretation.
So I was actually listening to an episode, Ewing Norton. He was talking about anxiety, something unfortunately, I know up close and personal. And he was saying it was a big breakthrough for him to recontextualize nothing about the physiological stimulus needed to change for him to change his relationship with it. He just had to tell himself, oh, this means that I'm focused, I'm ready, and this is a good thing, and this is engaging my brain and make sure I take this seriously and I'm gonna move forward.
And he said that that's completely changed his sense of what anxiety was. So that is. And it's. It's very similar to what I do when either I feel anxiety coming on or like when I've smoked and it.
I can feel it, that the panic attack, trying to become the thing that's hijacked my brain. So the way that I think of these things, this is a means of processing data. It's an interpretation of a stimulus and the way that I will often change my psychological interpretation of the stimulus. And this is what I'm trying to get to.
Psyche. That's me, is psyche. It is my sense of what I call the overwatch mechanism or my sense of self that there feels like. I know this isn't true, but it feels like there is a part of my being that exists above whatever physiological state I'm in.
So there's a watcher that sees me being drunk. There's a watcher that sees me being high again. I know that's not accurate, but that's how it feels. Now, that watcher can send a signal, a signal.
I almost always send this. Don't be a pussy. And so I'll just tell myself, ah, like, knock it off. You don't.
You're not gonna let yourself be paranoid. Stop that. I once had a vagal response to getting a bunch of injections in my wrist. And I didn't know it was a vague response the first time happened.
So the second time happened and the guy was like, basically just being weak. I was like, hold on. And because I now had a new context for the same thing, I just told myself, absolutely not, like, you don't get. Because I felt like I was about to black out.
I was like, 100% not. Like, this is not a physiological thing, a psychological thing that to Me, that overwatch mechanism, however badly I just explained it, is psyche. It is the ability to interpret what is happening to you rather than be swept up by it. And so if you are swept away by, oh, I just smoked weed and I'm feeling paranoid and now I'm having a full blown panic attack and I cannot get above it.
That's a problem. Now, the one I struggled with is, by the way, I think that's actually a really interesting idea, Tom. I think that's pretty cool. I'd like to, I'm continue to think about that through the lens of how I imagine the relationship.
We have to mind all drink substances. You know, for me, just like I never drink to the point of even remotely getting like. I mean, I've had too many, if I even start to notice it. I also have very high tolerance alcohol.
So if I have two glasses of wine, like I feel literally nothing, nothing, nothing. There is not even a semblance of a buzz that I would feel. But that's an interesting point. Right?
Is when you get to the point where that, that third eye of yours is not able to see what's going on. Yeah. I would argue that that's not a place you ever want to be. Agreed.
Regardless of substance. Agreed. The only time that I cannot extract myself from it is if my anxiety runs wild. Now, there is one thing that I've managed to peg that I'm like, this causes anxiety and this makes me so sad because I love these so much.
But Diet Monster, now, I ran this experiment so for a long. Monster is taurine, caffeine. What's the substance? Those are the two that I could tell you offhand what all the other.
I mean, it's just long. But is that different? I don't drink energy drinks particularly. But is that different than Red Bull?
Is that different than Monster is the only one that I enjoy enough that I've drank so many of them is good enough that you can drink the can and therefore hit that critical dose of whatever's in it. Yes. And I don't feel it on the day, I feel it the next day. So it's just really took me a long time to realize what it was.
But I went through a period where admittedly I was under so much stress that I actually thought, this is just. I actually have reason to be anxious. And then I was like. But there was one time I turned left instead of right and I had this massive spike of anxiety and it was literally just, exit parking garage.
And so I was like, like, that's When I realized, oh, this is generalized anxiety because there's no reason I should be anxious. And so that put me on the path to, okay, there's something I'm. I've either allowed my anxiety to hardwire itself so deeply in my brain that now I'm interpreting stimuli in a way that doesn't make any sense, or I. Because I was asking myself, what advice would you give if somebody came to you this.
And I'm just like, I would tell somebody, this is your diet. Just because I. You. You are helping slowly maybe bring me this, but I just come from a place of.
Anything bad in your life. You're your catch diet. It's your diet. Like, I just, I have such a knee jerk reaction, especially given my family history.
Um, and so I was like, what are you consuming a lot of? Stop consuming it. See what happens. And I was like, well, I drink a hell of a lot of monster.
And I was read a day. The big ones, the big ones. And they were amazing. I love them so much.
Uh, but I just cut them out and it went away. And I was like, wow. And then I literally. What's interesting to me, Tom, in that story is not that you had that response, it's that you didn't have that respons on the day of and it was the day after.
That's actually kind of interesting. Interesting. And it makes it really hard to steer by, especially because I have a shockingly poor ability to recognize patterns. So that was hard.
That took me a while. And it was really just whittling things down. Like, I know what I'm consuming a lot of. And yeah, finally cutting out one out.
Because again, I would cut it out, but nothing would change on the day. So I thought, okay, well what was that? And then once I was like, leave it out for three weeks and see what happens. And then that was when I was like, well, I really think this is.
And then I would. I tried reintroducing it and came back. So I was like, wow, there's just no way around that. Because I love them.
So I would keep them if I could. Can you imagine if we had a little cave friend here and we gave him some monsters Cheeses, man. So we gotta play this game sometimes. I love playing this game.
You go back in time, little time machine, 10,000 years ago, you bring one of our ancestors forward, assume he can speak our language, we'll give him that, that, like, just imagine, imagine. And how much does he stroke out over the things he sees? Like, think about him like the next time you're sitting in traffic, annoyed. Just imagine he's sitting in the passenger seat.
Like, how are you explaining this thing to him that he is sitting in? That's crazy. It's crazy. And that's what makes me want to save the modern world.
I am absolutely here for it. I'm here for flying cars. I'm here for going to Mars. I'm certainly here for virtual worlds so you can explore that, have more depth than the real world.
But I also want to recognize that there's a biological reality to be faced about what we do. Speaking of that, what do you think is a thing that we think of as wisdom right now for health longevity that we're going to cringe at in five to 10 years? You know, we're in the process of doing this, but essentially, I was just talking with your wife before we started today, and we were talking a little bit about menopausal care. So I think we're in the process of making this transition.
But if you ask me this question, five years ago, I would have said 100%. The denial of HRT for women. Right. So this.
This idea that women should suffer their way through menopause and they should never be given hormones. Like, there's still a lot of doctors who believe that. And I hope that in five years, we look back at that as just about the way we would now look back at how they did surgery in the 1880s without anesthetic. Like, Tom, how do you feel about having an amputation where I'm going to give you a piece of leather to bite on?
Sign me up. Right. That's pretty savage. And that's.
I hope that in five years where we have that view of if you have a woman who wants HRT and you deny it to her, that's, in my mind, that's. That's that level of savage. And the reason why now that they deny it is they worry it increases cancer risk. Yeah, there's.
We still have this kind of demonization of estrogen and this belief that HRT causes breast cancer, which, again, the data are really clear on this, that HRT does not cause breast cancer. And even if there's a tiny, tiny subset of women who are going to see a slight increase in the risk of getting breast cancer, there's no increase in the risk of mortality from breast cancer. And even that risk of the increase of breast cancer is almost assuredly due to a formulation of synthetic progestins that were used years ago that aren't used today. So in its current form, you know, it's just.
There's just no evidence that this is happening. In fact, the evidence is probably a little bit the opposite. So that'd be interesting. You know, as far as, like, on a more fringe side, I don't know if in five years, this is one of the things you can't prove a negative, but I'm pretty skeptical of the mass use of stem cells.
So I think that. I think there's probably some utility to stem cells in certain situations. But, you know, when you look at the ubiquity of which stem cells are proposed as the solution for everything, you know, intravenous stem cells, exosomes, all these things, like, I like the total lack of data supporting any of this stuff, and I just hard time believing it. So there's clearly indications where I think stem cells could be beneficial.
My hope is that it gets studied so that we could understand how to standardize stem cells and how to, you know, use them correctly, make a drug out of it, Right? So if you want a drug approved by the fda, you feel that demonstrate that you can produce the drug the same every time under GMP conditions. Can't do that. Can't have a drug like in the wild west here.
So, like, we have to be able to do this thing with stem cells. So my belief is if we did that by stem cells, we would probably find a handful of things for which it works and it'd be a pretty narrow indication, and a whole bunch of things for which it's total malarkey and we should just stop stealing people's money. Well, that makes me sad, Peter. I would really like to think that there's a magic hero.
Speaking of magic heralds for aging. So the last time that you and I met, you said you were on the cusp of thinking maybe testosterone replacement therapy might be a thing you want to do. Have you done it? Yeah.
So last. So over the winter, I just continued to watch my testosterone levels go down and down and down. And so, yeah, about six months ago, I started hcg. So.
So HCG is. It's a hormone that mimics luteinizing hormone. So our pituitary gland makes two hormones. Liningizing hormone, follicle stimulating hormone.
They tell the testes to make testosterone. So the first question I had, which is, again, the question we ask with our patients, right? If you have really low testosterone, is it primary or is it secondary? Is it low because the testes can't make anymore, or is it low because the pituitary gland is not making the signal to do that.
So one way to test that is to actually take hcg. And if you take HCG and your testosterone goes up, then, you know, the problem wasn't that you couldn't make it. The problem is that you're not getting the right signal for whatever reason. So in my case, it turned out that it worked.
So when I took hcg, which is, you know, a small peptide that you take, so you're doing, you know, sub Q injection a couple times a week, my testosterone got to normal. And that obviously tells me that the reason I've had a couple years of super low testosterone is something in my pituitary. And given that virtually everything I think I'm managing well. The one thing that I would point to is probably stress.
I would say probably, you know, stress is the one thing because, you know, my sleep is actually pretty good. I don't tend to struggle with sleep. I, you know, by all the means that I track it. It's really something that I've optimized.
Not perfect every night, of course, but on balance, it's not a chronic issue. Whereas for many people, you know, poor sleep is the cause of central hypogonadism, which is a technical term for that. So, yeah, I'm just walking around the world testosterone now. Now, is that where you would start?
Any guy that has low testosterone, is that the first thing you test or. It's not necessarily. Every case is a little bit different. But that would be a discussion which is, you know, where, you know, for me, again, it doesn't really matter.
I could take testosterone as well because I'm not planning to have kids. And all the things that would come my way if I stopped making my own testosterone, which is what would happen if you took exogenous testosterone for a few years, you would. You would stop making your own testosterone. That's not actually a concern for someone my age, but I don't know, for whatever reason, just psychologically, it was a little bit of an easier pill to swallow to start with this kind of indirect way.
Now, would HCG pop you for doping? Yep. It would still be a WADA band drug. Got it.
Okay, very interesting. What do you consider normal for testosterone for guys? Well, it's a bit complicated because, I mean, I can give you obviously the percentiles. Right.
So for, you know, a guy my age. And by the way, I don't think normal is what we want aspired. I think we want to aspire to top quartile. Right.
If you're going to bother treating somebody, you should at least take Them to the, I think the sort of top quartile. So you know, I just want to see my total tea above 800 and my free tea, you know, kind of above 16. Wow. You got your total T over 800 with HCG.
Yeah. Starting with a T of 300. Wow. Yeah, maybe 275.
I mean pretty low total T maybe 275 with a free T of about 5 and then went to a total T of about 800 and a free T of. You know, actually the first time I took it I overcooked it a little bit. I was more sensitive to the HTG. I ended up going to like 1200 total tea with a free T of 22.
So I backed it down and now I think I walk around probably about 800 with a free tea of 15ish. Does it feel different? It does. I definitely feel better.
I would say the two areas where I've noticed the biggest difference is I genuinely just feel better from a mood perspective. And secondly, I'm definitely stronger in the gym. What about sex drive was never an issue before. So that wasn't like despite my low T, I didn't, I didn't really struggle for libido.
It's interesting. So when I was in my, call it early to mid-20s, my poor wife, the friction burns. But as I've gotten older I've hit a more reasonable pace. But I've always wondered if I started doing some sort of testosterone replacement therapy, would I go back to my poor wife's suffering or does this stay somewhat balanced?
Like how does that depends on how high your levels go, how high the testosterone level depends how. It would depend a little bit on how high you are now. And here's the part about this, it's a little bit complicated. What we can't measure is the thing that matters most.
So your testosterone and your free testosterone are just proxies for what we most care about. What we most care about is how many molecules of testosterone are binding to your androgen receptors. So it's androgen receptor saturation that is the variable of interest. The people who win, they're not smarter, they just absorb more ideas faster.
There are thousands of books right now on business psychology, leadership books that have already changed how the best operators think. And everyone you haven't read is a gap between you and the people who have. This is a bandwidth problem, not a discipline problem. And that's exactly what Blinkist was built to solve.
It takes the world's best non fiction books and distills them into 15 minute summaries you can read or listen to scattered over 9,000 titles. We're talking atomic habits, Thinking fast and slow. The hard thing about hard things? Books you know you should have read by now.
Done in the time it takes to drink your coffee. And because it's built for mobile, your commute becomes a classroom. Blinkist also sends you personalized recommendations so you actually build a daily learning practice instead. Instead of just planning to do one.
Start your free trial and save an extra 30%. Go to blinkist.com impact Again, that's blinkist.com impact Start learning aggressively today. It will change your life. Let's talk about what is going on in your closet this time of year.
Most people realize half of what they own isn't even worth keeping. Things that look fine in the store, but they don't hold up. Pieces you keep but never reach for. I know that one all too well.
Spring is when you see it all clearly. You don't need more, you need better. That's where Quince comes in. I picked up one of their cashmere sweaters and the quality genuinely shocked me.
It's incredible. It's incredibly soft, well made. You can feel the difference immediately. That's 100% Mongolian cashmere for you.
And that is the same material luxury brands charge a fortune for. Refresh your wardrobe with Quince. Go to quince.comimpactpod for free shipping and 365 day returns. Now available in Canada as well.
Go to Q U I n c e.com impact pod for free shipping and 365 day returns. Quince.com impact pod let's talk about the most powerful force in wealth building. Compounding. There's a reason that people have talked about this forever.
You put an asset to work, it generates more of itself. Over time, the curve bends way in your favor. Every serious wealth builder knows how compounding works. Most gold owners, though, have completely opted out of it.
Monetary Metals changes that equation. They've built a way for your gold to earn yield paid not in dollars but in more physical gold up to 4% per year. More ounces stacking every single month. Compounding in one of the only assets that can't be printed, inflated or debased.
And storage and insurance are included, by the way, so hidden fees don't eat your yield. The average person is passively absorbing inflation. This is how you can pound your way out of it. Click the link in the show notes or visit monetary-metals.comimpact to learn more.
This is a paid advertisement, so if a guy has symptoms and his androgen receptors are not saturated, that's the guy you want to be replacing. Conversely, if a guy has no symptoms, or if a guy has symptoms but his androgen receptors are fully saturated, you need to look for something else. And we don't have a test that measures that. So the way testosterone works, of course, testosterone comes into the cell, binds the antigen receptor, and then that testosterone androgen receptor complex is what goes into the nucleus and it binds as a transcription factor that increases gene transcription.
So all those things are happening in a favorable way if you have a sufficient amount of that. But if you don't, that's where you're deficient. So when people have really low T, we just kind of make the assumption that, hey, they probably have not saturated their androgen receptors. Where it gets a bit complicated is when guys have like a T of 700, a free T of 10, and they have symptoms.
And at the end of the day, I just think sometimes you just try it empirically. You just say, well, if we boost you such that your free t goes from 10 to 20 and you don't feel a difference, this is not a T problem. You shouldn't be on T now. Testosterone, from what I've read, is dropping like crazy.
Minaret across averages, all time lows. What's happening? Should they be on hcg? I'm guessing answers, no.
There are plenty of things I can do before they get there. Well, the most common, we think that the most common reason for that trend, and it is a pretty significant trend. So for context, I think the average 30 year old today has about the same level of testosterone as the average 50 year old 30 years ago. God.
Yeah. So that's, that's pretty clear. We actually look those data up and the best explanation for it honestly is increasing body fat. It's just that men are having, men have more body fat today than they did 34 years ago.
And so body fat is doing two things to lower testosterone. The first is it's increasing the aromatization of testosterone. So you're making more estrogen from your testosterone. Estrogen is very important.
You want to actually make estrogen, but it needs to be in a balance, right? So if you're making more estrogen than you should, you're depriving yourself of the substrate testosterone as you want. And the second thing is more body fat is going to be accompanied by other metabolic changes that are going to lead to more like more sexual mind globulin, kind of like less available testosterone. So you're going to make less testosterone, you're going to utilize less testosterone because of more being aromatized.
And so that's probably the single biggest driver of what is obviously happening, which is a trend towards lower and lower testosterone. And then today for giving age. So the fat itself is sending out a signal that's causing the body to react differently to testosterone. Well, so fat is fat, fat is inflammatory, right?
So the more inflammation you have, the more metabolic disease you have, the more that process of making testosterone is going to get interrupted. So something in that hypothalamic pituitary, gonadal access is being disrupted. It's producing less testosterone. So you're gonna make less testosterone.
The testosterone you make, you're gonna spend. You're gonna put more of it into estradiol. So it's kind of like a double whammy. All right, at what number do you want to see men looking at their testosterone?
Like, what, what number is too low? And is that entirely tied to age? Well, it's highly correlated with age. So we, you know, I think we've, I know we've written about this, where we've got, we kind of go through the table.
So by decade, what are the percentiles? So you can, you know, somewhere on our site lives this table where you can see by decade of life, this would be bottom 25%, second quartile, third quartile, top quartile. Um, so very critical age response. Again, I, I think that it would always be desirable to be in the top quartile for your, for your age bracket, which by definition means three quarters of men are below that.
Now, why not push a 6 year old into a 25 year old top quartile? Well, there, there are side effects of testosterone, so you have to be kind of mindful of the side effects. Right. So one of the side effects of testosterone is you're gonna make more dht.
So hydrotestosterone is a hormone that is more potent than testosterone. So it's an even more potent binder of the antigen receptor. And it does some really good things, but also it does some desirable things. So one of the things it does, it's going to accelerate hair loss.
So if a guy is particularly concerned with androgenic hair loss, that's going to increase. It also increases the size of the prostate gland. And so when a guy is 60, his prostate gland is already kind of getting big enough on its own, so we don't really need to put too much more fuel on that fire. Now, of course, there are ways around that because you can take a drug that blocks the conversion of testosterone to dht.
And of course, many men do this. So drugs like finasteride and dutasteride, which are commonly used either for hair loss or. Yeah, so propecia would be the hair loss version of finasteride. Doesn't finasteride, though have like potentially huge sexual consequences?
Yes. So some men have really nasty side effects to finasteride, where they have a total loss of libido and it kills their mood and stuff like that. And fortunately, well, fortunately for most men, if they stop the drug, the symptom goes away. But there is a small subset of men who appear to maybe be irreversibly damaged by that.
Fortunately, that number seems to be very small. It's called Post Finasteride syndrome pfs. And it definitely is. You know, it's a controversial topic in the urology literature.
But it's another reason why I don't just recommend those drugs willy nilly. I think I want to make sure that if a guy's taking one of those drugs, he really kind of understands the potential risks of it. But there are other drawbacks to taking those drugs. Right?
So you can take that drug. Let's say you're in the majority of the cases of folks who, they don't have that side effect. Right. They don't have post syndrome.
But now something else happens which is you've already artificially suppressed your psa. So now one of the things we want to be, you know, making sure of an older man, or basically any man above 50, basically is what's happening with respect to your risk for prostate cancer. When I just took away one of my best screening tools. That's not the other world, if the doctor understands that.
But most doctors don't know that when you put a man on finasteride or dupasteride, you have to be much more diligent about how you screen for prostate cancer and you're looking for much smaller changes in PSA to cause alarm. Very, very interesting. Going back to something that potentially is cringeworthy down the road or maybe is amazing. Do you know Brian Johnson and his don't die movement?
I don't know him, but I've heard it. What do you think about that? I don't think that's on. So it's tied to AI basically.
So he's saying, hey, now is like the period you don't want to die. Because if AI comes in and does what it's going to do, then we really might be able to find patterns of biology that right now are just completely invisible to us that will allow us to radically extend human life. So the question then really becomes about AI. Do you see anything in AI that you think is promising that could have a substantive impact on the way that we do health now, even if it's not to radically extend lifespan that may radically extend health span or help us overcome diseases like cancer or whatever?
I mean, I would put myself in the camp of being very excited about AI, which is not a hard camp to be in today. I think most people are. I also think we're an AI bubble. So this, I think where we are for AI right now is where the Internet was in the late 90s.
Do you mean as an investor, though, or as in the ultimate. So, absolutely, as an investor, I think it's a totally overpriced bubble that's going to crash. But also in terms of expectation and timeline of expectation. Right.
So if you think back, and you're old enough as well to remember in the mid-90s, as we were heading into the new millennium, I mean, people just thought the Internet was going to change the world in ways that it hasn't quite. And yet there are ways in which it has changed the world that nobody could have predicted at the time. Let's take two obvious examples, Uber and Airbnb. Like, nobody thought about that at the time, right?
People thought about commerce, and that was a big deal, but even took longer there for it to really become the profitable, you know, disruptive technology. So I just think there's a parallel between how we think about AI and how we thought about the Internet. I think AI will be more important than the Internet. So I think when we're looking back in, you know, 50 years, I do believe, and I don't think I'm alone in believing this thing, AI will be an even bigger disruption, hopefully in a positive way, than the Internet was.
But I think there are certain things that will take longer than we expect. So to your question, where do I think it's going to help? I think. I think there's some really unsexy things it could do to help that, you know, this would be a good time for people to go and get a snack or go take a leak.
But, like, if you think about healthcare, for example, like adjudication and reconciliation of healthcare billing is arguably the stupidest, most inane, like, part of this country, right? Like, the amount of dollars that are wasted, the amount of brain damage that is inflicted on our species trying to reconcile medical payments and stuff like that, it's impossible by the way, nobody knows how to do it. Like, if I gave you an eob, an explanation of benefits and like walked you through what was in there, you'd be like, dude, why don't you just talk me in Japanese? Like I literally know what you're talking about.
You don't have much to pay. No. By the way, you're getting ripped off. So AI should solve that problem.
Like we need to train AI to basically solve the entire system of how medication is educated. I would love for AI to absolutely eviscerate PBMs, pharmacy benefit, benefit management entities. So these are the companies that live between the payers and the employers. They're responsible for totally ripping off everybody in America.
These are the biggest cancer in the healthcare system today. So if we could just eradicate that and actually have a pricing optimization run by an AI that is not skimming off literally tens of billions of dollars to itself, that would be incredible. None of those things really help you live longer. So the question of how are we going to live longer?
Probably the most exciting thing that AI has done at the moment. Again, not that sexy if you don't understand the biology. But do you know how a protein works? Like the idea of what a protein is in the body?
Yes, but good lord, as a person at a 80,000 foot view, not in any way magnificent. All right, so you eat a steak right now, right? So you're eating and a steak is great because it has all the amino acids in it, all the essential amino acids. So you just ingested a whole bunch of protein that gets broken down into these building blocks called amino acids.
There are 20 of them. When your body wants to make something, a lot of things your body makes are proteins. They can be structural proteins like collagen, hair, muscle. They can be functional proteins like enzymes.
Your DNA gets transcribed into rna. That RNA gets turned into a signal to make a protein and it strings together the amino acids. Now, by itself, that's not the protein. It has to undergo structural change and what's called post translational modification.
So sometimes it gets turned into a helix, sometimes it gets turned into something that folds and that's called the secondary structure of the protein and ultimately gets folded over onto itself into the most complicated three dimensional structure you can ever imagine with thousands of amino acids. Okay. Until AI came along, we had absolutely no clue what the relationship was like between the string of amino acids and the final shape of the protein. And that is a big part of what made biotech really, really hard.
So we could look at a receptor and Say, man, I'd really love to have a drug that fits into that receptor and turns it on. I know the shape of the receptor and I therefore know the shape of the thing that needs to fit in it. In other words, I have the keyhole. I can impute what the key is, but the key is a complicated structure and I have to from the key impute what the 4,000amino acids were that went into it.
That had to be done by trial and error. Woof. Today, that is done with AI that's actively happening right now. Is this Alphafold or let's talk about who's profiting from your personal information.
Someone is making money from your data every single day. You just don't see the transaction. Data brokers collect your personal information and sell it for profit. Your address, your phone number, your email, your Social Security number.
Yikes. It's bought and sold constantly without your knowledge or consent. Scammers buy it for identity theft. Spammers buy it for targeted harassment.
Criminals buy it to find you. And the crazy part is, it's completely legal. The solution is Incogni. Incogni tracks down your data across hundreds of data broker sites and removes it automatically.
You authorize them, once they handle everything else, they remove your data and keep it removed with repeated requests, plus with customer removals. You send them any link where your data appears and their team will remove it. They can't harm you if they can't find you. Go to incogni.com impact and use code impact for 60% off an annual plan.
Try it risk free for 30 days. That's I N C-O-G-N-I.com impact and use code impact. It's never too early to plan your summer story in Europe with WestJet. From rolling country signs to cobblestone streets.
Begin your next chapter. Book your [email protected] or call your travel agent WestJet, where your story takes off. Okay, how's that being used right now? Other than now?
We know, but how do we get that into. So it's basically changing pharma. It's basically changing drugs as we speak. Yeah.
Now, I want to be clear. Everybody said, oh my God, Alpha Fold is going to make drug discovery take one year instead of 10. Not really, because what they're forgetting is the long piece of this can't be done yet by AI, which is the clinical trials. So it's totally shaving off all this time at the beginning and saving a ton of money.
But the long hole in the tent of pharma is not the drug discovery, it's the clinical trial. You're still going to have to test this in humans in a very careful way that is, you know, incremental. Like, first, let's make sure it's safe. Okay, let's do a dose escalation.
Okay. Now let's make sure it kind of works. Okay. And it kind of works, and it's kind of safe.
Now let's do a bigger trial over a longer period of time with enough people in it that if it really works and it's really safe, we're comfortable approving it. So all of that still has to happen. I'm sure AI will come up with ways to help make that better. But again, as nerdy as that example is, I mean, the folding pattern is pretty freaking cool.
Yeah, that stuff is unbelievably cool. And this is one of those things. At the risk of me dragging people into my madness, this is why I'm so obsessed with game development. When you start doing game development, you realize that all, all, all of life is simply based on a set of rules.
And once you know those rules, you can actually begin to predict what will happen. But the joy of being human is you can't predict all the things that are going to happen. And so now, as you create a virtual world, you can set up these rules that are so complex and interact in these really fascinating ways that you effectively. I don't think people are gonna like this sense, but you effectively get to play God.
So as a developer, you play God because you're creating the set of rules. But as somebody exploring those worlds, imagine creating a world that is based on a set of rules, and it will spawn literally an infinite number of worlds. And you can have a world that is an entire universe unto itself that you will get to explore uniquely. Nobody else will ever encounter that world unless you want them to.
And once you understand that the sophistication with which you create these virtual worlds is tied to what the AI can predict in our world. So now you could. Now in the future, you'll be able to apply AI like AlphaFold to a virtual world where it's building up from amino acid building blocks, but creating creatures in your world that are, like, built literally down at that level. And so it becomes fascinating.
I mean, obviously this is, this is, you know, an incredible topic to explore. It's one for which. So let's take a step back so people can kind of understand the complexity of the problem. I know you understand this, but I think it's important for Listeners to understand we're about 25 years out from the sequencing of the human genome.
Right. So human genome project was completed, I think, about the year 2000. And that was heralded as the single most important breakthrough in all of human health. But it hasn't really panned out.
In other words, that didn't come close. Didn't come close to being as helpful to human health as the discovery of insulin, penicillin, sterile technique, sanitation, like, you know, the elucidation of the human genome has saved fewer lives in the last 25 years than 10 minutes of what sanitation, antibiotics, and insulin have saved since we've been talking. Okay, so why is that? Well, it's because of something you sort of alluded to, But I'm going to twist it and make it my words.
The rule book doesn't transmit. So what we coded was the human genotype. There are 3 billion base pairs that define you and define me, by the way. The only difference between you and me is a couple percent.
So it's virtually identical genetically. But obviously, if you look at us, we look nothing alike. And we both have two eyes. You know, we stand on our feet like you can recognize that we're the same species.
We otherwise look quite different. So what we have no clue of, for the most part, Is the relationship between the genotype and the phenotype. It's why we can't go and make something genetically. So as an example, let's assume that I had all the control to manipulate the genome that I wanted, which I don't, by the way.
But let's assume CRISPR was so perfect that I could. And it is, by the way, CRISPR is probably, at this point good enough that we can change any gene. What we can't do is insert it. So we're good at developing payloads.
We still have horrible delivery systems. Let's assume all that's done. Let's assume I can make any gene of any size, edit any gene of any size, and deliver it through your entire genome if I want to. That's a big if.
Not clear that's ever gonna happen. Let's grant that it's gonna happen. If you said, peter, I want to be smarter, can you genetically manipulate me to be smarter? We would say we don't have the foggiest clue what genes are responsible for being smart.
Okay, can you make me taller? Never mind me. My growth plates are fused. Can you make my kid taller?
No. We have no clue what genes, right? We can't even. We wouldn't even know what Genes to adjust the eye color of your offspring.
Think of how trivial these things are. And we, we do not understand the relationship between genotype and phenotype. And so it just speaks to how complex this problem is. Now is AI going to help us on that front?
I sure as hell hope so. But if we don't have that, we don't even have a prayer of basically doing AI clinical trials. But here's the other thing. Genotype is only part of the equation.
It's also adaptation. So it's the whole nurture nature problem. So why is it that you could take three genetically identical embryos and put them in three totally different environments and they're end up producing three pretty different people that by the way, might respond different to three different drugs. And why is that?
Well, it's differential gene expression. This is now getting into what's called the epigenome. And you could say, well Peter, can we figure out that? Okay, but like it's getting awfully complicated.
So look, the beauty of futurism is it makes us all look stupid because all this, all the stuff we sit here and think is going to happen, we can't imagine the things that are going to happen. But my intuition is that there are lots of really incredible ways that AI is going to make what I do so much better. But, but, but they're, they're not as like wham, bam, super sexy as maybe people would like to believe, at least not in the foreseeable future. Do you use AI right now?
Yes, use it quite a bit. You know, definitely. Still kind of struggling with some of it. Like I think, honestly it's just, you know, I would, I would describe AI right now as a really unhelpful analyst.
Like it's an analyst, it functions as an analyst, but it's like the fastest working, fastest reading, least intelligent analyst I have. Like there's some value of that, right? Amazing at spinning out a first draft of thought. But like, you know, we have five research analysts on our team.
Like they're all here. GPT4 is like way down here, right? Yeah, I. For the type of work we do, obviously GPT4 is way above all of us in certain areas.
So the type of work we do, it's borderline moron. I will be very surprised if in call it three years that for an analyst position that it's not able to match or surpass the brightest that you have. That's going to come down to the training set. So this is for certain things it already does.
Like you know, I could Argue maybe for coding and things like that. With training, set aside the real challenge of the work we do. It's very difficult to train an AI on why? Because how do you.
It just hasn't been done. Like we could do it, but like you have to quit our jobs and just work on training the AI. So nobody's out there, to my knowledge, who's trying to train an AI to do what we do, which is how do you critically read papers and you know, how do you read 10 papers on a subject matter, go and read all the references, be critical of which studies suck, which ones are good, what the limitations and strengths are of each paper and all of that stuff. I mean, what AI could do?
If AI could do that, it could clearly write better than us, it could clearly write faster than us. It's just, it's going to take a while for us to train it how to think. And it's not been from a lack of trying on our end. Like, I really pushed GPT4 to try to get smarter on stuff that I wanted to do.
I jam a lot of papers into it and it's great at summarizing. It's just not a synthesizer. It's just not smart in that regard. Yeah, I'll be very interested to see if, as the cost of compute goes down, which that's a huge question mark, whether we're going to run out of computer or not.
But let's say that the cost of compute goes down, you get more people training models. If you're able to reduce the amount of data that the models need to be trained, which it seems like they will be able to now all of a sudden put it in the hands of a team like yours and thousands, maybe tens of thousands of other teams out there. I think it'll be really interesting. Even if it only gets more efficient at the level of model training, we can see some pretty big breakthroughs because now you have a lot of people training these and depending on whether they get made proprietary or open source, there's just so much benefit to humankind, obviously, if we can begin to recognize patterns.
I mean, that anything where it really seems obvious that this is a pattern recognition game, to me, it is only a matter of time for AI will be able to do it for things that aren't pure pattern recognition. Okay. And maybe some of what you're doing isn't just pattern recognition. And so there becomes a judgment thing.
I don't know, I don't know the level of data that would be required for the genome type. I Forgot how you said that to the phenotype. That translation like is like, if we were to get everybody, billions of people, would that be enough for AI? Yeah, that would be awesome.
So if you had the genome, if you had the 3 billion base pairs of every one of the 8 billion of us on this planet, and you could code the phenotype, that's what you feel to do. So how are you going to do, how are you going to quantify and code phenotypes? So that's what we do for you. Height, weight, skin tone, eye color, iq.
Like, what are the parameters we're put in there? I don't know, like we want to be able to put hundreds of thousands of parameters in there. We don't know how to quantify. What can we measure and quantify to put in there?
Vertical jump. Well, I don't know how I know your vertical jump is what it is because you hurt your knee 10 years ago versus what your genetic capability was. Right. So this is the problem.
Yeah, look heard. And that for sure is going to be a limiting factor. But when i1 is, I think about this on longer timelines, a lot of this goes away. But when I also think about, you can do this prospectively, right?
So if starting today, every kid that was born, you took their genome and you could map some quantity of their phenotype, we came up with like the 50 to 100 metrics we wanted and we just tracked everybody serially over time. I mean, sure, you could prove something really remarkable, no doubt. And then also, I think we're kicking off so much data now. If we were aggregating all that data to your point, you're not going to have everything that we ever care about, but you certainly would be able to get things like sleep data.
So this person, how do they respond to sleep, how do they respond to stress, heart rate variability? There, there are so many things that we kick off now. It's not going to, you know, I mean, it's going to be less than probably a percent of the things we ultimately care about. But, but you can really begin to zoom in on things, especially again over time.
So it's like, all right, if you're tracking what a kid's doing from whatever age 13, even if you start clocking them at 18 and you collect data on them for 20 years, including things like that, you would get in the census. So how much money do they end up making now all of a sudden? I'm not even saying this doesn't lead to something dystopian. But you'll be able to figure out a lot of these things in their genotype are coded to these outcomes or we see patterns in these.
Then it begs the question, to what end? Yeah, I will answer that question. And this comes down to I think society ends up bifurcating. This is again where we get into my personal hobby horses here.
But I think society bifurcates are going to be people that want anything to do with this and they're going to be people that break on the technological side. And now all of a sudden you are quite literally crafting superhuman people. And how do we bifurcate this society? Is it a geographic bifurc?
So yes, I think this is where I break company with Bali, who I think you know, you don't know Balaji Srinivasan. Oh my God. Two people who I think would find each other fascinating. So Balaji has a whole thesis that we're going to break into what he calls network states where you won't be geographically bound, you're going to be bound by effectively ideology and that you'll see a weakening of governments as money decentralizes.
Anyway, I don't think humans are wired like that. I think ultimately we respond most strongly to the people that we're around and I think we respond very negatively to over isolation. And it does not yet seem that the connections you get online can replicate the things you need in real life. So I have a feeling that unless we overcome that, you're not going to see that play out.
What you will see is a further segregating by geography. Now how big that geography will have to be is at city versus rural probably that seems like the sort of play throughout human history is that you see people aggregate in cities and they think about the world very differently than people that aggregate out in rural areas. So my gut instinct will be that you'll get something like the Amish. And in my mind I was clothing as puritans, people who literally use the word I am a Puritan.
My body is pure. I do not put technology in my body. I do not interface with AI. And so they'll break somewhere along the 90s effectively.
So I think about it, they'll want technology. They're not going to want to go too far backwards. They will go to the 90s and say, okay, this is a state of technology that is so useful. They'll go back from here.
They will go back from here. Just like we're going to stop, I think stop letting kids have social media before they're 16. That'll probably end up being a little bit later. Again, these are factions.
This won't be universal, but I think parents are waking up to the fact that, that it is parental malpractice to let your kid suffer the psychological consequences of having social media too early. More and more stud come out. It'll just be impossible to ignore. It'll start getting mandated.
You'll start seeing people really embrace that as literally Elon Musk continues to do brain computer interfaces. It won't be long, 15 years before you've got people that legitimately are getting upgrades. There's nothing wrong with them. They just want a brain implant because it allows them to see it infrared or whatever.
And so you start seeing people do that and you will get people that respond violently to that. This is not what God intended. It will break in very similar ways to religion. You're already seeing a geographic segregation based on right and left, which I worry about that when you start, like in the next three years.
Probably not in the next three years. Yes, like, this stuff is going to get exacerbated. So I think we have enough things with modern technology that is both outlandishly amazing and so will attract people like me and really detrimental and people are going to respond and they're going to go down that path. So anyway, I think anything extreme like this is going to push people farther and farther into those two camps.
So, yeah, I think the ultimate segregation will be geographic. But this comes down to ideology. How far people embrace AI being the most important. Interesting thesis.
It is going to get fascinating, that is for sure. All right, as you look at some of the things right now today in terms of positive, negative. So modern ills, overfeeding being, I think maybe the most profound. But I have Ozempic.
Why can I drug my way out of this? Can I pour more modernity on modernity to solve the problems? It looks like you can. I mean, it really looks like.
And remember, Ozempic is just sort of the first effective group of those drugs. So the GLP one's been around for a decade and they've been largely ineffective for weight management. In fact, they were never brought on as weight management drugs. They were brought on as drugs for improving insulin sensitivity.
But something about semaglutide as Ozempic was the first time and it was like, oh my God, I've got these people with diabetes. Their diabetes is getting better, but they're losing an unbelievable amount of weight. And so that realization happened four years ago. And now we've got a second drug that's even better than Ozempic, called Tirzepatide or Wegovy or Mounjaro.
And it's even more effective. Right? Fewer side effects, even more effective. And then we've got another one that's not approved yet, but it's in phase three called Retatrutide, that is even more potent than both of these with no more side effects.
So if you actually go down the list of all of these drugs out there, there's, you know, there's 30 more of them in the pipeline. So we have no shortage of drugs coming at us that are going to help people eat less. And now. And these drugs are also being targeted at reversing sarcopenia.
So it's not just how do I eat less, it's how I preserve more muscle. Now, the challenge is we still don't know if these drugs preserve strength. And right now, the early versions of these drugs appear to increase muscle mass, but not strength. So it doesn't appear to be a functional gain and actively increase muscle mass.
Yeah. So there's one drug called Bima that actually increases muscle mass even in the setting of using semaglutide, which decreases muscle mass all things equal, because you're losing muscle mass and fat mass. You're getting leaner, by the way. So it's decreasing fat mass at three times the rate it's decreasing muscle mass.
So you're, you're still, you're getting. Your body composition is improving. So, you know, that begs the question that you might not have a complete pharmacologic ill to training like you still have to train. It seems that at least based on what we know today, the only way to actually get stronger is to put the muscle under stress like you.
It's not enough to just take a drug that increases the size of the muscle or prevents the muscle from shrinking. That shouldn't be that surprising. It's hard to imagine that there was going to be a hack around getting stronger. But it really seems that these drugs have done something that no other class of drugs have ever done before, which is to be simultaneously reasonably safe.
I say reasonably because I don't think we know enough yet to know that they are completely safe. And I don't really know what completely safe means. I don't think there's a single drug that is completely safe, like Tylenol, Adzole, Aspirin are not completely safe. So.
So, you know, with any drug there's going to be a side effect. But when I say reasonably safe, I mean it is not. It does not appear to be the case that these drugs are causing cancer or pulmonary hypertension or, you know, some sort of reversal damage in the brain or something like that. Where we certainly had concerns with previous drugs that had really good efficacy for weight loss, like fen phen, but they were killing people eventually.
So you have this drug that's safe and effective, and that's a first. We just haven't had that in the weight space. There are a lot of things that are being touted about GLP1 agonists as like panaceas for all sorts of conditions. But we did a pretty thorough review on this, and most of the evidence that these things have benefit for, you know, heart disease, kidney disease, dementia, seem more related to the fact that you're losing weight and you're getting more metabolically healthy.
Because these drugs are not just weight loss drugs. What they're also doing to people is really increasing their insulin sensitivity. So with those two things happening, it doesn't surprise me that people are having less sleep apnea, fewer heart attacks, less kidney disease, less hypertension, you know, less cancer. I think we'll see all of those things happen.
But it begs the question, would you get those same benefits if you used sleep, nutrition and exercise to achieve the same spot? And the answer is probably. But how many people can do that without the assistance of these drugs? That says nothing of the cost of these drugs.
We can get into a much lengthier discussion about, you know, if we truly wanted to put every person who's overweight on this country on one of those drugs, well, healthcare system can't support that. And what's the trade off of the cost of doing that versus the cost of letting those chronic diseases come to rooster 20 years down the line? So there's, there's a whole bunch of really complicated economic and sort of social questions around this. But I think the more important question really comes down to what's the safety profile, what's the efficacy profile?
Wow, I did not realize that there were that many more coming out. I had a very negative view of Ozempic. I would have told people, hey, stay away from it. There are going to be consequences that we can't anticipate.
Listening to that description, maybe you heard what I wanted to hear, but it sounds like I've got something that's like a steroid that I'll have to lift because I don't get strong. And I do. I still have to put the energy. But if I've got something that's helping me get leaner while adding muscle Mass, which is the Holy Grail.
As long as I take care of making sure the muscle mass can actually contract hard lift heavy things and lower heavy things. I'm pretty excited about that. Yeah. Although I don't see how you're going to need to take one of these drugs.
I mean, you're already pretty lean. I'm probably 12% body fat, let's say. Why would you want to be leaner? Not.
Is that a joke? No. Have you seen Fight Club here? Like, I could look better, I'll tell you that.