On a related note, AlphaGo did not "solve" Go -- and no system ever will. (Checkers is solved, but Chess and Go are simply too large.)
---
RT @pfau
And I generally dislike claims that a field is "solved". Vision is far from "solved" despite great results on ImageNet. Similarly, protein folding is not yet "solved", and probably never will be - but it is now "solved enough" that really exciting things become possible.
https://twitter.com/pfau/status/1333469467846844417