The a16z Show - a16z Podcast: Defeating Aging with Aubrey de Grey

Episode Date: November 19, 2015

There are those who would say that Aubrey de Grey is out to cure death, but what this former artificial intelligence specialist turned gerontologist is really focused on is health -- and the side effect of health is living a lot longer. In this segment of the a16z podcast we talk with Aubrey de Grey on the subject of aging and health, and how his training as a computer scientist helped him approach the problem in a different way from traditional biologists. The intersection of software and biology, and how this “troublemaker” from the computer science world is trying to keep us all healthy for a very, very long time. Stay Updated:Find a16z on YouTube: YouTubeFind a16z on XFind a16z on LinkedInListen to the a16z Show on SpotifyListen to the a16z Show on Apple PodcastsFollow our host: https://twitter.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to the A16Z podcast. I'm Michael Copeland. There are those who would say that Aubrey de Grey is out to cure death. But with this former artificial intelligence specialist turned gerontologist is really focused on is health. And the side effect of health is living a lot longer. In this segment of the A16Z podcast, we talk with Aubrey de Grey on the subject of aging and health and how his training as a computer scientist helped him approach the problem in a different way from traditional biologists. The intersection of software and biology and how this troublemaker from the computer science world is trying to keep us all healthy for a very, very long time. Aubrey, welcome.

Thanks very much for having me. We want to talk about the intersection of computer science and software and biology. And I think there's a misconception out there that you're out to cure death, but you shifted from software verification to this idea of really health, but extending life. And How did you make that transition and why? Well, first of all, thank you for putting in that clarification, because it's normally something that I need to put in in interviews, that yes, I'm just a real medical researcher. I work on health, and any longevity benefits that may arise from this work

are a fight effect of health, because, of course, being sick is what kills people. Yes. So, and of course, journalists, you know, tend often to sensationalize that. They think it sells more papers if they talk about immortality and other stuff like that. But how did I make the transition? Well, it was quite interesting, actually, and a lot of luck was involved. Essentially, in 2000, when I was in the middle of my work in software verification, I met, and shortly afterwards married, a biologist from the US who was on sabbatical in the UK.

She was a full professor at UC San Diego. She's quite a lot older than me. And through her, I, first of all, learned a lot of biosephableness. biology over the dinner table, as one does. Right. But also, after a couple of years, it began very gradually to dawn on me that we were never talking about aging, which was really quite bizarre, I thought, because, you know, I'd always gone through my whole life, assuming that everybody understood that aging was the world's

most important problem, the source of the world's greatest amount of suffering and so on. But it turned out that my wife and indeed all the other biologists I was meeting were actually of a very different persuasion. They thought that aging was not very interesting and not very important. And I was absolutely appalled, and it actually took me another year or two before I really came to terms with it. But eventually I decided that even though I was already working in artificial intelligence research, which I viewed as a humanitarian exercise. Making machines smart enough to relieve us of the tedium of having to spend our lives doing things that we wouldn't do if we weren't being paid in order to make the work around. Yeah, so I wanted that to end.

but I always knew that that was only the world's second most important problem. And so after I kind of got over the shock of finding that most people didn't think that way, which I'd never dawned on me until the age of 30 or whatever, yeah, I switched fields. And I was in a very fortunate position there. I was working in the University of Cambridge on a bioinformatics project that was a nice way of combining my formal training in computer science with my newfound informal training in biology.

And that project was very undemanding. It left me a lot of spare time, and of course access to all the university facilities and libraries and so on, and paid the bills. So I was able essentially to do research in my spare time. In fact, the reason I originally took the job was so as to resume my artificial intelligence research, which I had to put on hold for a year or two on account of lack of funding. But when I decided to switch fields, of course, it was just switching what I did in my spare time. a time, so it wasn't something risky. And did your AI background and your software background, did that bleed over into longevity?

And how did those two things match up, if at all? There was a lot of overlap there. It was actually one of the big reasons why I decided to switch fields, but why I decided that I had a respectable chance of making a significant contribution in biology. I realized that a lot of the reason why people... were not making progress in postponing the ill health of old age was because they were going about it more as, you know, lifelong basic scientists and not so much as technologists with a more goal-directed kind of way of thinking. I felt that there was a good chance that I would be able to

bring in new ways of addressing these issues and thinking about them and maybe come up with ideas that would actually be promising. And sure enough, that's how it turned out. And how did you, I mean, There's sort of a religious, I don't know, I want to say, but like, you know, chemistry versus computation. Like the way forward is chemistry, how dare you think that, you know, just compute can solve our problems or figure it out. But were you accepted into the biology world, or were you seen as somewhat of a heretic?

I actually was accepted pretty well pretty quickly. In the first five or so years that I was working as a gerontologist, so basically the second half of the 90s, I was publishing stuff that was relatively harmless. It was like new explanations for other people's data that interpreted them better. And this was actually very well received. I was able to gain quite a reputation for myself. And in fact, the fact that I didn't have the regular traditional experimental training

in how to work a pet and so on, well, somewhat in my favour. You know, people thought, well, this guy, you know, he's come from nowhere and he's having these ideas that we ought to have had really. And, you know, he must be very smart. So that was, so I rose to a level of, you know, general acceptance and recognition pretty quickly, especially since I was taking the trouble to integrate myself a lot, going to conferences on my own hook and so on. Right. It was only after the year 2000 when I started putting forward my particular ideas for how we might do something about aging, that people started to think of me as more of a troublemaker.

and in some cases use my lack of experimental training against me, as a way of trying to imply that because I didn't know how to work a pet, therefore, or everything I said was nonsense. Well, let's get to your ideas on how to extend life and what's caused people to, you know, some people think it's completely off base. It's Hocum and others, although others have tried to disprove it and really can't.

So first of all, let me point out that you are backsliding a little bit by talking about my efforts to extend life. You've got to remember it's a side effect. Yes, sorry. Your efforts to make us healthy. That's right. To keep us healthy. Keep us healthy, yeah.

All right. So, well, right, yes. It was very controversial at first. And, of course, this was kind of no surprise to me, because when I had this big kind of eureka moment in 2000 of how to go about this, the realization was a radical departure from what people had been thinking before. Essentially, gerontology for decades had been built on the, concept of trying to work with the variability in rate of aging that we see in nature.

The fact that some species age a lot more slowly than others, even within a species, some individuals age a fair bit more slowly than others. If we could understand that phenomenon really well, then maybe we could translate that understanding into something therapeutic. And it hadn't worked, of course. Nobody was really getting anywhere. And indeed, until the 1990s, pretty much everyone had even given. up on trying to admit that that was the ultimate goal of gerontology.

There were some breakthroughs in the late 80s and early 90s, which kind of changed that, but it was a bit of a force-dorn, actually. And I came along, and basically my new idea, in very simple terms, was just that rather than trying to clean up metabolism and thereby slow down the rate at which the body creates damage to itself as a result of its normal operation, rather than doing that, the idea was to actually repair damage, so to go in one step down the road, so to speak, to do periodic repair,

not necessarily completely 100% comprehensive, but fairly comprehensive, so as to maintain a level of damage in the body that it was within the tolerance of the body, because the body set up to tolerate a certain amount, right? And that's why nothing really goes wrong until late middle age. So this idea only for, flu, it only made sense because I was able to identify other areas of biology that gave rise to

practical options for really implementing this, for doing the damage repair. And these other areas were areas that gerontologists didn't know anything about. They'd never come across to them. When I started talking about them, they thought that I was talking about stuff they didn't need to know about. So it was all very difficult. What were some of those areas? Well, I mean, there was one example, one big example that is a component of SENS is not even from anything medical at all.

SENS is actually your nonprofit, but also an acronym for... Yeah, it stands for strategies for engineered negligible senescence, but you really don't need to know that. Well, no, you do. But, yeah, I mean, so one example is not even from anything biomedical. It's from environmental decontamination, the idea of using bacteria as a source of genes and enzymes that can break down material. in the body that we don't have any enzymes to break down, thereby, of course, eliminating this stuff and stopping it from accumulating to an eventually toxic level. You know, nobody heard of that.

People were pretty intrigued by it, but they didn't really think it was something that made much sense in medical terms, or at least some of them didn't. Of course, in any field, there's a range of degrees of dogmatism, and some of my more vocal detractors certainly just push, who pooed pretty much everything they didn't, hadn't already thought of. But in about 2005, I was able to kind of smoke out the opposition, so to speak, which had previously been happening at a level of kind of off-the-record ridicule. And there were a couple of major running battles that I pretty comprehensively won

in terms of demonstrating that the ideas that I was putting forward were indeed very plausible and that the conclusion that they were not plausible that other people had been expressing was essentially a result of ignorance. At the core, it's this idea that you can repair the damage that's been done. Can repair? Basically all of it, yeah.

I mean, the idea is you don't need to repair all of it, but you do need to repair most of it. The damage comes in a variety of different types, different categories, and within each of the categories, there may be a lot of examples some of these things may add up, but basically any of the categories can kill you on its own. How do we go about doing this repair?

Well, of course, the different categories have very different approaches. So one example that's very familiar to everybody is the way of repairing the type of damage that I call simply cell loss. So cell loss is just cells dying and not being automatically replaced by the division of other cells. And various aspects of aging are predominantly driven by that. Parkinson's disease is a fine example where there's a particular part of the brain in which neurons of a particular type tend to diet an unusually rapid rate and that eventually stops that part of the brain from working right. The fix is, of course, stem cell therapy. You put cells in that can know how to divide and differentiate to replace the cells that the body is not replacing on its own. Right, become the cell that you need.

That's right. And stem cell therapy for Parkinson's disease is a very viable concept. It's actually, there are a couple of clinical trials going on right now that people are very optimistic about. So, yeah, and that's the kind of idea. But a lot of the other things I suggested were much further afield from what gerontologists had even heard about in the news, let alone in conferences. Things like, for example, well, as I mentioned, finding bacteria that can break down substances like oxidized cholesterol that drives arthritis. sclerosis. The idea would be then you'd find the bacteria, you'd then identify the genes and enzymes that they had that allowed them to break this stuff down, and then you'd modify those genes so that you could put them into human cells and protect the human cells by giving them this augmented garbage disposal capacity. It took us a long time to get it to work. I put the idea forward for the first time back in 1999. We started working on it in about 2005, and it was only 2012 that we were. were able to demonstrate a really powerful proof of concept, showing that we could actually

protect cells in cell culture from otherwise lethal amounts of this particular toxin. So that's why it was hard battle credibility-wise. But things are getting better all the time in a large number of these areas just because we are getting to proofs of concept. So the way you have gone about this, like you say that people didn't hear about this in conferences. far a field that it just didn't occur to him. And somehow it occurred to you. But what that sounds like a little bit is the internet. And it sounds like, and by that I mean this ability to

move information, you know, from one place to another and that everybody has access to it. So in healthcare, for example, just the sheer ability to look across a huge data set of people and patients and outcomes and say, oh, that's what happened. Do you think that, that, you think that what you did personally as this kind of, you know, human version of the internet, does that accelerate? I mean, so now as more data and more information is kind of more widely available, and ideally as silos are broken down more, and maybe they're not being broken down fast enough, do we accelerate this sort of ideas and the approaches that you have put forth? Interesting question. I would say yes and no. I mean, certainly the availability of information

online is an enormous opportunity for people to come out of left field and have new ideas. But then, you know, in most sciences and in fact in most fields of technology, one has to be pretty knowledgeable already in order to actually have new ideas that are not completely broken. You know, I would not have been able to come up with sense. And if it hadn't been for spending those initial five years in the field, just learning, just going around and listening a lot and thinking. And, you know, the silo question comes later, really.

The silo question, the silo problem arises really when we look at the ways in which ideas are taken, in which big ideas are taken forward. Because, of course, getting something to actually work once you've decided what you're trying to get working, it involves a long string of solving little problems. and it takes time and money. And the biggest obstacle, really, in science to getting stuff done like that is the fact that the overwhelming majority of science is funded by peer review. Peer review is an absolute catastrophe when it comes to doing anything high risk, high reward.

And also, for that matter, doing anything cross-literciplinary. Because people tend to play it safe and or get, and they're worried about getting attacked essentially? Well, it's kind of. I mean, basically, peer review is an us covering process, right? It's both the people who actually are providing the money and the people who are selecting who to get the money out of the many, out of the far too many applications. You know, they need to protect their reputation somehow. And it's just too easy to be cautious and to favor incremental stuff that is within the remit of what people, what the applicant has already done, you know, very close to it.

So it's an enormous stifling influence on research. And it's particularly bad these days when funding is so short, when there's so much, you know, when the payline, the proportion of grants that are actually funded is so low. Again, coming from the AI and software world and working on those kinds of problems, how do you compare that to working on the problem of sort of humans and health? And the subtleties of the human body and biology versus the subtleties of, you know, clearly AI is a very hard problem too.

But for those folks in the technology world who are increasingly going to cross over into the health side of things, what are the things that they ought to keep in mind and what's so hard about it, honestly? Well, right. So, I mean, I think the big thing that a technologist, an engineer of any kind has as a starting point is, the understanding that any machine, and of course the human body is just a machine, right, any machine has moving parts, it does damage to itself as a side effect of its normal operation, so one can use the same principles, the same like top-level principles, to postpone the ill health of old age as one might use to, you know,

keep a car going longer than it was designed to go. That's the easy part. then the hard part is that because the human body is such an astronomically complicated machine and because we understand it so poorly, one has to be quite ingenious in identifying approaches to extending its healthy lifespan that are likely to work despite our ignorance. In other words, to essentially leave well alone as much as possible and only interfere with things that are unlikely to have unwanted side effects. And how do you run those as kind of, I don't know, theories or lines of code or lines of thought?

I mean, how do you kind of get to the right answers? Well, I mean, probably the biggest thing that, the biggest component of the sense concept that allowed the whole thing to fly originally was my realization that we have this window of opportunity afforded by the fact that the body is set up to tolerate a certain amount of these various types of damage. Right. which means that we can infer that these various types of damage are inert until such time as they've accumulated to a certain threshold of abundance. Now, inert means they're not participating in metabolism. So if we just target those initially inert phenomena, then we have a good chance of not having unwanted side effects,

not disrupting the, you know, labyrinthine, network of processes that keeps us alive. Where are we now in that, and kind of your continuum? Like, you know, I know that the therapies sort of aren't ready yet, but how far along are we? And as a side effect, you know, in your mind, how long can people live? So the, how far along we are in developing these things, of course, the different types of damage, there's a different answer to that question. So in the case of stem cell therapy, of course, there are quite a number of stem

self-therapies already in clinical trials. I mentioned the case of Parkinson's disease. There are plenty of other, of course, aspects of ill health that don't have to do with aging that are also amenable to clinical trials using stem cells. Most of the other things are a good deal less far advanced. Some of them are partly advanced. So, for example, the elimination of amyloid, which is a kind of molecular waste product that accumulates outside in the spaces between cells.

in some cases, well in one case, Alzheimer's disease, that's also very much in clinical trials and we basically solve the problem. That can now be eliminated. It doesn't have much effect on Alzheimer's disease on its own, but it's fair bet that in combination with other therapies that will be developed in the future that fix the other aspects of Alzheimer's that it's going to be very useful. There are other amyloids, though, in other tissues that accumulate and cause other problems in aging, and we haven't made much progress in those areas, so that's a lot of.

actually one area that we're funding precisely for that reason. You know, since Research Foundation exists and indeed it was constituted as a charity, specifically because not a lot is happening in a lot of these areas. We're being neglected far too much, and somebody needed to step in and actually kick them along the road and get them to a sufficient level of proof of concept that other people would get interested, and we've been very successful in doing that. Are you, in some sense, in competition with the Ray Kurzweil's singularity view of the world? I mean, we're either going to become, you know, machines, you know, that have the, I don't know, brains of a human or sentience of a human, or we're going to, as a side effect, live a lot longer and therefore we don't need to have the singularity.

So I wouldn't call it competition. It's more of a race. You know, Ray and I are, we know each other well, of course, and Ray is very much interested in regular biomedical approaches as well. So when he talks about how to live long enough to live forever, it's the phrase that he likes, he talks about these bridges. There's things that you can do today that will postpone the ill health of old age somewhat. And he is actually a lot more optimistic than me with regard to how much we can postpone aging with stuff that already exist today. But then Bridge 2, as he calls it, is almost identically sense. It's basically using high-tech biotechnology to repair damage. and Bridge 3 is the one that you're really referring to,

the increasing use of what we might just in general call non-biological solutions to medical problems, especially focusing on the more miniaturized stuff like nanotechnology and then eventually perhaps even on transferring consciousness to a different substrate, the concept of uploading. So the reason I call it a race rather than a competition is because we just don't know what's going to actually prove to be implementable sooner. And do you care, I mean, if, you know, damages repaired biologically or with these micro machines? Well, I kind of care. You know, I'm quite sentimental about being made out of meat.

But at the same time, you know, if push came to shove and the work that we do and other people do on the biotech side, started to, you know, hit diminishing returns and basically run into the sand and work on uploading or in other ways of reinforcing the, the health of the individual through non-biological means actually moved forward relatively rapidly and ended up being the solution that got there first, then that's fine with me. Yeah. It's the end result that matters, I guess. So let's say whichever one works, we're going to have a lot more people living a lot longer. Rents are high enough in the Bay Area, not to mention just food or climate change.

How do we account for all these people living for so long? Well, so the concept that if we defeated aging, we would have a terrible problem of other population is probably the number one knee-jerk concern that people raise. And it's so insidious and so persistent that we eventually resorted to the option of actually funding a forecasting group in Denver that have over the past 30 years developed. very well-regarded system called international futures. We actually funded them to extend the versatility of their system so that it could explore the concept of sense,

the concept of actual rejuvenation biotechnologies that would restore the health of people who are already in middle age and keep it there. And, of course, we knew what the answer was going to be, more or less, namely that the consequences of that for the trajectory of world population or indeed of populations of regions was actually much more modest,

much less frightening than people would normally think. Plus also, of course, we're interested in the solutions. So we've got an overpopulation problem today. But the problem is not that we have 7 billion people. The problem is that we have 7 billion people who are all creating a lot of pollution because of fossil fuels and such like. So, of course, the solution to that

could be have fewer children or don't cure aging, but it could also be invent new technologies that increase the carrying capacity of the planet. And that's, of course, exactly what we're doing. We're having a burgeoning of renewable energy. Quite soon we'll have nuclear fusion one way or another in terms of agriculture as well. You know, we obviously need a lot of land right now

to create enough food for everybody, but that's changing with the development of artificial meat and so on. So it seems to me pretty damn clear that the increase in the carrying capacity, planet over the next century will way outrun the increase in the actual population of the planet. Do you feel like we're at a point in time where technology and kind of the things that you're studying, whether directly or related, will kind of speed things up? It seems to me that, you know, when we look at transportation, we look at lots of things that, that at this point, software is

helping us to really accelerate some of the things that we want to do. Do you, in the near-term hope to see or think that you'll see advances across the board in health care? Well, we're certainly seeing that already. I mean, of course, a lot of this involves enabling technology at the level of informatics and also at the level of simple hardware. You know, so we've got better techniques for sequencing now. We've got better techniques for modification of the genome, you know, with things like CRISPR, for example.

And, of course, we've also constantly got new advances in computation. interpretation of what we know about the genome and the epigenome and the microbiome and so on, what these things are doing. So yes, I think that there's definitely a very heartening and accelerating increase in our ability to maintain health and maybe restore health as a result of all of this. But I think we've always got to remember that things like the ability to sequence things really fast and really cheaply, or the ability to process things, usually doesn't actually underpin the fundamental breakthroughs. Right, there's all this new data, but then what the hell do we do with it? Right.

What it does instead is it makes things easier and faster. Having the genome, even the original one genome, you know, Craig Venter's genome, has definitely facilitated a lot of research, but on its own it doesn't cure things. Right. Well, Aubrey de Grey, I wish you health and much of it, and for many, many years. Thanks for joining the A16Z podcast.

My pleasure. Thanks for having me.