From newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rpi!think.com!mips!mips!munnari.oz.au!uunet!trwacs!erwin Mon Jun 15 16:05:04 EDT 1992
Article 6247 of comp.ai.philosophy:
Path: newshub.ccs.yorku.ca!ists!torn.onet.on.ca!utgpu!news-server.csri.toronto.edu!rpi!think.com!mips!mips!munnari.oz.au!uunet!trwacs!erwin
>From: erwin@trwacs.fp.trw.com (Harry Erwin)
Newsgroups: comp.ai.philosophy
Subject: Re: Transducers
Message-ID: <631@trwacs.fp.trw.com>
Date: 14 Jun 92 00:56:00 GMT
References: <60806@aurs01.UUCP> <1992Jun11.182144.12157@mp.cs.niu.edu> <60810@aurs01.UUCP> <1992Jun12.192537.32302@mp.cs.niu.edu>
Organization: TRW Systems Division, Fairfax VA
Lines: 39

rickert@mp.cs.niu.edu (Neil Rickert) writes:

>In article <60810@aurs01.UUCP> throop@aurs01.UUCP (Wayne Throop) writes:
>>
>>The other issue is that, presumably, Neil thinks that current machines
>>aren't up to frog-level of capabilities.

>  Yes.  That was my point.

>>                                          I'd be interested in an
>>elaboration of this, in particular what do frogs got (in their functional
>>capabilities) that at least some current computer processes ain't got
>>(or at least got things very closely akin)?

> Here goes.  Please understand that this is a gross oversimplification.

> Let's look at a hierarchy of intelligence.

>  0:	A rock.  Not much here.
>  1:	A single celled creature (a protozoan for example).  Not bad
>	considering that it has only one cell.
Interestingly, we're having trouble understanding level 1. We know that we
can train Paramecia to run a very simple maze. The neuroskeleton is
apparently made up of actin microtubules, but we can't get close enough to
understand how it transmits and learns information. We do know that the
neuronal neuroskeleton is involved in some fashion with intra-neuronal
data processing, but that's about it... On the other hand, as long as you
avoid sub-neuronal issues, we seem to be getting moderately close to
understanding how owls operate. One interesting point is that some simple
data processing primitives require complex networks to implement. An
example is estimation of the time delay between sounds reaching the two
ears. In owls, this involves a fairly complex net involving both sides of
the brain. 

Cheers,
-- 
Harry Erwin
Internet: erwin@trwacs.fp.trw.com



