bobgian@psuvax.UUCP (01/08/84)
When I said in the course description that "this course has no teacher"
I meant it! I will lead discussions here at Penn State and function
as a "mailing list moderator", but I'm HARDLY the world's expert on all
the syllabus topics. (I'm an Instructor here while finishing a PhD at
MIT.) One reason for trying an experiment like this is to increase all
of our exposure (and thus learning rate) to material in these areas.
I hope the REAL experts will come forward and share their ideas and
experience with me and all the others following the course.
To anticipate a few questions:
How can people who are not Penn State students participate?
(i.e., does it cost, do we get a certificate at the end, etc?)
No cost, no certificate. The "netwide" component is simply for the sake
of fun, learning, and sharing of our collective expertise.
What background do you require of the participants: a little
previous exposure to AI topics, a thorough grounding in AI,
an afternoon with LOGO?
The background of the Penn State students is one term of an introductory
undergrad AI course. (I posted the syllabus and some exams from that
course to net.ai about two weeks ago.) I hope (this is ambitious, I know)
to make the course understandable and useful for rank novices while still
interesting to grad students and pros.
Is it possible to merely audit your seminar, to do the readings
and follow discussions, without being required to produce a
large system or other project?
That is exactly the idea. That applies to the Penn State students too.
There are NO requirements for the course other than a sincere interest
in participating in an experimental endeavor. No assignments or projects
or exams required. Hopefully, many people WILL have the time and interest
to contribute ideas, proposals, rebuttals, suggestions, criticisms, software,
literature references, and so on, but nobody (even the local students) will
HAVE to do anything. This is education for the pure fun of it.
Well, how SHALL we do it? I'm considering two alternatives: (1) creating
a USENET newsgroup for unrestricted discussions (the "worldwide seminar"
to which anybody can listen and speak) or (2) asking people to mail
material to me. There are obvious pros/cons:
(1) pro: Everybody can participate on an equal basis, with no
"censorship" by me. (Who am I to censor, anyway?)
(1) con: Potentially huge traffic volume ==> large phone bills.
(2) pro: I can "filter" and edit submissions to correlate closely
with the course at Penn State and impose some "order"
on the whole thing.
(2) con: What right do I have to "bury" other people's submissions
just because I differ with their "relevance" evaluation?
PLEASE LET ME KNOW (mail) IF YOU HAVE STRONG FEELINGS ON THIS ISSUE!!!
Regardless of the (1) vs (2) choice, I intend to "abstract" out the
best of the submissions, edit them together with local originations
into coherent "position papers", and repost (with full credit to all
authors, of course) to net.ai and AIList. That way USENET and Arpa
members can participate without having to "wade through" all the "raw"
discussions. That will keep the burden low (and interest high, I hope)
among our net experts who are likely to be very busy people.
Bob
--
Bob Giansiracusa (Dept of Computer Science, Penn State Univ, 814-865-9507)
UUCP: bobgian@psuvax.UUCP -or- allegra!psuvax!bobgian
Arpa: bobgian@PSUVAX1 -or- bobgian%psuvax1.bitnet@Berkeley
Bitnet: bobgian@PSUVAX1.BITNET CSnet: bobgian@penn-state.csnet
USnail: 333 Whitmore Lab, Penn State Univ, University Park, PA 16802