I think Richard's point is interesting.
I think the brain probably does NOT do implicit probabilities, at least not with high precisions. If it does, I see no reason why such information is not available to consciousness.
The brain probably performs inference using categories like "some", "most", "a few", or "likely", "highly likely", etc. Such categories may be regarded as probabilities with extremely low precision (like 1-2 bits).
So, the distinction between Richard's categorical inference and numerical inference is one of precision.
Now the question is, does using categories improve performance greatly? How much resource can we free up by using 1-2 bit precision instead of floating points?
YKY
To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
