Gary Marcus

I am very familiar with DARPA’s program. Yes, it desires to push beyond DL, but DL is very much the baseline. As for China - and it is difficult to get uncensored information- they seem to have a maniacal DL focus.

But this conversation is distracting: my simple original point is that there is a huge pile of money at national levels being poured into some very specific A.I. approaches...and that is a little concerning to me, not unlike what happened in Japan.

You went off on a tangent suggesting I condemned general funding for research..that is not at all what I was saying. Don’t overthink the topic: my spidey sense is tingling, that is all.

If your concern is that there's going to be a lot of wasted money then I think that is misplaced. I guarantee that many of these nations will 'waste' a ton of money for minimal results. That's just the nature of basic research.

No, I am mostly concerned of the opportunity cost...as @GaryMarcus has observed, a lot of these new and heavily funded efforts are proceeding with vigor but with hardly any understanding or appreciation for the history of AI

What great success of AI history are people forgetting? The high development cost and higher ownership cost of hand-coded knowledge bases?

Google Search was built on classical AI and hand-coded knowledge bases. Not too shabby. So was Siri AFAIK, though embeddings surely play a role now. Weather-prediction is still classical models, not deep learning. Best driving systems still hybrids, as far as I know.

PageRank was definitely not "classical AI" or a hand-coded knowledge base. This is patently ridiculous. Knowledge Graph came much much later.