4 ms·
Depends on the domain. Why did DeepMind pick Go as a challenge rather than Chess? Because with Chess you can store the entire search space in memory. With Go
by op03 6y ago
Depends on the domain.
Why did DeepMind pick Go as a challenge rather than Chess?
Because with Chess you can store the entire search space in memory. With Go you can't.
- dreamer7 6y ago> Because with Chess you can store the entire search space in memory. That is incorrect. Quoting from [0] > Chess has approximately 10^120 game paths. These positions comprise the problem search space. Typically, AI problems will have a very large space, too large to search or enumerate exhaustively. But as Go has an even larger search space, it makes sense that DeepMind sees it as a bigger challenge. [0] - http://www.cse.uaa.alaska.edu/~afkjm/cs405/handouts/search.pdf http://www.cse.uaa.alaska.edu/~afkjm/cs405/handouts/search.p...
- op03 6y agoDeep Blue brute forced that search space. With 1997 hardware. They didn't have to look at 10^120 to beat Kasprov. "The system could search to a depth of between six and eight pairs of moves—one white, one black—to a maximum of 20 or even more pairs in some situations" - https://www.scientificamerican.com/article/20-years-after-deep-blue-how-ai-has-advanced-since-conquering-chess/ https://www.scientificamerican.com/article/20-years-after-de...
- edgarvaldes 6y agoIs chess already solved?
- aj7 6y agoEffectively, yes. But I attach no greater philosophical significance to that than to say drilling holes is solved because most people can go buy a nice electric drill foR $30-$40. Flying to the moon is solved, too.