No-limit Texas Hold Em is one of the most complex and popular form of poker and 2 years ago a AI was reported that could beat the best human player at it, but in last Thursday's issue of the journal Science there was a development that was new for 3 reasons:
1) Previously the AI could win only at one on one play but this new program (called Pluribus) could play at a superhuman level against multiple human expert players and win, and that is much more difficult than one on one play. 2) The program started out knowing almost nothing about Poker but taught itself to play at a superhuman level in 8 days and did it without human help. 3) Unlike previous programs this one does not require massive computing power for the training or for the actual play. During training it played against itself trillions of times and needed 512 GB of memory and 8 days on a 64-core Intel Xeon E5-8860 server for a total of 12,288 CPU core hours. If you bought that much processing power on cloud computing it would only cost you $144. After the training period was complete and a winning strategy was found on how to play Poker at a superhuman level then during a real game in real time against real opponents the program only needed 128 GB of memory. Superhuman AI for multiplayer poker <https://science.sciencemag.org/content/early/2019/07/10/science.aay2400> John K Clark -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/CAJPayv37ofS01YqLPA3mFaht9QZPDn9zm4ab%3Ddpa6kXC_bL%3DKQ%40mail.gmail.com.
