Genius Dogs

John K Clark (
Fri, 10 Oct 1997 10:10:12 -0700 (PDT)


On Thu, 9 Oct 1997 Hal Finney <> Wrote:

>Just simulating the *brains* which have existed would not be enough
>to evolve intelligence.

I agree.

>A brute force technique, then, would have to simulate not only the
>brains, but potentially a significant subset of the biosphere. This
>would be many orders of magnitude more difficult. Volume
>considerations alone would suggest 4 or 5 orders of magnitude

I don't agree. If we could simulate brains, the most complex object in the
known universe, then I don't think other objects would be an insuperable
problem, nor do I think volume is necessarily proportional to simulation
difficulty. We wouldn't have to simulate an environment exactly as it was on
any given day on Earth, we'd only need to make it logical and complex enough
to be interesting.

Take the weather for example, even today we have pretty good models of it,
we can't make a good prediction about what it will do a month from now and I
doubt we ever will, but that's because of our ignorance of initial conditions
and the weather's exquisite sensitivity to it. If you gave a meteorologist a
description of the atmosphere for the next month based on a good computer
model and a description of what the atmosphere actually did the two would be
very different, but unless he had access to external instruments he couldn't
tell which one was real and which one was the computer model. The model
provided a completely logical, complex, and realistic description of what the
atmosphere might have done, it's just that instead it did something else that
was also logical and complex.

Another thought, perhaps we could give our virtual creatures an advantage
nature figured out how to give real ones, instead of relying exclusively on
Darwin's style of Evolution with its horrible random mutation and natural
selection, perhaps we could also have Lamarckian style Evolution by the
inheritance of acquired characteristics. It would be wonderful if that the
muscular body you worked so hard to develop, would make your children strong
too. How splendid it would be if your offspring were born knowing everything
you struggled to learn during your life. Unfortunately it doesn't work that
way in our world, Shakespeare's daughter had to start from square one and
learn how to hold a pen and make sense from squiggles written on paper,
Einstein's son had to reinvent the wheel and rediscover that 2 +2 = 4.
Besides being less cruel, Lamarck's type of evolution would be much faster
than Darwinian evolution, that's why social evolution is so much swifter than
biological evolution.

John K Clark

Version: 2.6.i