New research using theoretical computer science to understand evolution

DPhil candidate Artem Kaznatcheev is the author of a paper in the journal 'Genetics' which uses theoretical computer science to understand evolution. His research shows that computational complexity can act as an ultimate constraint on evolution and thus enable open-ended evolution.

    

Evolution is often conceptualized as a population climbing up a fitness landscape until it reaches a local fitness peak. Experiments have shown that these landscapes can have a rich combinatorial structure. Artem argues that this combinatorial structure can make hill-climbing difficult, effectively transforming the mountain into a maze. On these hard fitness landscapes, no evolutionary dynamic can find the equilibrium of a local fitness peak. At least not in polynomial time.

    

Artem notes that this perpetual maladaptive disequilibrium can result in a slow power-law decay in selection strength that is consistent with ongoing long-term evolution experiments. Such effectively unbounded growth in fitness can help us understand how evolution can continue to innovate. This is especially useful when thinking about the evolution of costly learning or the maintenance of cooperation in populations of complex organisms like humans.

    

Read the full paper here: