The current state of the longevity industry is both worse and better than what the public narrative suggests. The problems are harder, the state of our knowledge more limited, and too much money and attention is thrown into too many dead ends. But at the same time our understanding of the difficulties is sharper than ever and we have developed a new set of extraordinary tools.
If the situation is bad, the options have never been better.
Four bad news
Almost nothing works and nothing works very well
The baseline of the longevity industry is that we have no breakthrough intervention with a proven, significant impact on either healthspan or lifespan beyond known commonsense advice (moderate exercise, enough sleep) and taking advantage of standard medical advice and resources. In a way this is a trick observation: every time we develop something like that it becomes part of the standard. What's more significant is that there's nothing like that in the research pipeline, much less clinical trials. We can do impressive things with model animals, little of which translates well to humans.
That's compounded by a second problem:
Basic research has been devastated
The most influential actor in basic longevity research in the world is the US National Institutes of Health and the large number of public and private organizations that collaborate with them or are downstream of their research. As expensive and risky as drug discovery can be, it would be entirely impossible as an industry without, to a large degree, the US government subsidizing the long, even riskier, entirely unmonetizable basic research that has to happen before a pharmaceutical company can even begin. For various and horrible reasons the current US administration has caused huge damage to its funding, organizational capabilities, and human resources, which even in the best case will take a long time to reverse.
But we have AIs now, right?
LLMs are a dead end for longevity tech research
Some days it can look like half of the Internet consists of discussions about what ChatGPT or Claude can or cannot do. To keep this manageable I'll just point that the people who actually do biomedical research — as opposed to churning out filler papers, managing AI companies, or talking about the future of AI — haven't found those AIs very useful. That's because you can't solve problems in biomedical research by searching and summarizing the existing literature: the truth is not out there online, it's in there, in our bodies. What limits research speed is the quantity and quality of experiments, not the speed with which we can search for papers or draft new ones.
The last problem is related to this one, but subtler because it's cultural and political rather than just technical.
The longevity tech narrative is being hijacked
There's always been an uncomfortable split in the longevity field, with the bulk of advances coming from basic research but attention focused on often overhyped practices and products. What's different now is that attention and, increasingly, investment is being captured by a different set of actors: high-profile companies or individuals already wealthy from other businesses — think "tech billionaire" — ignorant or skeptical of the difference between biomedical research and software or finance, and with the political and media power to be taken at face value.
The short-term problem is that this drives money and attention to dead ends. The long-term problem is that the sometimes abhorrent ethics and politics of some of these actors risk tainting longevity tech as a field for the long term.
So much for the bad news.
Three good options
A cognitive boost to the market
The most obvious opportunity in the longevity tech market is the need to make it clearer: a shared and evolving understanding of what is possible today and what isn't, what it means to say that something works, what are the unknowns, limitations, interactions, and redundancies. A longevity tech market that really works for consumers is the only one that can grow beyond periodic boom and bust cycles, and the unavoidable complexity at the intersection of biology and biotechnology calls for radically smarter ways to communicate options, assist decisions, and platform processes in order to make the best of our existing tools and knowledge.
Increasing the limits of manageable complexity is absolutely necessary, because future developments will come much faster.
A cognitive boost to research
The gap between the astronomical complexity of human biology and the limits of what we already know means we can't use AIs to fill in the blanks. But there is an inefficiency we can use AIs of other types to help us with: too many experiments in longevity tech are designed and run in ways that minimize or eliminate any chance of making significant advances from them, partly because academic and industry incentives mean publishing a paper or a press release that can be spun into a positive headline is seen as more important than moving towards more effective tools, partly because good experiment design is not easy, specially for something as complex and stubborn as aging. There's no silver bullet for research, but organizations that apply new AI-level tools to design and analyze experiments have a speed advantage that in the context of longevity tech is the difference between some movement and none at all.
Reframing longevity tech as the necessary dual of AI
If the bad news is that lots of capital that should ideally be allocated to longevity tech is going into forms of AI that are irrelevant to it, the good news is that this means lots of capital is being allocated to the pursuit of the perceived as impossible. For good and for ill we are at a point in history where deep strangeness and existential stakes do not deter but spur vast investments and frantic development. One way or another the current dynamics of the AI industry are unsustainable. Whatever players are left afterward the generational lesson is that cultural drivers of economic ambition have moved from the sci-fi product to the sci-fi scenario. In that sense few other industries are as well-positioned to take the lead as longevity tech — if it can become mature enough to be clear about its complexities, pragmatic about its ambition, and ruthlessly ethical as a matter of self-preservation.