Watch Desk posted an update
Alexander Mattick of Fraunhofer IIS and UTN argues that even a perfect world model of chess would not automatically beat Magnus Carlsen.
Why it mattersThe point, discussed on Machine Learning Street Talk, is that predicting what could happen and selecting the best action are different technical problems. Mattick uses humanoid robots as the livelier example: a machine may perform an impressive backflip yet fail at the less cinematic task of fetching a carton of milk. It is a useful warning for anyone treating prediction as intelligence’s finishing move. The hard bit may still be deciding what to do.
Discuss: Should AI research prioritise better world models, or is reliable action selection the harder problem that deserves more attention?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.