Author: Matthew Hull
Date: 19:01:47 08/05/02
Go up one level in this thread
On August 05, 2002 at 17:56:59, Dann Corbit wrote:
>On August 05, 2002 at 17:27:39, Matthew Hull wrote:
>
>>On August 05, 2002 at 17:13:12, Dann Corbit wrote:
>>
>>>On August 05, 2002 at 16:28:56, Russell Reagan wrote:
>>>
>>>>On August 05, 2002 at 16:06:26, Dann Corbit wrote:
>>>>
>>>>>Computers have absolutely no cognition
>>>>>of any thing at this point in time. Before long, that will change.
>>>>
>>>>How will it change? Are there any particular technologies you forsee being big
>>>>players in the change? Or is it just a feeling that since technology is
>>>>advancing, it's bound to happen sooner or later?
>>>
>>>The brain of a honeybee does about 10 gigaflops (about the same power as the
>>>retina of your eyeball, for that matter). When computers average that power, it
>>>will be possible to begin rudimentary experiments in _real_ machine
>>>intelligence. Not far away.
>>>
>>>At some point, computers will have more processing power than a human brain. By
>>>that time, it will be possible to make a machine that thinks.
>>
>>I would submit that if they are Turing machines, the quantity of them will not
>>suffice to overcome basic computability problems. Penrose has demonstrated this
>>in his book, The Emperor's New Mind.
>
>That's OK. We (as humans) can't overcome the basic computability problems
>either. I don't see why the machines should have the upper hand in that.
That's a nice little bit of sophistry there.
One thing is for sure, Turing machines can't get you there ("there" being strong
AI), no matter how fast they are or how many you can get to work on a problem.
Because processing power is not even the problem. The problem is in the
limitations of algorithmic processes. There are IMPORTANT classes of problems
that can't be solved algorithmically. This is _the_ critical limitation of the
Turing machine. No two ways about it. :-\
Regards,
Matt
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.