[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: what most every language is missing :-)

Dan Sugalski wrote:
> With more interesting systems it can also be used as a spot to split
> the execution stream and assign weights to the new streams, deferring
> final evaluation of the results until you've enough information, or
> have reached a stopping point and have to generate some random
> numbers, to collapse the system to a final result.

With that, I imagine a neural network.  Perhaps this line of thinking
could produce programs that deal not with "factual information", but
with opinions:

	sunset =? pretty -> true (in my opinion)
	sunset =? pretty -> false (in my opinion)

There are certain pieces of factual information that contribute to
the result, along with random/arbitrary values to fill in where
facts are unavailable.  You could even include an amount of
"social" influence -- if another computer/execution stream answers
the same question first, then that answer could play a role in
the outcome.

Of course, I'm not a neural network researcher (not now, at least),
and maybe this is all trivial to those who are...?  :-)

 -- Trevis