[extropy-chat] Is Many Worlds testable?

[John K Clark]

Russell Wallace

> what's your proposal for using an arbitrarily small amount of _energy_?

I'd love to take credit but it's not my proposal, nor is it new. In 1961
Rolf Landauer did extensive work on the thermodynamics of computing. He
proved that you always need energy to change one bit of information, but the
amount of energy needed to change that one bit can be made arbitrarily small
if you use a computer that is reversible; you just need to slow it down.
Today's computers are not reversible because there is no reason for them to
be, in our big clunking machines the theoretical limits from the
thermodynamics of computation are just not important compared with other
factors. But a nanocomputer would almost have to be reversible, otherwise it
would generate so much heat it would be more like a bomb than a computer.

  John K Clark