zlacker

There was a brief discussion/effort to reimplement Eurisko on LessWrong about a decade ago[0], of course it rapidly devolved into whether such effort would promote the end of the human race.

[0]: https://www.lesswrong.com/posts/t47TeAbBYxYgqDGQT/let-s-reim...

>>janzer+(OP)
The hardest part of this is there just aren't enough implementation details to reverse engineer Eurisko.

Especially the heuristics are only ever described very abstractly.