Hi Rohin, are older version of the newsletter available?
Also:
This sounds mostly like a claim that it is more computationally expensive to deal with hidden information and long term planning.
One consideration: When you are exploring a tree of possibilities, every bit of missing information means you need to double the size of the tree. So it could be that hidden information leads to an exponential explosion in search cost in the absence of hidden-information-specific search strategies. Although strictly speaking this is just a case of something being “more computationally expensive”, exponential penalties generically push things from being feasible to infeasible.
Hey Jess, as Ben mentioned I keep all newsletter-related things on my website.
I agree that in theory hidden information leads to an exponential explosion. In practice, I think you don’t need to search over all the exponentially many ways the hidden information could be in order to get good results. (At least, you don’t need to do that in order to beat humans, because humans don’t seem to do that.)
I think overall we agree though—when I said “it wasn’t clear how to make things work with hidden information—you could try the same thing but it was plausible it wouldn’t work”, I was primarily thinking that the computational cost might be too high. I was relatively confident that given unbounded compute, AlphaGo-style algorithms could deal with hidden information.
Hi Rohin, are older version of the newsletter available?
Also:
One consideration: When you are exploring a tree of possibilities, every bit of missing information means you need to double the size of the tree. So it could be that hidden information leads to an exponential explosion in search cost in the absence of hidden-information-specific search strategies. Although strictly speaking this is just a case of something being “more computationally expensive”, exponential penalties generically push things from being feasible to infeasible.
They’re all available at his LW profile and also at his offsite blog.
Hey Jess, as Ben mentioned I keep all newsletter-related things on my website.
I agree that in theory hidden information leads to an exponential explosion. In practice, I think you don’t need to search over all the exponentially many ways the hidden information could be in order to get good results. (At least, you don’t need to do that in order to beat humans, because humans don’t seem to do that.)
I think overall we agree though—when I said “it wasn’t clear how to make things work with hidden information—you could try the same thing but it was plausible it wouldn’t work”, I was primarily thinking that the computational cost might be too high. I was relatively confident that given unbounded compute, AlphaGo-style algorithms could deal with hidden information.