[ExI] Virtual Reality is where the aliens are (spike)

[Keith Henson]

On Sun, Aug 23, 2015 at 5:00 AM,   "spike" <spike66 at att.net> wrote:

>>...Is it safe to assume that advanced future computing will generate a lot
> of waste heat?... BillK
> _______________________________________________
>
> If they actually do anything, photons do create waste heat.  The third law
> applies to them as well.
>
> But we know what you meant: photons are a way more efficient way to transfer
> signals than pushing electrons down a wire: entropy production is way lower.
>
> The reason I focus on your comment is not to strain at gnats BillK, but
> rather because I have really been pondering how to calculate temperature
> gradients in an MBrain using entropy.

This has been a concern of mine since the subject first came up--hmm,
more than two decades ago.  You want to do a lot of waste heat
generations computing in a small space to minimize speed of light
delays.  But you have to have a large area to radiate the heat.  Black
box, if it was in space, it would be sunlight in one side and low
temperature thermal radiation out the other.  You can't do more
computation per square meter than can be supported by the energy you
can harvest from that square meter.  The alternative is to make
concentrated power and use small scale intensely cooled hardware.  But
I suspect that on a large scale, you can't do better than whatever
limit is set by computer sandwiched between a sunlight converter
surface and a radiator.

> If I can bound the problem using the
> third law, we might at least get a best-case scenario.  This technique is
> what led me to the insight that MBrains would overheat unless you reflect
> most of the energy out of the system.
>
> By my understanding of entropy and the third law of thermodynamics, an
> MBrain not only has the option of turning its star into a photon rocket, it
> is required to do so.  If that notion is correct, then an MBrain would be
> detectable.  That we don't see one is evidence that MBrains do not currently
> exist, or if so, there aren't many of them.
>
> That is an insight I regret Robert Bradbury did not live long enough for me
> to share with him, for I didn't discover it until a few months after his
> untimely passing.

I wonder if it would be useful to set up a Skype chat room to discuss
these problems?

Keith

> spike