Research
Life runs far from thermodynamic equilibrium, burning fuel to sense, copy, move, keep time and build order. Thermodynamics draws the map of what such a system can do, and kinetics chooses where on that map it sits. We look for laws that hold whatever the molecular details, laws that tell us what physics permits or forbids and how a living system can be controlled.
Thermodynamic constraints
How does nonequilibrium driving shape and limit what biochemical processes can do?
Carnot bounded every heat engine without knowing how any of them was built, and we ask the same question of the biochemical processes that keep cells alive. Kept far from equilibrium by a steady supply of free energy, these processes give rise to self-organized order, such as spatial patterns, reliable clocks and accurate copying. We look for the rules that decide which kinds of order a given drive can sustain, and what each of them costs, whatever the molecular details.
Selected papers
Response and control
How do living systems respond to change, and what are the general principles of their control?
Far from equilibrium, a system is controlled through its drive, and fuel, gradients and the rates they power are the knobs that cells, drugs and experimenters turn. Classical response theory, built on the fluctuation–dissipation theorem, does not reach these systems. We look for general principles that govern how a cell, or any driven system, answers a change of any size, and how far, how fast and how precisely it can be controlled.
Nonequilibrium anomalies
Which equilibrium intuitions fail when a system is driven, and what replaces them?
At equilibrium, energy and entropy alone decide which states are populated, and kinetics only sets how fast that balance is reached. A steady drive, even a plain temperature difference, ends this division of labour, and kinetics starts to decide the outcome. We look for the minimal ingredients of each such effect, asking first what equilibrium alone already does and then what the drive adds and at what price.
Selected papers
New directions
-
Physical principles of computation and learning
Cells, physical systems and neural networks all compute and learn, and we ask what thermodynamics says about the energy, speed and reliability of these processes.
-
Science in the age of AI
We use AI to help discover theory, and we use statistical physics to understand AI.