Re:The Myth of a Superhuman AI
by blamestross on Wednesday August 25, 2021 @08:17PM
(#28308949)
I'm in total agreement about the potential of growing AGI out of these methods, but there will be bottlenecks well before the gods of the singularity come knocking.
Re:The Myth of a Superhuman AI
by blamestross on Wednesday August 25, 2021 @07:30PM
(#28308534)
> What is the difference between experiencing reality and a game?
Finding out the consequences of an action is a lot more expensive in reality than a simulation of a game.
Re:The Myth of a Superhuman AI
by namedlambda on Thursday August 26, 2021 @07:03AM
(#28313203)
There is nothing fundamentally different between an infinite horizon game and reality.
The difference I can see is that there is no one explicit objective function, but this doesn't stop generally capable agents [1], and doesn't imply that inverse RL is not possible.
> The ability to play go very well doesn't enable an AGI to better understand reality.
I disagree, model based RL constructs a model of the agent's reality and can use it to plan ahead, train the agent, or do some form of monte-carlo tree search. The latter is something very similar to how we imagine and think about the future.
[1] https://deepmind.com/blog/article/generally-capable-agents-e...