Hacker News Reader

A trip down memory lane. Some of the “hot” technologies took off (CPU architectures, genome computations, ray tracing); some lingered on in the same early 2000s state (genetic algorithms, etc).

I, for one, am slightly puzzled why genetic algos never became mainstream in any computational field.

View on HN

Computer programming is hard enough when you know how the algorithms work, so the idea of genetic programming as "randomly generating program text and mixing them together till it works" is definitely more than anyone wants to deal with.

Genetic programming is also a machine learning technique where you generate random answers then mix them together until it increases the fitness function. Since it knows nothing about the actual problem space, it runs very slowly yet doesn't work better than something more sensible like simulated annealing.

I don't use them much myself, but genetic algorithms (and "evolutionary computation (EC)" more broadly) have become pretty big in some areas. In AI there seems to be a bit of a geographical tilt, with EC being bigger in Europe and Asia than North America. And in application areas, EC seems to be bigger in areas considered "engineering" rather than "computing" (in terms of publications, IEEE-affiliated rather than ACM-affiliated venues).

Curious to know, thanks.

Reply on news.ycombinator.com