Open Side Menu Go to the Top
Register
How far away are we from online poker being completely doomed by AI? How far away are we from online poker being completely doomed by AI?

05-23-2018 , 03:04 PM
It is clear that the world is moving towards automation and AI doing every task that humans do. This is not true just for easy repetitive tasks, but for more complex tasks too. Even jobs such as doctors, lawyers and engineers are all at risk of automation over the next several decades.

Online poker is without doubt moving in this direction as well. It is already partly there, but clearly not fully there. The question is, in how many years will the entirety of online poker be completely unbeatable for a good, solid human player? When will online poker pros as we know them today become extinct as a profession?

These things happen very quickly as well. 5-10 years ago, people never thought a computer would beat a human at the game Go. Now we have AlphaGo which crushes the best human players, then AlphaGo Zero which crushes AlphaGo, then Alpha Zero which crushes anyone at any game which is programmed into it.

This is very fast advancement. Technology power increases exponentially each year. So once one game becomes solved and completely unbeatable for a human player, there could easily be a domino effect whereby all games become solved in a matter of a few years. Already we have Libratus which can crush any human at HU NLHE, we can't be that far away from every game type being completely solved.

What are your opinions on this issue?
05-23-2018 , 03:11 PM
10 years at the most.

and if cell phones are not banned in live environments, that will also be a problem in live games.
05-23-2018 , 03:23 PM
Quote:
Originally Posted by D33P
10 years at the most.

and if cell phones are not banned in live environments, that will also be a problem in live games.
Interesting. Perhaps someone with experience of AI/machine learning could give us a more precise figure and their reasoning behind it.
05-23-2018 , 03:33 PM
As soon as skynet becomes self aware online poker is officially unbeatable
05-23-2018 , 03:39 PM
Quote:
Originally Posted by TheGreatCornholio
As soon as skynet becomes self aware online poker is officially unbeatable
Skynet mentioned one post later than when OP started a very similar thread 2 years ago. Disappointing.
05-23-2018 , 03:39 PM
At SSNL its already dead due to AI and botting, almost impossible for new players to move up the limits now.
05-23-2018 , 03:44 PM
Quote:
Originally Posted by .isolated
Skynet mentioned one post later than when OP started a very similar thread 2 years ago. Disappointing.
If I would have seen the thread earlier Skynet would be posted as comment #2. Oh well, I’ll try harder in a couple of years when the next thread gets started by OP
05-23-2018 , 03:49 PM
About tree fiddy.
05-23-2018 , 04:16 PM
Basically now, you can still make money, but the games are noticeably more boring now considering how many bots you are playing.the bots are getting better and better and rake is sky high, I def wouldn’t quit your day job.
05-23-2018 , 04:22 PM
Quote:
Originally Posted by MatthewNL
At SSNL its already dead due to AI and botting, almost impossible for new players to move up the limits now.
Very interesting. Whenever I see threads on forums saying 'I'm stuck at X limit, what am I doing wrong?' where X limit is always small stakes/micros, you get a lot of obnoxious replies like 'lol improve, get better, you're bad at poker'. But never an explanation similar to what you said (mixture of bots/AI + high rake).
05-23-2018 , 04:26 PM
do we have evidence of 'rampant botting' on pokerstars or are we just talking out of our asses
05-23-2018 , 04:30 PM
Quote:
Originally Posted by omnishakira
do we have evidence of 'rampant botting' on pokerstars or are we just talking out of our asses
No one even mentioned pokerstars. But I wouldn't be surprised if there was a decent amount of botting going on there. Just not as much as other sites.
05-23-2018 , 04:32 PM
pokerstars is what matters when we discuss the state of online poker, no one cares about what happens on ipoker or ongame

Quote:
Originally Posted by SpinMeRightRound
I wouldn't be surprised if there was a decent amount of botting going on there. Just not as much as other sites.
any reasons to believe this or you just feel like its the case?
05-23-2018 , 04:42 PM
Quote:
Originally Posted by omnishakira
pokerstars is what matters when we discuss the state of online poker, no one cares about what happens on ipoker or ongame
I'm sure plenty of people care what happens outside pokerstars. The whole American market does, that's for sure. You basically stuck 2 fingers up at them by posting what you just did.

Quote:
any reasons to believe this or you just feel like its the case?
Because pokerstars cares about botting more than other sites, so it goes without saying they catch more bots and it deters botters from playing there. But no doubt some go undetected, you'd be stupid to think pokerstars is completely bot-free.
05-23-2018 , 04:46 PM
a better question to me: is it possible to intentionally develop a game that's as easy as hold em' for beginners to understand (i.e. not omaha) but is exponentially more complex for bots? or is any game thats ultimately developed just effectively 4 hours of alpha zero away from GTO?

the new game, as far as im concerned, could develop anything, a new deck, modified rules, etc.
05-23-2018 , 04:48 PM
Quote:
Originally Posted by SpinMeRightRound
Very interesting. Whenever I see threads on forums saying 'I'm stuck at X limit, what am I doing wrong?' where X limit is always small stakes/micros, you get a lot of obnoxious replies like 'lol improve, get better, you're bad at poker'. But never an explanation similar to what you said (mixture of bots/AI + high rake).
I'm sure there's some botting going on but that doesn't mean those "obnoxious replies" aren't valid comments. Chances are if you're losing at micro or small stakes, you're not trying. Some people just aren't capable of winning, either because for whatever reason things just don't click or more likely, they're misapplying concepts which is costing them tons of bbs.
05-23-2018 , 09:30 PM
Quote:
Originally Posted by theskillzdatklls
a better question to me: is it possible to intentionally develop a game that's as easy as hold em' for beginners to understand (i.e. not omaha) but is exponentially more complex for bots? or is any game thats ultimately developed just effectively 4 hours of alpha zero away from GTO?

the new game, as far as im concerned, could develop anything, a new deck, modified rules, etc.
Just off the top of my head, how bout this:

The big blind's hand is face up and he gets to see the flop without putting in any extra preflop money.
05-24-2018 , 01:44 AM
Quote:
The big blind's hand is face up
According to The Theory of Poker and its fundamental theorem, that would allow people to play perfectly against them now wouldn't it?
05-24-2018 , 01:48 AM
Quote:
Originally Posted by David Sklansky
Just off the top of my head, how bout this:

The big blind's hand is face up and he gets to see the flop without putting in any extra preflop money.
Wasn't there a challenge where 1 guy had AA dealt face up and the other random cards and the guy with any 2 won?
05-24-2018 , 02:18 AM
It already came and went, sorry
05-24-2018 , 03:07 AM
I think that one of the mistakes in the thinking here is that "quality of AI" is a meaningful risk of "dooming" online poker.

You don't "doom" online poker by having one brilliant poker AI that can defeat Tom Dwan. Rather, you "doom" online poker by having widespread AI that is actively used in many games.

AI has been good enough to defeat enough poker players for a decade (or more, in some games). It hasn't. That should give us all hope.
05-24-2018 , 04:38 AM
Quote:
Originally Posted by SpinMeRightRound
It is clear that the world is moving towards automation and AI doing every task that humans do. This is not true just for easy repetitive tasks, but for more complex tasks too. Even jobs such as doctors, lawyers and engineers are all at risk of automation over the next several decades.

Online poker is without doubt moving in this direction as well. It is already partly there, but clearly not fully there. The question is, in how many years will the entirety of online poker be completely unbeatable for a good, solid human player? When will online poker pros as we know them today become extinct as a profession?

These things happen very quickly as well. 5-10 years ago, people never thought a computer would beat a human at the game Go. Now we have AlphaGo which crushes the best human players, then AlphaGo Zero which crushes AlphaGo, then Alpha Zero which crushes anyone at any game which is programmed into it.

This is very fast advancement. Technology power increases exponentially each year. So once one game becomes solved and completely unbeatable for a human player, there could easily be a domino effect whereby all games become solved in a matter of a few years. Already we have Libratus which can crush any human at HU NLHE, we can't be that far away from every game type being completely solved.

What are your opinions on this issue?
AI is not a problem in online poker, because players can always agree not to use any AIDS.
05-24-2018 , 05:04 AM
the answer to this question is much more technical as the posts so far suggest, but you can't expect much from an OP with the following premise:

Quote:
Originally Posted by SpinMeRightRound
Already we have Libratus which can crush any human at HU NLHE, we can't be that far away from every game type being completely solved.
a wrong assumption (can't crush any human, doesn't even play hunl but rather a one stacksize game) and the implication that, once heads up is solved, the rest can't be too far away will induce a bunch of speculation and skynet jokes but prolly not a factual discussion.
05-24-2018 , 06:14 AM
Far more worrying to me than widespread AI is the advancement of real-time GTO solvers, and their increasing availability. AI is still a way from being practical for a "recreational" cheat, but real time GTO is out there now, and the only thing stopping it from spilling over from hypers and spin n goes is processing power. Couple of years and standard PCs will be able to handle the calculations at speed, then it's GG online poker, it was fun.
There was a thread somewhere else on the NVG forum but it's been closed for some reason; I was interested to see what people thought we could do to fight them/ beat them.
05-24-2018 , 08:17 AM
let’s get some dealer’s choice in the house!
Closed Thread Subscribe
...

      
m