Likely AGI architectures?

post by Anirandis · 2020-07-29T03:49:47.822Z · score: 3 (2 votes) · LW · GW · None comments

This is a question post.

Suppose we were to build an AGI this century. Should we expect it to use the current machine learning paradigm, or should we expect some new form of architecture to find more success? If so, what could it look like?

Could you feasibly hard-code cognitive skills into a system, sidestepping the "learning" process in current deep learning systems? Could something like expert systems feasibly come into the forefront?


None comments

Comments sorted by top scores.