Computing Evolution

[Derek Broughton]

Christopher Chan wrote:

> Brian Fahrlander wrote:
>> Chan Chung Hang Christopher wrote:
>>> We cannot even fully understand how the brain works let alone how the
>>> mind exists and you want to talk about artificial INTELLIGENCE?
>>>   
>>     Did we understand how the brain works when creating a desktop
>> metaphor starting with simple switches and vacuum tubes? People tend to
>> think a) Everything they do has a clear goal, and b) It's worth reaching.
> 
> Haha, that's it, compare making a bunch of logic gates to the pipe dream
> of creating an intelligence. A mathematical challenge to creating a
> mind. People really should start using proper descriptions of
> preprogrammed 'intelligence' instead of spouting nonsense.

Really?  Neural nets, while not able to produce true artificial intelligence, 
are entirely capable of a number of the basic actions we _think_ must be 
part of intelligence, and they do it without clear goals.  So, no, there's 
no reason to assume we'll need to fully understand the brain before we can 
talk about (let alone create) artificial intelligence.
>> 
>>     It's not about neurons; it's about logic. 

No.  It's very much about neurons.

>>     Objective logic. When the
>> newspaper comes do we ever answer the phone to retrieve it? When the
>> waiter brings food, do we go to the garage to get it installed?  Life as
>> we know it has a 'way' about it.

Indeed it does - and logic rarely enters into it.  It's about training 
neurons to act in certain ways.  The first time the newpaper comes, we 
probably look out more than one window first, then we go to the right door.  
Eventually, we just go straight to the right door.