[ExI] alpha zero

[Dave Sill]

On Thu, Dec 7, 2017 at 5:30 PM, John Clark <johnkclark at gmail.com> wrote:
>
>
> ​
> On Thu, Dec 7, 2017 at 2:02 PM, Dylan Distasio <interzone at gmail.com>
>  wrote:
>
> ​
>  Deep learning neural nets appear to bear little resemblance to how
> biological nervous systems actually work.
>
> ​As far as Chess​
>
> ​Go and Shogi are concerned it works far better than ​
> biological nervous systems
> ​.​
>
>

Yes, in simple, well-defined domains. Computers are incredibly fast at math
but that doesn't mean they're math geniuses. I can't do billions of
floating point operations per second, but I can explain to a child in terms
it will understand what "addition" means. A CPU has no understanding of
what it does. Likewise, AlphaGO has no understanding of the games it plays.
It can't explain its strategy--it has none, it just "knows" what usually
works--and that's excessively anthropomorphic, it knows nothing: it just
does what it was programmed to do.

It a clever and useful technique but it's a far cry from a general
intelligence that can interact directly with the world where the rules
aren't all known, and communicate with other intelligent entities, evaluate
novel situations, and solve complex problems.

-Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20171207/0ab2cd8f/attachment.html>