[alicebot-archcomm] Long Term Fate of AIML

[Christopher Fahey [askrom]]

> Everyone from my psychiatrist to
> journalists to the CEOs of software companies was telling me, 
> "get this guy off your Board."


So now it's "your Board", huh? Is that what you've been telling people? 

I'm very sorry to have to write this email, but I fear that if I and
others don't speak up now and begin to develop alternative ideas to
AIML's current disastrous course, we will simply watch all we've worked
for quickly spiral into nothingness.

I'm greatly concerned that some members of the AIML Architecture
Committee, and the AIML community in general, have unquestioningly lined
up behind Richard in his latest haphazard and obviously unilateral
plans. It's totally depressing to me. I have completely lost confidence
in the ability of the Alicebot AI Foundation (and particularly in Dr.
Wallace) to develop the AIML community or the AIML specification in a
positive direction any more.

I am firmly of the opinion that AIML is now destined to one of these two
fates:

  1) Richard, acting as a one-man foundation (but perhaps eventually
with some poor souls to help him) will flail away at trying to get
people to regain confidence in the now-floundering AIML spec. This will
fail and AIML will freeze at it's current version 1.0. Developers of
Programs D, E, P, etc, will no longer have any confidence in a stable
AIML standard to develop towards. Current and future AIML bot authors
will learn to live with the current August 2002 versions of the various
AIML parsers. In fact, I think this is effectively where we are now: I
can't imagine the wisdom of continuing to develop an AIML parser project
(or any AIML project) at this point since it's become clear to us all
that AIML is ultimately under the control of whatever entity Richard
Wallace chooses to call his Board, even if it's just him.

  2) One or more alternative AIML organizations will emerge, possibly
some flavor of open-source but also possibly commercial. Such
organizations won't have Richard Wallace to kick around, but instead
will endeavor to agree on an AIML standard between them in much the same
way corporations today work with each other to establish standards.
Richard will do battle with this organization or coalition, claiming
that he owns the acronym AIML and the concepts AIML reflects. I hope
that Richard will not do this and will stand behind his original opinion
that AIML was a gift to the world. Richard might even support this or
these organizations (his support is certainly meaningful in a symbolic
way), although after these past few weeks I doubt any of them would
offer him any decision-making capability.

It is in Richard's best interest for #2 to happen. It is better to be
like Rasmus Lerdorf (inventor of PHP) who has allowed his role in the
PHP project to become less and less central as others have come along to
develop it further
(http://safari.oreilly.com/main.asp?bookname=progphp&snode=11) than to
be known as the guy who gave birth to, nurtured, and then destroyed a
great idea.

All I care about is the AIML spec. It's the core of what the foundation
was supposed to evolve and support, and to me everything else was
extraneous. I never understood why we needed to have a foundation in the
first place, and quite frankly I think the alicebot.org site was doing
too much: promoting AIML, promoting Richard Wallace, promoting ALICE,
teaching AIML, providing AI news, being the home page for Program D
development, etc, etc. IMHO, the alicebot.org site (or whatever
non-Richard-Wallace-centric successor) should look something like this:

    http://www.w3c.org/MarkUp/html-spec/html-spec_toc.html

I only hope that the various alicebot.org mailing lists manage to
survive long enough for somebody to set up a successor group for the
community, because I strongly fear that Richard's current actions will
lead to alicebot.org's total destruction.

-Cf

[christopher eli fahey]
art: http://www.graphpaper.com
sci: http://www.askrom.com
biz: http://www.behaviordesign.com