With AlphaZero proving chess is far from solved, what insights you think we will discover later? Imho, Mobility calculations of today engines are too flaky and unreliable. AlphaZero knows how much mobility it needs and which moves steal mobility from opponent. Current tactical ability of programs doesn't understand positional sacrifices well: AlphaZero picks worse positions with great tactical potential which it exploits later.
Name:
Lain2018-01-03 16:29
Stockfish developer Tord Romstad responded with "The match results by themselves are not particularly meaningful because of the rather strange choice of time controls and Stockfish parameter settings: The games were played at a fixed time of 1 minute/move, which means that Stockfish has no use of its time management heuristics (lot of effort has been put into making Stockfish identify critical points in the game and decide when to spend some extra time on a move; at a fixed time per move, the strength will suffer significantly). The version of Stockfish used is one year old, was playing with far more search threads than has ever received any significant amount of testing, and had way too small hash tables for the number of threads. I believe the percentage of draws would have been much higher in a match with more normal conditions."[5]
Name:
Anonymous2018-01-03 18:48
>>2 AlphaZero still searches orders of magnitude LESS moves. Its highly selective at what it searches. The hardware is necessary being normal CPU/GPU can't handle modern neural networks.
Name:
Anonymous2018-01-04 7:07
The main flaw in program like Stockfish is that they have universal one-size-fits-all evaluation: its eval weights are not fit to the position, but to generic class of position like "we have a bishop pair, ignore how strong they are in this closed maze of pawns, just add their values"
Name:
Anonymous2018-01-04 7:55
There will be a cold war between the USA and China, and the world map will be redrawn.