PERSON
John Holland
The computer scientist who turned biological evolution into an algorithm—inventor of the genetic algorithm, architect of the theory of complex adaptive systems, and the thinker who best explained why the most powerful things in the world are grown rather than designed.
John Holland is the philosopher of grown intelligence. In 1975, in a difficult book called
Adaptation in Natural and Artificial Systems, he took the logic of natural selection—variation, recombination, selection—and wrote it down as a procedure a computer could run, an algorithm that found good solutions to hard problems without knowing in advance what a good solution looked like. He called the result a
genetic algorithm, and it was an act of unification: the natural and the artificial, as his title declared, were two instances of one process. That unification is the seed of everything Holland contributed, and its relevance to the present moment is not that he predicted
large language models—he did not; he died in 2015, before the largest models existed—but that he understood, earlier and more rigorously than almost anyone, the kind of thing they are. The most consequential systems we have built are not designed but grown; they