SUO: Re: OpenCyc Announcement
Give away Cyc? I'll believe it when I see it. Read the fine print.
Doug has also, apparently, announced that this give-away will be done
using DAML+OIL, which is not capable of expressing about 90% of the
content of the Cyc ontology.
On the other hand, there is no doubt that there is a LOT of stuff in
Cyc, and much of it is very thoroughly debugged. Cycorp is also
working very hard right now on building a usable NL interface, by the
way, in the context of the DARPA-funded RKF (rapid knowledge
formation) project, so I suspect that a critique written in 1996 is
now largely irrelevant. I know they have every confidence that they
will become the 'standard' or at least the state-of-the-art for
effective use of the 'semantic web', rather in the way that Dolby
became the de facto standard for audio technology just by being so
much better than anyone else at just the right time. However, being
the world's largest repository of 'common sense' is one thing, and
being a usable industrial standard is something else.
If the SUO sees itself as competing with Cyc, then we might as well
all just save our effort and do something else. If we see ourselves
as having fun, of course, then that is a different matter altogether.
Pat Hayes
---------------------------------------------------------------------
IHMC (850)434 8903 home
40 South Alcaniz St. (850)202 4416 office
Pensacola, FL 32501 (850)202 4440 fax
phayes@ai.uwf.edu
http://www.coginst.uwf.edu/~phayes