Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How A.I. Conquered Poker

nytimes.com

Post reply on HN
Print view View original

11–20 of 188 posts

Re: How A.I. Conquered Poker

#11

Post by hammock »

HALtheWise wrote: post #8
hashberry wrote: post #2

> he opened a computer program called PioSOLVER, one of a handful of artificial-intelligence-based tools So I checked out this tool, and the team describes themselves as "programmers interested in algorithms"[0] ... what is the difference between A.I. and algorithms? [0] https://www.piosolver.com/pages/about-us

The taxonomy I personally use is: * Every computer program has algorithms, it's an extremely general term for "the idea behind how the computer will solve the problem". Advanced algorithms are typically those that took a lot of human effort to come up with. * Machine Learning refers to a specific class of algorithms where the computer automatically figures out (part of) what it should do based on data. * Deep learnin…

Clearest and most useful descriptions I've seen by far. Thanks for sharing

Re: How A.I. Conquered Poker

#12

Post by im3w1l »

bluecalm wrote: post #9

I am the solver programmer mentioned in the article. I think it came out pretty decently. Maybe not all the details are right (I was late answering fact checking email) but I think it's a decent read and the author has done a good job!

How perfect is the play that piosolver calculates? Do you make any simplifying assumptions?

I noticed in the article that piosolver can incorporate a mixed strategy with multiple bet sizes. Pretty neat, is that a recent feature? As I rememember it, a few years back people would discuss what the optimal bet size was in various situations with the implicit assumption that there is just one.

Re: How A.I. Conquered Poker

#13

Post by ModernMech »

hashberry wrote: post #2

> he opened a computer program called PioSOLVER, one of a handful of artificial-intelligence-based tools So I checked out this tool, and the team describes themselves as "programmers interested in algorithms"[0] ... what is the difference between A.I. and algorithms? [0] https://www.piosolver.com/pages/about-us

Algorithms are provably correct. If you've got no proof of correctness, the best you have is a heuristic.

Re: How A.I. Conquered Poker

#14

Post by ewhauser421 »

bluecalm wrote: post #9

I am the solver programmer mentioned in the article. I think it came out pretty decently. Maybe not all the details are right (I was late answering fact checking email) but I think it's a decent read and the author has done a good job!

Congrats - PIOSolver is an amazing program. What are the future plans for it? It would be great if you didn't need a PhD to set it up though. I keep wondering if I'm actually training myself using the right solutions or not based on the setup :)

Re: How A.I. Conquered Poker

#15

Post by HWR_14 »

btheshoe wrote: post #7
marcodiego wrote: post #6

Is there any remaining logic game which AI hasn't become superhuman? IIRC Monte Carlo tree search has done wonders in this area.

Imperfect information games that require agents to cooperate with each other haven't been solved yet. (multiagent cooperation is still a hard problem) Hanabi is a big(ish) research area. Contract bridge has had some work done on it.

I'm surprised that Hanabi is such a huge research area. Compared to bridge or poker, it seems like a far more simple game.

Re: How A.I. Conquered Poker

#16

Post by ska »

HWR_14 wrote: post #15
btheshoe wrote: post #7

Earlier quoted context omitted.

Imperfect information games that require agents to cooperate with each other haven't been solved yet. (multiagent cooperation is still a hard problem) Hanabi is a big(ish) research area. Contract bridge has had some work done on it.

I'm surprised that Hanabi is such a huge research area. Compared to bridge or poker, it seems like a far more simple game.

> I'm surprised ... it seems like a far more simple game.

This probably shouldn't surprise you. If you are researching how to attack a new/difficult class of problems, typically you look for the simplest versions of that class first.

Re: How A.I. Conquered Poker

#17

Post by bluecalm »

ewhauser421 wrote: post #14
bluecalm wrote: post #9

I am the solver programmer mentioned in the article. I think it came out pretty decently. Maybe not all the details are right (I was late answering fact checking email) but I think it's a decent read and the author has done a good job!

Congrats - PIOSolver is an amazing program. What are the future plans for it? It would be great if you didn't need a PhD to set it up though. I keep wondering if I'm actually training myself using the right solutions or not based on the setup :)

The plans are to start turning it into real company which hopefully means easier to use software, more learning material and better communication. Till now it was mostly two guys working from home and overwhelmed by it all.

Re: How A.I. Conquered Poker

#19

Post by bluecalm »

im3w1l wrote: post #12
bluecalm wrote: post #9

I am the solver programmer mentioned in the article. I think it came out pretty decently. Maybe not all the details are right (I was late answering fact checking email) but I think it's a decent read and the author has done a good job!

How perfect is the play that piosolver calculates? Do you make any simplifying assumptions? I noticed in the article that piosolver can incorporate a mixed strategy with multiple bet sizes. Pretty neat, is that a recent feature? As I rememember it, a few years back people would discuss what the optimal bet size was in various situations with the implicit assumption that there is just one.

Adding multiple betsizes was one of the first features we added in 2015 - so not very recent, no. The one true correct bet size is a seducing idea but that just doesn't exist for most situations. You can try finding the best one with the assumption you can only use one but it's not the best idea in my opinion. Main reasons are that that's the sizing most people will be most ready to play against and EV difference between possible bet sizes is minimal.

What happened in chess is now happening in poker: at first people were all about the best theoretical move and then gradual shift to "not necessary the best just sound enough but less likely to be analyzed by the opponents" started happening.

As to the first part of your question: the solutions are pretty much perfect. The only assumption is about possible bet sizes. As to accuracy we measure it in theoretical exploitability (how much a perfect adversary who knows our strategy exactly could win against us). You can easily go so low that even a theoretical perfect adversary wouldn't get close to beating the rake vs the solution, even at high stakes.

Re: How A.I. Conquered Poker

#20

Post by isaacg »

marcodiego wrote: post #6

Is there any remaining logic game which AI hasn't become superhuman? IIRC Monte Carlo tree search has done wonders in this area.

As far as I know, in Magic: the Gathering, the best bots are far worse than most players. Part of the difficulty is that the rules are so complicated that there are only a couple of complete rules implementations. Beyond that, it's an imperfect information game with far more actions per game than poker, so optimal-solver techniques haven't seen success.
Post reply on HN