Benja comments on An angle of attack on Open Problem #1 - Less Wrong

30 Post author: Benja 18 August 2012 12:08PM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (84)

You are viewing a single comment's thread. Show more comments above.

Comment author: Benja 19 August 2012 11:47:54PM *  0 points [-]

I will show that my AI will be able to accept itself as a safe rewrite

Now again unsure whether I can show this. Argh.

ETA: The argument I want to be making is the one from the third paragraph in the proof sketch. The problem is the quantification over p -- I need to show in PPT that "for all p, if PPT proves 'p is safe for K steps', then p is safe for K-1 steps". But the axioms of PPT (v0.2) don't allow for parameters (free variables) in the statement C.

Comment author: Benja 22 August 2012 06:31:16PM 1 point [-]

I seem to have fixed the remaining problems; new proof attempt here.