> this kind of AI research is seriously out of vogue I would think this is a consequence of it being a very hard problem. ML gets all the industry funding and

> this kind of AI research is seriously out of vogue I would think this is a consequence of it being a very hard problem. ML gets all the industry funding and publicity because you get results that are immediately useful. You can work on General AI for decades and come up empty handed since there doesn't seem to be an incremental approach where intermediate results are useful. So it's closer to esoteric mathematics or philosophy in terms of "being en vogue". So I see this mostly as a reflection of the academic landscape in general. Funding is more focused on application and less on theoretical / fundamental research.