From: Stephen Reed (reed@cyc.com)
Date: Sun May 19 2002 - 23:01:10 MDT
Hi, I'm Steve Reed and I've been working at Cycorp in Austin for the past
three years. My most recent accomplishment to advance AI was the release
of the OpenCyc 0.6 beta this April at SourceForge.
At Cycorp I was the first manager of the programming department, the first
project manager of the Darpa Rapid Knowledge Formation project, and the
instigator of OpenCyc. My current job is manager for Cycorp's
participation in the Darpa DAML project. I am fifty years old, having
received my first AI training back in 1972 with Herbert Gelertner. I kept
up with the field by reading and hobby programming, as my career developed
in commercial (COBOL) data processing. In the late 80's I started
smalltalk and OO programming, and my java programming (a cyc KB-store
clone) in 1998 brought me to the attention of Doug Lenat, who
offered me the dream job of a lifetime at Cycorp.
Cycorp is fairly well known but as an employee I can tell our story from
the perspective of the members of this mail list. We are working hard to
create a real AI. Along the way we have created a solid deductive
inference engine, the world's largest commonsense knowledge base,
incorporated a planner, and with a staff of computational linguists, made
a good start towards a hookup of Natural Language processing and the Cyc
KB.
As it turns out, the KB and the Cyc ontology are useful components
for the kinds of software that Darpa wants to create in the wake of
9-11. Our funding consequently is mainly governmental and our customers
are primarily military and intelligence community researchers -- but we
are not a classified facility and much of what we create for Darpa ends up
in OpenCyc.
In my opinion, Darpa is not yet ready to fund the Singularity mission,
although as it gets closer, the US military will catch on first and
provide funding before the commercial sector. The safety of the nation is
almost priceless.
Cycorp is spending substantial effort on our mission. We have a
staff of over 60 people in Austin, including a large number of PhD
philosophers, computational linguists and programmers. Doug Lenat is our
president and is in this for the long haul.
Cyc is programmed in SubL, using a commercial Lisp and a dialect named
SubL, whose purpose is the easy translation into C for deployed platforms,
such as OpenCyc. We have not released the source code for the runtime
library, the object store, inference engine and all the html gui
tools. But OpenCyc is mostly the same binary executable as Darpa gets and
it is freely distributable. The "open" part of OpenCyc is the KB content,
and I expect that you all see the value in that. The logical scripting
language, CycL is what you use to assert new knowledge and to make
queries.
Regarding progress towards the Singularity, my intention is make OpenCyc
available to those who can add knowledge, especially
how-to/problem-solving knowledge when the point comes that Cyc can "do
things". I would like to add to Cyc those features that enable/begin what
is called here Seed AI and Friendly AI. The DAML project seeks to
create/promote the Semantic Web, and I will position OpenCyc as a Semantic
Web Server, connected to a "Knowledge Grid" via a (super) peer-to-peer
network, in the remaining years of the project.
My view is that achieving an AI will continue to be hard work, with
many false starts, dead-ends, refactorings and rewrites. I dream that we
have the hardware now to create an AI, if only we could harness enough of
it -- with the right (seed) software.
Regards to all,
-Steve
--
===========================================================
Stephen L. Reed phone: 512.342.4036
Cycorp, Suite 100 fax: 512.342.4040
3721 Executive Center Drive email: reed@cyc.com
Austin, TX 78731 web: http://www.cyc.com
download OpenCyc at http://www.opencyc.org
===========================================================
This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:39 MDT