That’s the infamous 120 pages draft (or has it been updated?) I had started reading it some time ago, found some basic errors, lots of rambling and decided it wasn’t worth my time.
The linked article says something like: Superrationality is not necessary. All you need is to realize that in real life, there are no single-shot PDs, and therefore you should use the optimal strategy for iterated PD, which cooperates in the first move.
That’s simply refusing to deal with the original question and answering something different instead.
No, it says that one-shot PD is rare, and when it actually happens defecting is indeed the correct choice, even if it is counterintuitive because we are much more accustomed with scenarios that are similar to iterated PD.
That’s the infamous 120 pages draft (or has it been updated?)
I had started reading it some time ago, found some basic errors, lots of rambling and decided it wasn’t worth my time.
Anyway, I found a good critique to superrationality: http://www.science20.com/hammock_physicist/whats_wrong_superrationality-100813
The linked article says something like: Superrationality is not necessary. All you need is to realize that in real life, there are no single-shot PDs, and therefore you should use the optimal strategy for iterated PD, which cooperates in the first move.
That’s simply refusing to deal with the original question and answering something different instead.
No, it says that one-shot PD is rare, and when it actually happens defecting is indeed the correct choice, even if it is counterintuitive because we are much more accustomed with scenarios that are similar to iterated PD.