Any progress towards AI at all? was Re: [extropy-chat] Futures Past

[Russell Wallace]

On 10/10/05, Mike Lorrey <mlorrey at yahoo.com> wrote:
>
> Well, these days researchers are building insect and fish equivalent
> brains and robots. I suspect we could get a pretty accurate estimate of
> the appearance of human equivalent AI (not brain scanned) by tracking
> how quickly research robots develop along the evolutionary path between
> cambrian period insects and present day humans.


I suspect we couldn't, because researchers don't even try to duplicate _all_
the functionality of any life form (it would be entirely impossible with
today's technology anyway), and once you're going for only partial
functionality, then apparent progress depends more on how narrow a slice you
accept, than on the sophistication of the machine. (The extreme case being
when human-equivalent AI was considered to be around the corner based on
computers that could do algebra and play chess.)

Remember that Moore's Law predicts human level general processing power
> coming around 2025-2030, but not the software or hardware logic that
> makes grey matter human. Ten to the whatever transistors or petaflops
> or what have you doesn't mean much sitting in a jar.


Yep.

- Russell
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20051010/63f51614/attachment.html>