Seminar in Artificial Intelligence, 3.0 ETCS / 2.0 Std. (WS 2015/16)
Did Gödel show that intelligence cannot be fully mechanized?
This course will be held in English.
Last update: October 7, 2016 (CF)
Check regularly for UPDATES of this page!
Organisational meeting (moved) takes place on
Friday, October 21, 2016, 13:00-13:30
Von Neumann Seminar Room,
Favoritenstraße 9-11, ground floor
Topic and Prerequisites:
Gödel's famous incompleteness theorem
is one of the most frequently (mis)used results in the context of
discussions about the possibility of 'strong AI'.
Thousands of attempts to show that Gödel's results have severe consequences
for the prospects of 'mechanizing intellingence' have appeared in books,
papers and contributions to discussions on the internet. Everyone
seriously interested in Artificial Intelligence should be able
to assess this debate.
The main aim of the seminar is to learn to critcially and
different kinds of interpretations of Gödel's theorem. Needless to
say that this calls for some familiarity with basic
notions from mathematical logic and a
careful analysis of basic concepts in AI and related fields.
Background of all presentations and discussions will be the book
Torkel Franzen: Gödel's Theorem - An Incomplete Guide to its Use and Abuse
which does not presume any specific previous knowlegde of its readers,
but is also highly esteemed among experts.
Wellesley, Massachusetts, A K Peters, Ltd., 2005. x + 172 pp.,
Click here for infomation of the publisher
There is no need to buy the book.
Electronic version available here (password protected).
- Participants are required to prepare a
short paper (8-10 pages) on
the topic assigned at the
- It is strongly recommended to use
the preparing seminar paper.
In any case, the paper has to satisfy usual academic standards
regarding citations, references, structure etc.
Participants should bear in mind that
their papers will be carefully checked by at least some
of their colleagues (see below)
and, moreover, will be
made available to everybody on the Web.
- Important: deadline
for the submission of seminar papers is currently planned for
Thursday, December 1, 2016
Send your papers as PDF to
email@example.com (subject: Seminar in AI essay).
- The presentations are on the same topic as the seminar papers.
Note that the talks take place after the papers
will have been made available to all participants.
(In January - details will follow later)
- Each presentation has to be accompanied by a one page handout.
The handout has to contain your data (name, Matrikelnummer, etc)
and should concisely
sum up the key concepts introduced in your talk. It also has
to contain full references to relevant literature.
(You can either bring along enough copies of the handout at the
presentation yourself or send a PDF to
firstname.lastname@example.org (subject: Seminar in AI handout) a day in advance.)
- Each participant will be assigned two papers of other participants for
The purpose of this part of the seminar is to train your
ability to evaluate and comment on seminar contributions by
other participants in a systematic and structured manner.
(See Guidelines, below).
Reviewing abilities are of utmost importance in a scientific
context. All researchers, including `beginners', are regularly
asked to formally act (anonymously) as referees/reviewers of
workshop abstracts, posters, research papers, research proposals
etc. submitted by other researchers. Peer reviewing - as this
part of scientific working is called - is an essential ingredient
of scientific progress in general and of each researcher's individual
career in particular.
- You will receive an Email containing two numbers,
Monday, December 5, 2016.
These numbers refer to the
two essays that you are assigned to review, as listed under
be below under `Further meetings' as downloadable PDFs.
- You should write a detailed referee report on the two assigned
essays using the prepared Report Form (click here).
You may add (but not delete) as many additional lines as you wish.
- IMPORTANT: Prepare your reports in ASCII format (i.e., pure
text: no .doc, .pdf, .ps files!) and send them
as separate emails to
(Subject: SemAI Report No. ...).
Don't forget to add the corresponding number of the reviewed essay
to the subject field of your email.
- Deadline for submitting the reports is
Monday, January 9, 2016
- IMPORTANT: the evaluation of your own seminar essay
by other participants will not be the basis for the
overall evaluation (credits) of your performance in the seminar.
Rather, the quality of your own evaluation of other essays will
influence the credits assigned.
The dates will be fixed at the first meeting (see above).
The evaluation (Benotung)
will be based on the quality of the following four
types of contributions:
- your seminar talk,
- your seminar paper,
- your evaluation of other seminar papers,
- your contributions to the discussions.
Seminar papers, but also other comments, requests, etc., should be addressed to
Check regularly for UPDATES of this page!
[ LVAs 185/2
| Abteilung 185/2
| Institut 185
| TU Wien
| Server home page