The quoted paper is from 2018. Evolutionary programming is IMO similar to a search algorithm. The biosphere is not. It's not searching; it's removing unfit matches. Which is far less efficient.
I'm a big fan of evolutionary programming; it's just inefficient in the past. I think LLM agents might just be the little advantage they need, like guided missiles with GPS.
Evolutionary programming is hard though. I think it might answer OP's question - it's something that's difficult enough for most people to avoid, but there would be greatly increased interest in it.
Comments
BTW, on the topic of the fading of genetic algorithms, here is an interesting recent take: https://statmodeling.stat.columbia.edu/2025/04/17/what-happe...
The quoted paper is from 2018. Evolutionary programming is IMO similar to a search algorithm. The biosphere is not. It's not searching; it's removing unfit matches. Which is far less efficient.
I'm a big fan of evolutionary programming; it's just inefficient in the past. I think LLM agents might just be the little advantage they need, like guided missiles with GPS.
Evolutionary programming is hard though. I think it might answer OP's question - it's something that's difficult enough for most people to avoid, but there would be greatly increased interest in it.