4 ms·
Well.... There's the possibility that there's a certain degree of myopia in the AI field as a whole. As in, we know that there are some pretty gaping holes in o
by marvin 7y ago
Well.... There's the possibility that there's a certain degree of myopia in the AI field as a whole. As in, we know that there are some pretty gaping holes in our models and understanding, and most of the effort is spent on refining approaches that we have already validated.
Maybe a better analogy would be "specialize physical theories that work in all environments, not have to be adapted separately for the ocean, space, the atmosphere, the forest" etc.
- b3kart 7y agoThere's certainly a lot to be done, and we'll likely need new approaches. But is it really productive to be tackling a general problem when we don't even know how to solve specific sub-problems? Especially if solving said sub-problems would bring a lot of value in its own right.
- marvin 7y agoIt’s basic research. No one knows anything about which approaches will work. If a genius millionaire technologist wants to dedicate all their time and effort to any novel approach, I’d strongly endorse it. (Not that it matters; they’ll do it anyway). I feel likewise for any research effort; it’s not like this will put all cancer research on hold, or more immediately practical AI research. It’s just a few hundred people globally :) And it’s such promising technology.