From: J. Andrew Rogers (andrew@ceruleansystems.com)
Date: Tue Jun 01 2004 - 00:07:06 MDT
On May 30, 2004, at 6:57 AM, fudley wrote:
> The fastest signals in the human brain creep along at a top speed of
> about 100 meters a second; many are slower, some much much slower
> (chemical signals moving by diffusion). Light travels at 300,000,000
> meters a second, 300,000 times faster than the very fastest human brain
> signals. At the extreme two areas of the brain could be about a foot
> apart, so if two AI’s were 300,000 feet apart, about 150 miles, they’d
> have no more problem with fractured identity issues than we have, and I
> expect you could go many times that distance before things got serious.
You rephrased the question and dropped the most important variable
somewhere in the process.
Unless you've somehow figured out a way to push a billion lambdas over
your communication medium, we are a long way from doing what the brain
can do. You can't multiplex tens of billions of channels over a single
channel and expect to see the same latency as a single channel.
Multiplexing has a latency price tag, but you are treating it as though
it comes free. It is non-trivial to do a brain's worth of message
delivery on the same piece of modern silicon, never mind over a 1000
miles.
j. andrew rogers
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:47 MDT