How the Next Decade Could Add 50 Years to Your Life — Dr. Derya Unutmaz
Immunologist and aging researcher Dr. Derya Unutmaz argues that exponential advances in artificial intelligence could soon let us cure most diseases and even reverse aging. His core message: stay healthy for the next 10–15 years and you may reach 'longevity escape velocity,' where each year you live adds more than a year to your life.
Overview
In this wide-ranging FoundMyFitness conversation, immunologist and aging researcher Dr. Derya Unutmaz explains why he believes the coming decade may be the most consequential in human history. He describes how reasoning models are already compressing months of biological data analysis into minutes and helping design better experiments.
A central idea is the 'digital twin' — a data-rich simulation of a person that could shorten clinical trials from years to weeks and enable truly personalized treatment on demand. He distinguishes artificial general intelligence from superintelligence, outlining the missing pieces: persistent memory, self-learning, and physical intelligence. On cancer, he explains why it is so hard to treat and why immunotherapy, CAR-T, and personalized mRNA vaccines point toward near-total cures.
He is strikingly optimistic, arguing the only real existential threat is human misuse, not the technology itself. On aging, he frames the body as a self-repairing system that gradually loses information, and discusses prevention, cellular reprogramming, and the special challenge of the brain. He closes with practical guidance on building a personal 'mini digital twin' from everyday biomarker data.
Key quotes
5This is probably the most critical time in human history. So try not to die for the next 10 years.
We will come to a point where every year you live is going to add more than a year to your life.
There is only one existential threat to humanity, and that is humanity.
Cancer is going to be 100% curable, probably in less than a decade.
Every day counts now, in my opinion.
Key ideas
8Exponential, not linear, progress
Our minds expect the next decade to resemble the last, but Unutmaz argues the next 10 years may exceed the last century. He frames 'longevity escape velocity,' coined by Aubrey de Gray, as the point where every year lived adds more than a year to life.
Months of analysis compressed to minutes
Reasoning models can now analyze millions of biological data points and return not just results but genuine insight. He describes a 112-minute run that produced a 40-page report on an immunological dataset, including which questions to ask next.
The digital twin
A data-rich simulation combining genome, metabolism, immune system, and microbiome could predict a drug's effect on an individual. This could shorten clinical trials from years to weeks and move toward treatment on demand.
General intelligence vs superintelligence
He defines artificial general intelligence as the ability to transfer knowledge across domains, and superintelligence as the combined capability of all humanity. The missing levels are persistent memory, self-learning, and physical intelligence.
The only threat is humanity
Rather than fearing the technology, Unutmaz argues that throughout history human intelligence has caused the most harm. He sees the technology as an enabler that gives people 'superpowers' as long as humans keep their agency.
Biological intuition and the last 10%
Most expertise is knowledge-based, but the final 5–10% depends on intuition built over decades. He describes a complex experiment whose outcome a top model predicted with near-100% accuracy — matching intuition he spent 30 years acquiring.
Why cancer is hard, and how it is changing
Cancer is hundreds of diseases whose cells are part of us, making them difficult to target. Immunotherapy, engineered CAR-T cells, and personalized mRNA vaccines now train the immune system to recognize an individual's specific mutations.
Predicting disease years before it appears
A UK Biobank study used a few hundred data points per person to forecast around a thousand diseases before onset, with cancer and neurodegenerative disease predictable roughly a decade ahead. This shifts medicine from sick-care toward prevention.
Practical takeaways
6- 1
Use the strongest reasoning model 20:00
For complex health questions, prefer the most advanced 'thinking' or 'pro' model that reasons at length, rather than a fast instant response.
- 2
Advocate for yourself in the clinic 1:14:00
For complex or ambiguous cases, it is reasonable to ask whether the most current tools and second opinions were considered — advanced models can catch non-obvious diagnoses.
- 3
Track continuous data to catch trends early 1:58:00
Continuous signals such as glucose reveal patterns long before a disease diagnosis; more frequent data beats a single yearly snapshot.
- 4
Build a personal health record 2:28:00
Keep your labs, supplements, sleep, steps, and diet in one consistent place so a model can compare before-and-after and personalize its suggestions.
- 5
Learn your own set points 2:30:00
A few repeated lab measurements let you establish your personal normal range instead of relying only on population averages.
- 6
Optimism itself supports healthy aging 2:33:00
He notes that optimism and resilience — the ability to absorb setbacks and keep going — are among the best things one can do for aging.
Topics & chapters
15The 10-year thesis
Unutmaz opens with his provocation: stay alive for the next decade and you may gain 50 extra years.
Exponential progress and escape velocity
Why the next 10 years may outpace the last century, and what Aubrey de Gray's 'longevity escape velocity' means.
How artificial intelligence already accelerates biology
Reasoning models turn months of data analysis into minutes, including a 112-minute run that produced a 40-page insight report.
The digital twin
A data-rich simulation of a person could compress clinical trials to weeks and enable treatment on demand.
From general intelligence to superintelligence
Definitions of artificial general intelligence and superintelligence, and the missing levels: memory, self-learning, and physical intelligence.
The case for optimism
Why the only existential threat is human misuse, and how the technology can act as an enabler that gives 'superpowers.'
Biological intuition and the last 10%
Whether models can match the hard-won intuition of a 30-year lab veteran — and the moment one nearly did.
The clinic: from unethical to malpractice
Why he argues advanced tools should be standard practice, referencing misdiagnosis statistics and early cancer detection.
Choosing the right model
Generalist vs specialist models, why he prefers the strongest reasoning model, and services that pick the best model for doctors.
Why cancer is so hard
Cancer as hundreds of diseases, and the immunotherapy and CAR-T revolution that reframed treatment.
mRNA vaccines and cancer on demand
Personalized mRNA vaccines, the Australian dog case, and predicting side effects before they happen.
Predicting disease years early
The UK Biobank study, continuous glucose monitoring, and the shift from sick-care to prevention.
Bios singularity: a 25-year-old prediction
How Ray Kurzweil inspired a blog forecasting disease cures by 2035 and age reversal by 2045.
The hallmarks of aging
Aging as gradual information loss and declining resilience, and why prevention is easier than repair.
Reversing aging and the brain
Cellular reprogramming, the special challenge of the brain, and building a personal 'mini digital twin' today.