What I think really makes P=NP interesting is that proving them equal is seemingly within range of many programmers. All you need to do is take an exponential algorithm for one of the reference problems and optimize it down to polynomial time. The existance of the algorithm would prove equality, and net the author a cool million.

Easy to consider, perhaps a little harder in practice 🙂

DjamÃ©says:btw, there also made a movie about P =NP

http://en.wikipedia.org/wiki/Travelling_Salesman_(2012_film)

Mehmet Suzensays:Lance Fortnow has also written a recent review paper on CACM called

The Status of the P Versus NP Problem .

Paul W Homersays:What I think really makes P=NP interesting is that proving them equal is seemingly within range of many programmers. All you need to do is take an exponential algorithm for one of the reference problems and optimize it down to polynomial time. The existance of the algorithm would prove equality, and net the author a cool million.

Easy to consider, perhaps a little harder in practice 🙂

Paul.