[ExI] fun outsider's view on ai

spike spike66 at att.net
Tue May 10 03:54:24 UTC 2016



From: extropy-chat [mailto:extropy-chat-bounces at lists.extropy.org] On Behalf Of William Flynn Wallace


>…OK, now I am getting somewhere.  At the start I assume you define just what it is that you want it to do.  What the whole process seems to me to be is to create a machine that can duplicate a human, only it is faster, not different in quality in any way. (Is this correct?)  And that's because we can't think of any way to think other than human because we are human and limited by that…  BillW



Not necessarily faster than a human, but perhaps more scalable.  We don’t know for sure what it will do.  If we manage to create a sim of a brain, we can be pretty sure it will be slower than humans, because of the signal latency inherent in running a bunch of processors in parallel.


We have no concept of a machine which can come up with an idea then code it in software.  But we are living inside a machine that does exactly that.  So… we can use existing machines to try to create a simulation of a human brain, then see if we can get it to write software, and if so, can it optimize itself.


Human brains don’t really scale, but if we could sim a small brain, such as a nematode, then we could perhaps sim the brain of a bee, a mouse, a human, and perhaps we could keep scaling it right on up.


We don’t know what will happen. 







-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20160509/293f996b/attachment-0001.html>

More information about the extropy-chat mailing list