Re: Future Tech.

Anders Sandberg (nv91-asa@nada.kth.se)
Thu, 8 May 1997 13:52:55 +0200 (MET DST)


On Wed, 7 May 1997, MikeRose wrote:

> On Wed, 7 May 1997, Anders Sandberg wrote:
>
> > ... the newspapers are uncritical. The artificial brain (the "Soulsucker
> > 2000 Chip") was intended as a dramatic and humoristic ending of their
> > talk, but some journalists thought it was reality, and then other
> > journalists copied them, and so on.
>
> And thus, an urban legend was born. The fact it stuck, however, must mean
> there is some human desire for it to be true.

Clearly. And I think the discussions on this list is good evidence that
there are many who desire this kind of technology.

> How true is the old adage that a computer can only be as clever as the
> programmer who designed it?

No, it is a fallacy which is surprisingly common. The oldest and most
obvious counterexample is IMHO Samuel's checker program, which was
significantly better than him. I'm constantly amazed (and dismayed) by
what the neural networks I play with do, and sometimes they come up with
obvious solutions to problems I did not think of.

> How far has self-learning computing progressed?

Depends on what kind of learning you refer to. I don't know enough about
classical AI, but EURISKO-style heuristic competition is quite similar to
genetic algorithms, a field where there is plenty of exciting results.
Neural networks learn fine, but we still have to find good input
representations and architectures for them.

-----------------------------------------------------------------------
Anders Sandberg Towards Ascension!
nv91-asa@nada.kth.se http://www.nada.kth.se/~nv91-asa/main.html
GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y