Peter
Fedichev

In 2015, someone showed Peter Fedichev a list of animals that don't appear to age. Naked mole rats. Certain bats. Organisms that seem to shrug off the biological wear-and-tear that kills everything else. For most scientists, this would be a curiosity. For a theoretical physicist who'd spent years modeling quantum gases and condensed matter systems at the University of Innsbruck, it was a provocation: if the math describes everything else, why not this?

He published his first aging paper that year. Three years later, he co-founded Gero in Singapore with Maxim Kholin. The premise was simple and radical at once: biological aging is not fundamentally a biochemical problem. It is a problem in non-equilibrium thermodynamics - the same mathematical territory as heat engines, entropy, and complex physical systems failing over time. The drug discovery world mostly disagreed, or ignored him. Pfizer eventually came around. Then Chugai Pharmaceutical, the Japanese arm of Roche, signed a joint research deal in July 2025 worth up to $250 million in milestones and royalties.

The career pivot was not painless. Before Gero, Fedichev was among the top-2 cited Russian physicists under 35. He had published extensively on ultracold quantum systems, run his own computational drug design company (Quantum Pharmaceuticals, applying quantum mechanics to protein-ligand binding), and held positions at the Kurchatov Institute, the FOM Institute in Amsterdam, and the University of Innsbruck. He was, by any measure, succeeding at a prestigious field.

He left it anyway. Not for the money - longevity biotech was not obviously a wealth-generating proposition in 2018. He left because the question was better. Why do organisms age? Why do some not? What does the data actually say, when you look at it through the lens of statistical physics rather than molecular biology?

The answer Gero has been building toward involves three variables that Fedichev calls stress, noise, and damage. Stress covers reversible physiological disruptions - the kind your body bounces back from when you're young and increasingly fails to recover from as you get older. Noise describes the small internal fluctuations in biological systems that accumulate over a lifetime. Damage is the slow accumulation of irreversible structural deterioration. Together, they form a framework borrowed from the physics of complex systems near failure. It is not a metaphor. It is a model - quantitative, falsifiable, and increasingly validated by real-world longitudinal medical data.

Gero's platform applies this framework to real-world datasets - electronic medical records, wearable sensor data, longitudinal biomarker profiles. The output is not another aging clock that tells you your biological age and stops. It is a model of systemic resilience: how well your body responds to perturbation, and at what rate that capacity declines. That rate, Fedichev's team has shown, predicts mortality more accurately than chronological age, and the model can be run on something as simple as an iPhone's built-in accelerometer.

GeroSense, launched in 2018, did exactly that - predicting biological age from wearable data years before the feature appeared in any major consumer device. A 2021 Nature Communications study from the team proposed that human resilience declines with age and may fundamentally limit maximum lifespan - a finding that reframed how the longevity field thinks about intervention timing and therapeutic strategy.

The scientific case got strong enough that Pfizer came knocking in 2023 for a collaboration on fibrotic disease targets. Then Nature made Fedichev co-editor of its 2024 Gerophysics collection alongside Brian K. Kennedy, effectively acknowledging that "gerophysics" had become a real field. Then in May 2024, he entered a formal public debate on defeating aging at the Foresight Institute and walked out with a $10,000 check after expert judges voted 42-38 in his favor.

The Chugai deal in July 2025 was a different order of magnitude. A joint research and license agreement with a subsidiary of Roche - one of the world's three largest pharmaceutical companies - validated not just Gero's science but its commercial model: use physics-informed AI to identify therapeutic targets that traditional biology-first approaches miss, then partner with established pharma to do the clinical work. The potential $250 million in milestones and royalties is, by biotech standards, a serious number for a company that has raised a total of $8.2 million in venture capital.

That same year, Gero launched ProtoBind-Diff, described as the first AI model capable of drug design using only protein sequence - no structural data required. If the claim holds up under scrutiny, it addresses one of the most persistent bottlenecks in structure-based drug discovery. The team, now 140 people, operates across Palo Alto and Singapore, with Fedichev also heading the Biological Systems Simulation Lab at Moscow Institute of Physics and Technology.

His son, reportedly, once asked what happened to the dinosaurs, then announced he wanted to fly to the moon and defend Earth. The ambition is apparently hereditary.