News for hackers. Stuff that matters.
Re:Competitive Self-Play by aeleos on Wednesday October 11, 2017 @02:53PM (#15452264)
The main issue with self-play is that unless done very methodically it can lead to behavior that does not learn what we want it to learn, but games the simulation and basically cheats. Its not a perfect solution and is just another tool being used to improve models. It can produce really interesting results especially in complex games but its also not perfect.
Re:Competitive Self-Play by gooseus on Wednesday October 11, 2017 @04:01PM (#15452928)
I agree... I feel like advanced games will always need guidance to alter the game such that they AI don't reach some local maxima through an exploit.

This happens all the time when making games for humans and is evident by how many balancing patches are made to new, highly competitive games (such as any Blizzard title).

The next logical step would be for another, impartial AI to observe the games and changing the rules and parameters intelligently as they evolve to guide the player AI toward the actual goal.

So, I'll just dive sideways right into a religious/philosophical thought based on the simulation discussion I've been having all over the place:

A universe-sized simulation built for a purpose, which requires simulated intelligence to carry out that purpose, would almost certainly include a God intelligence to alter parameters and induce suffering/hardship to direct the simulated intelligence toward that purpose.

Re:Competitive Self-Play by danohu on Wednesday October 11, 2017 @06:33PM (#15454129)
Earlier iterations are buggier and have poorer dev tools. So the God intelligence has more need to smite and command the AIs within the game.

After a while the bugs are ironed out, so God can settle back and gently tweak parameters at a distance.

Re:Competitive Self-Play by visarga on Wednesday October 11, 2017 @11:55PM (#15455554)
> After a while the bugs are ironed out, so God can settle back and gently tweak parameters at a distance.

That explains the hands off approach God has lately with the human society.

Re:Competitive Self-Play by Danihan on Wednesday October 11, 2017 @05:11PM (#15453531)
>would almost certainly include a God intelligence to alter parameters and induce suffering/hardship to direct the simulated intelligence toward that purpose.

Wow that's pretty deep, actually.

Re:Competitive Self-Play by vorotato on Wednesday October 11, 2017 @03:50PM (#15452834)
One could argue that "cheating" is a valid solution.
Re:Competitive Self-Play by tlb on Wednesday October 11, 2017 @04:34PM (#15453206)
Only when the learning environment is the same environment it will use when deployed. Agents for robots are often trained in simulation, where it's a common problem for the agent to exploit a physics bug in the simulator.
Re:Competitive Self-Play by recursive on Wednesday October 11, 2017 @03:53PM (#15452858)
You only get what you measure. If you can figure out how to measure the right thing, then that's what you will get.