[ExI] DeepMind wins Game1 in Go championship Match

[Dave Sill]

On Wed, Mar 9, 2016 at 8:22 PM, spike <spike66 at att.net> wrote:

>
> Dave Sill wrote:
>
> >…If all you want is a system that can play chess or go, that's clearly
> doable. But these are more simulated intelligence than artificial
> intelligence…
>
> Ja.  Any task transitions from intelligence to simulated intelligence the
> minute we figure out the algorithm to do it.
>
No, that's not what I mean. Building a table of all of the possible
tic-tac-toe games and playing perfectly by consulting it isn't
intelligence. Intelligence is determining that by constructing such a table
one can play perfectly.


> >… Produce a system that can be taught any game the way a human learns it,
> and can learn to play it well via playing and studying the game, and *that*
> will be AI…-Dave
>
>
>
> On the contrary sir, for this has been done, first with tic-tac-toe and
> later with checkers.  I haven’t checked on later progress, but I think the
> algorithms never get all that good just by looking at games.
>
Firstly, no, this hasn't been done. I said "Produce a system that can be
taught any game the way a human learns it". Has not been done. Second, the
fact that "algorithms never get all that good just by looking at games" is
exactly my point: there's no inherent intelligence in the algorithms.

Siri is already smarter and a better conversationalist than most teenagers,
> most of whom seem to communicate using a lookup table.  The point is we *
> *still** don’t have a working definition for what is machine
> intelligence, for we insist on constantly moving the goalposts.
>

Doing something that seems clever but isn't really is a neat trick, but
human-level intelligence is much, much more than that. The goalposts
haven't moved, we just haven't made any real progress in AI.

-Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20160310/911fe340/attachment.html>