Re: MEDIA: Forbes Magazine discusses AI and the Singularity (well, kinda)

Michael Nielsen (
Thu, 25 Jun 1998 15:46:15 -0600 (MDT)

On 25 Jun 1998, Anders Sandberg wrote:

> Michael Nielsen <> writes:
> > I would be interested to hear what people think about the possibility of
> > extending the reign of Moore's law beyond 2020, when we hit atomic-sized
> > components, dissipating kT Joules per operation, assuming it continues on
> > its merry path for the next 20 years.
> I think the dissipation problem can be dealt with using reversible
> logic, there is already a fairly large literature on the subject.

I think that is unlikely. The reversible architectures I am aware of have
higher fundamental error rates than existing irreversible architectures,
and will likely require considerable error correction.

Error correction, in turn, involves the dissipation of heat, by
elementary thermodynamic arguments. Essentially, error correction is a
procedure for lowering the entropy of a physical system (the computer).
The entropic cost is paid in heat dissipated into the environment.

Maybe I'm wrong, though, and there exist good architectures for reversible
computation with a low error rate. Does anybody know of one?

> As for size, it seems like we need a jump to nuclear densities
> ("femtotech") to go beyond nanotechnology. That might be a real limit
> since nuclear matter is so unstable at our energy/density scale:
> We don't have anything like Drexler's proofs of concept for femtotech.

Justification for another shot at the SSC? :-)

Attempting to develop such a proof of concept would be quite a challenge!

Michael Nielsen