Re: complexity and heat, an analogy in the history of science

Eric Watt Forste (
Mon, 18 Aug 1997 13:25:05 -0700

Nicholas Bostrom writes:
> "Heat", in modern thermodynamics, is used to refer to the thermic
> energy of a system.

Yeah, but people screw this usage up all the time because thermic
energy is not something that can be directly sensed without knowledge
of specific heat of the substance you are touching with your skin
(and knowledge of the temperature of your fingers or whatever you're
using as a thermal sensor). What we feel is the rate of heat
conduction across our skins, plus incident infrared radiation and
absorbed light, and that relates only messily and incompletely to
what modern thermodynamicists call "heat" in the strict sense.
Caloric lives on in human nervous systems, lurking in the currents
of "common sense".

It's a bit like trying to translate candelas into watts, or watts
into candelas. Candelas really are just measuring power, like
watts, but power seen through a inverted-U shaped filter that looks
only at the visible wavelengths of the full photonic spectrum.
Likewise, our current bodies only pick up on some of the phenomena
of heat, and we manage to piece together concepts of the "physical
reality" going on behind our sensorium. That real historical progress
is possible in this process of piecing together and testing our
perceptual filters and models of the world has always fascinated
me, and is still rather mysterious to me.

> On this list, people often seem to use the word "complexity" to
> mean:" system which I happen to find intellectually interesting".
> The people who do this then ususally assume without argument that
> the post-singularity future will be extremely complex in this sense.

One of the functions of gray-goo stories is to remind people
that we have no guarantee that the future will be very
interesting. I find that rather motivating. It's one of the
things that keeps my optimism dynamic rather than static.

> In my opinion, that is a possibility; but it is also very possible
> that it won't be, even if we assume that some form of highly
> intelligent life survives. We should be on our guard against possible
> biases that might make us inclined to believe in something just
> because it would be interesting it it were so. (Is this bias toward
> intellectual interestingness our equivalent to emotional biases
> more prominent in other groups?)

Yes, that's a very interesting question. ;^)

Eric Watt Forste ++ ++ expectation foils perception -pcd