For a decade or two, Caltech professor of physics John Preskill has given many excellent talks and lectures about quantum computing (and also about high energy physics, his original specialty). Much of this material can be found at his Caltech webpage, which is a pleasure dome for physics nerds.
Ten years ago, Preskill gave a wonderful talk, one with very little math, understandable by most physics undergrads, entitled
Quantum Computation and the Future of Physics (delivered at the Berkeley Workshop on Theory of Computation and the Sciences. – 10 May 2002)
The title alone invites endless speculation. Quantum computing is likely to change considerably the technology landscape, but what about “fundamental” physics? Will it change that too very much or not much at all?
All of the talk is interesting, but there is one transparency that I find particularly intriguing (transparency number 48 out of 49). Here it is:
So he thinks Matrix M-theory cannot be simulated efficiently by a quantum computer, although quantum field theory can be. Or at least he thought this 10 years ago. I wonder if he has changed his mind since then. I hope Preskill will be proven wrong (or right) in the near future by someone. String Theorists, get to work.
Should physical theories be required to satisfy certain complexity constraints, and can such requirements serve to winnow out a lot of bad theories? Maybe if the time evolution of a version of string theory cannot be simulated with polynomial efficiently, on a quantum computer, then that version of string theory should be considered kaput and verboten?