AI techniques in the Future

[chris-coleman]

Can anyone point me in the direction of any papers/articles discussing which
ai techniques will survive in the future and which will not.

Many Thanks
C

[Doc O'Leary]

In article <Cyro6.9010$925.8...@news6-win.server.ntlworld.com>,
chris-coleman <chris-...@ntlworld.com> wrote:

> Can anyone point me in the direction of any papers/articles discussing which
> ai techniques will survive in the future and which will not.

I don't think this needs any involved article. AI has become a
classification for a large number of small moving targets, so it's
unclear what techniques will eventually prove to be on the right path
and which are dead ends, to say nothing of approaches that have not
been tried yet.

I'd question your reason asking. Are you an author working on sci-fi
story? A new graduate looking to focus on a field of study? A game
developer writing a computer player?

[Don Tveter]

Anything that works now will work in the future. AI now does the
little specialized problems and even in the future the well-versed
practicioner of AI techniques will have to know them all so they can
choose just the right one.

If you take the original definition of AI -- the goal of creating
machines roughly equivalent to people -- probably nothing now will
be of much use in the future.

--

************************************************************************
The Pattern Recognition Basis of Artificial Intelligence
Backpropagator's Review NN freeware for UNIX and PCs
A Professional BP Version for X and W95
************************************************************************
Don Tveter d...@unidial.com http://www.dontveter.com
************************************************************************

[bob nerkul]

On 5 Mar 2001 11:51:14 +1100, "chris-coleman"
<chris-...@ntlworld.com> wrote:

Minsky's 1990 paper online here:
http://www.ai.mit.edu/people/minsky/papers/SymbolicVs.Connectionist.txt

nerkul