I can’t believe it. I successfully explained P vs NP to a smart theology student in under an hour while on holiday with her; I can’t believe Feynman wouldn’t be able to grasp it faster.

We get lots of kooks in crypto, as you can probably imagine. I didn’t realise that kooks tried to prove P != NP – I’m more used to seeing them present algorithms that puport to solve NP-complete problems in polynomial time.

]]>Thanks for the anecdote! I’ve also heard that Feynman couldn’t understand what P vs. NP is and why it’s an open problem. I don’t know if that’s true, but I hope it is, since it would imply the non-universality of Feynman’s brain.

]]>I’m inclined to agree with your criteria.

However, in connection with Baker-Gill-Solovay and Razborov-Rudich, it’s sobering to remember Feynman’s response when asked by Dirac if QED was unitary. Essentially, he said “I’ll write down the theory, and you can tell me if it’s unitary.” (I’m not advocating ignorance, just pointing out an empirical fact.)

Michael Nielsen

]]>No, I think this one’s the first.

]]>I’m reading it.

I simply don’t understand why these people don’t implement (in case of P=NP claims) their algorithms and run them on data… Duh! I suppose they don’t even know any language…

]]>———————

*I wonder why the authors of these papers aren’t aware of these stumbling blocks? Sloppiness?*

If they were aware of them, they wouldn’t be writing papers claiming to have proved P!=NP.

]]>If they were aware of them, they wouldn’t be writing papers claiming to have proved P!=NP.

]]>