DanielFilan comments on Alignment proposals and complexity classes

[DanielFilan]

I don’t see where the post says that humans can solve all problems solvable in polynomial time given access to an oracle.

From the post:

First, we’ll assume that a human, $H$, is polynomial-time such that $H$ can reliably solve any problem in $\text{P}$

[adamShimi]

Yes, this says that humans can solve any problem in P. It says nothing about using M as an oracle.

[DanielFilan]

It doesn’t talk about using M as an oracle, but otherwise I don’t see how the proofs pan out: for instance, how else is

Given $p$, return $M(p:s_0:[]:x_0:[\left]\right)$ where $s_0\in S_p$ is the initial state of $T_p$, $x_0$ is the empty tape symbol, and $:$ is string concatenation.

supposed to only take polynomial time?

[adamShimi]

You’re right, after rereading the proofs and talking with Evan, the only way for H to be polynomial time is to get oracle access to M. Which is slightly unintuitive, but makes sense because the part of the computation that depends on H and not on M is indeed polynomial time.