A mini-rant abut AI and longevity. (This is a slightly revised version of a post I shared on X today.)
They say "Artificial Superintelligence would take only a few years to cure cancer, solve longevity, and defeat death itself'.
This is a common claim by pro-AI lobbyists, accelerationists, and naive tech-fetishists. But the claim makes no sense.
The recent success of LLMs does NOT suggest that ASIs could easily cure diseases or solve longevity, for at least two reasons.
1) The data problem. Generative AI for art, music, and language succeeded mostly because AI companies could steal billions of examples of art, music, and language from the internet, to build their base models. They weren't just trained on academic papers _about_ art, music, and language. They were trained on real _examples_ of art, music, and language.
There are no analogous biomedical data sets with billions of data points that would allow accurate modelling of every biochemical detail of human physiology, disease, and aging. ASIs can't just read academic papers about human biology to solve longevity. They'd need direct access to vast quantities of biomedical data that simply don't exist in any easy-to-access forms. And they'd need very detailed, reliable, validated data about a wide range of people across different ages, sexes, ethnicities, genotypes, and medical conditions. Moreover, medical privacy laws would make it extremely difficult and wildly unethical to collect such a vast data set from real humans about every molecular-level detail of their bodies.
2) The feedback problem. LLMs also work well because the AI companies could refine their output with additional feedback from human brains (through Reinforcement Learning from Human Feedback, RLHF).
But there is nothing analogous to that for modeling human bodies, biochemistry, and disease processes. There are no known methods of Reinforcement Learning from Physiological Feedback. And the physiological feedback would have to be long-term, over spans of years to decades, taking into account thousands of possible side-effects for any given intervention. There's no way to rush animal and human clinical trials -- however clever ASI might become at 'drug discovery'.
More generally, there would be no fast feedback loops from users about model performance. GenAI and LLMs succeeded partly because developers within companies, and customers outside companies, could give very fast feedback about how well the models were functioning. They could just look at the output (images, songs, text), and then tweak, refine, test, and interpret models very quickly, based on how good they were at generating art, music, and language. In biomedical research, there would be no fast feedback loops from human bodies about how well ASI-suggested interventions are actually affecting human bodies, over the long term, across different lifestyles, including all the tradeoffs and side-effects.
It's interesting that most of the people arguing that 'ASI would cure all diseases and aging' are young tech bros who know a lot about computers, but almost nothing about organic chemistry, human genomics, biomedical research, drug discovery, clinical trials, the evolutionary biology of senescence, evolutionary medicine, medical ethics, or the decades of frustrations and failures in longevity research. They think that 'fixing the human body' would be as simple as debugging a few thousand lines of code.
Look, I'm all for curing diseases and promoting longevity. If we took the hundreds of billions of dollars per year that are currently spent on trying to build ASI, and we devoted that money instead to longevity research, that would increase the amount of funding in the longevity space by at least 100-fold. And we'd probably solve longevity much faster by targeting it directly than by trying to summon ASI as a magical cure-all. ASIs has some potential benefits (and many grievous risks and downsides).
But it's totally irresponsible of pro-AI lobbyists to argue that ASIs could magically & quickly cure all human diseases, or solve longevity, or end death. And it's totally irresponsible of them to claim that anyone opposed to ASI development is 'pro-death'.
Let's break this into two questions:
1. After a few years of ASI, will the ASI be able to stop or reverse aging?
2. After a few years of ASI, will hardly anyone die of aging related diseases?
Let's tackle number one first. It's true the ASI would not be able to do long term human trials the regular way. However, I think it could learn a lot from the data from running trillions of lab-on-a-chip experiments. I think it could develop nano bots that could remove cancer cells and repair aging related damage. And it could get quick feedback by making C. elegans, etc immortal. It might also be able to simulate biology from first principles in order to run the equivalent of decades long human trials.
I also think it could develop noninvasive (or at least non-destructive) scanning techniques that would allow someone's consciousness to be simulated. And even if that doesn't count, it might even be able to build up a new biological human that has equivalent consciousness to the original (which still may not count depending on one's values). There are likely many other routes to quick longevity that I can't think of but an ASI could.
As for the second question, would people allow the, e.g., repair nano bots into their bodies? One subquestion is whether countries would allow it. Based on current laws, probably not, though it's possible they would change quickly due to ASI (and people could go into international waters). Another subquestion is if it is legal, would people do it? Obviously some people would not, but if the alternative is a soon death, I think many people would.
An option to circumvent current laws would be the person dying, and then being fixed by the tech and revived. This could be thought of as like "fast cryonics."
I'm extremely skeptical that it would be possible to 'simulate biology from first principles' in any computationally feasible way, given the hierarchical complexity of biology across huge range of scales -- both spatial (from biomolecules to organelles to cells, tissues, organs, and organism) and temporal (from femtoseconds to decades). We just don't have any 'first principles' in biology that are analogous to physical laws that could be used to simulate planet formation or weather.
We also don't have the data required to 'run trillions of lab-on-a-chip experiments'.
And I cannot imagine any situations in which 'nano bots that could remove cancer cells' could be deployed in living humans without the first several thousands patients dying in surprising and gruesome ways.
I just don't see any of these suggestions connecting to the reality of human biomedical research.
It could simulate individual atoms, and work up from there. Yes, that is prohibitively computationally expensive now, but the ASI could scale up computation many OOMs and could probably model it much more efficiently than we could. It probably wouldn't need to model all the atoms because of sub modeling the different scales you note (like we do with weather). I'm not confident that it could do this in a few years, but I think it's generally a risky bet to say that ASI can't figure out something quickly.
I was saying it could gather data with those experiments, e.g. testing out many different nanobots/drugs on many different types of tissue. Why would it need to have data in order to run the experiments?
It could experiment on living cancer tissue in vitro. It looks like transgenic Zebrafish get cancer in 2-4 weeks. And as I commented below, it could wait until after death to try to fix - more than 20k people donate their body to science a year.
You do realize, I hope, that this all sounds wildly speculative to anyone who works in biomedical research?
It builds assumption on top of assumption.
Basically you're saying 'trust us, ASI can do ANYTHING it needs to do to gather ALL the data it needs, by any means necessary, to solve all diseases quickly, reliably, with no side-effects, no tradeoffs, and no catastrophic tragedies that would turn public opinion against the whole enterprise'.
That is not a compelling argument to me at all, and I think its implausibility undercuts the common talking point among e/accs and pro-AI lobbyists that 'ASI would cure death quickly and easily'
Well, not Aubrey de Grey. :) But seriously, let's say that one asked biomedical researchers, "Imagine a scenario where you had billions of researchers much more capable than the best scientists who ever lived thinking for centuries of subjective time and running trillions of in vitro experiments and billions of in vivo experiments on small animals and could create nano bots (e.g. white blood cells) and could experiment on thousands of recently deceased people, do you think they could solve aging?" I would be interested in the percentage of them who would describe this as wildly speculative.
To be clear, I disagree with high confidence that ASI would cure death quickly and easily, especially if that means death is actually cured, rather than we have a cure available. Indeed, catastrophic tragedies could turn public opinion against the whole enterprise. And I'm not claiming there would be no trade-offs, especially because many people say now they don't want to live forever. I'm also not claiming no side effects, but that the alternative of dying would be worse. I think we should pause at AGI because ASI would be dangerous. But if ASI were aligned, I do think it is plausible that it could quickly develop a cure for aging.