Four numbers define progress in Longevity
Audio Brief
Show transcript
This episode explores Martin Borch Jensen's strategic framework for the longevity field, moving beyond general optimism to define progress through specific metrics and economic levers.
There are three key takeaways from this conversation. First, the field must solve the longevity production function by increasing the rate of new ideas and shortening trial times. Second, unlocking major capital requires a clear demonstration event to prove that aging interventions are possible. And third, the role of AI in biology is currently limited by a lack of causal training data, which must be addressed to accelerate discovery.
To understand the longevity production function, consider four critical numbers. These are the output of healthy years added, the input rate of new therapeutic ideas, the probability of success, and the time lag for trials. Currently, the input is dangerously low, with only a handful of viable ideas per year. The time lag is decades long, and the perceived probability of success is near zero because no longevity drug has ever been approved.
This low probability of success creates a financing bottleneck. The longevity field currently receives less than one percent of biomedical funding because investors see the risk as too high. To unlock capital, the industry needs a "demonstration event." This means prioritizing short-cut clinical trials that focus on rapidly aging systems, such as ovarian function or lifespan in large dog breeds. A clear win in a shorter timeframe would prove intervention is possible and drastically shift investor sentiment.
Finally, while artificial intelligence has revolutionized molecular design, it struggles to predict biological outcomes. We face a data problem. We lack the training data to answer what happens to a human cell when a specific intervention is applied. To utilize AI effectively, the field must stop accumulating observational data and start generating foundational perturbation data. This involves mapping how biological systems respond to specific interventions, revealing the biological dark matter that current models miss.
Solving these structural and data challenges is the only path to moving longevity research from a niche field to a global imperative.
Episode Overview
- Martin Borch Jensen outlines a strategic framework for the longevity field, moving beyond general optimism to define progress through specific, measurable metrics.
- The discussion analyzes the current "production function" of aging research, identifying why the field currently adds only ~0.1 years of life annually and how to accelerate this.
- This episode connects the scientific bottlenecks of aging research—such as long feedback loops and lack of training data—with economic levers like funding dynamics and the role of artificial intelligence.
Key Concepts
- The Longevity Production Function: Progress in aging is defined by four numbers: the output (healthy years added), the input (rate of new therapeutic ideas), the probability of success, and the time lag for trials. Currently, the input is low (2-5 ideas/year), the lag is long (decades), and the perceived probability of success is near zero because no longevity drug has ever been approved.
- Financing Follows Faith: The longevity field receives less than 1% of biomedical funding because investors view the probability of success as effectively zero. To unlock capital, the field needs a "demonstration event"—a clear win in a shorter timeframe, such as extending lifespan in large dogs or delaying ovarian aging in humans—to prove that intervention is possible.
- The AI "Dark Matter" Problem: While AI has revolutionized molecular design (e.g., AlphaFold), it struggles to predict biological outcomes. We lack the training data to answer "what happens to a human cell if we do X?" To utilize cheap intelligence effectively, the field must generate massive datasets mapping cellular state transitions, revealing the biological "dark matter" that current models miss.
- Biomarkers as Lag Reduction: The biggest hurdle in longevity is the time required to prove efficacy (waiting for people to die). Validated biomarkers and "aging clocks" are essential to compress this timeline. However, current clocks are merely correlative; they need to be benchmarked against interventions known to extend life to prove they are causal measurements of biological age.
Quotes
- At 0:46 - "When I got to the National Institute on Aging, I learned that there was no plan. And there was no intention of making a plan." - Revealing the strategic vacuum in the institutional approach to aging that motivates the need for a defined roadmap.
- At 8:09 - "The field is so small that you can make a big difference individually... One very talented person [can add] 0.1 to 1% more output." - Highlighting the high leverage available to talented researchers or founders entering the nascent longevity space compared to established fields.
- At 13:42 - "If we don't have any idea of what to do, we cannot do anything... having things that we can put into that system is rate limiting." - Identifying that the primary bottleneck isn't just testing capacity, but the scarcity of high-quality therapeutic hypotheses at the top of the funnel.
- At 20:32 - "If these tests really told you your biological age, then we would all be crazy for not taking them all the time... your insurance company would probably force you to take them." - Illustrating the skepticism surrounding current commercial aging clocks and the gap between correlation and actionable medical validity.
Takeaways
- Prioritize "short-cut" clinical trials that focus on rapidly aging systems, such as ovarian function or lifespan in large dog breeds, to demonstrate proof-of-concept quickly.
- Invest in generating foundational "perturbation data"—datasets that show how biological systems respond to specific interventions—rather than just accumulating more observational data, to make AI models useful for hypothesis generation.
- Validate aging clocks by running experiments where animals are treated with interventions known to extend life, ensuring the clock reflects the biological reality rather than just chronological time.
