Re: "analog computer" = useless hypothesis?

From: James Rogers (jamesr@best.com)
Date: Sun Apr 01 2001 - 21:27:49 MDT


On 4/1/01 4:01 PM, "Eliezer S. Yudkowsky" <sentience@pobox.com> wrote:
> Anders Sandberg wrote:
>>
>> Well, remember that timing of neural processes is limited by the speed
>> of neural interactions (roughly ~1 millisecond). It is doubtful that
>> precision greater than that is possible without employing many neurons
>> doing clever things.
>
> If I recall correctly, bats can do ~1 microsecond discrimination as part
> of how they process echoes - specifically, IIRC, bats can discriminate a
> one-microsecond difference in echo arrival time between one ear and the
> other. This would, however, be in the category of "many neurons doing
> clever things".

Mammals (including the bats and cetaceans) all have the same detection
hardware (quite good as such things go in the animal kingdom), which has a
sampling rate of around 10 microseconds. All effective differences in
ability are based in how the neural hardware is organized. Microbats are an
extreme example of a highly optimized mammalian auditory cortex, being able
to detect phase variations of less than 100 nanoseconds (i.e. Differences in
signal arrival between ears). However, since phase discrimination is
limited in part by the maximum sampling rate of the sensor hardware,
microbats represent the rough theoretical limit for performance in mammals.
Certain types of fish can best this by around an order of magnitude.

Well-trained human ears probably have a phase detection limit of around 10
microseconds (back-of-envelope calc), though for most of us it is closer to
100 microseconds. It has been demonstrated that the human auditory system
is actually quite trainable in terms of its sensitivity, varying somewhere
around two orders of magnitude from poor to quite good relative to most
other animals.

-James Rogers
 jamesr@best.com



This archive was generated by hypermail 2.1.5 : Sat Nov 02 2002 - 08:06:48 MST