home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!cs.utexas.edu!swrinde!mips!mips!munnari.oz.au!ariel!ucsvc.ucs.unimelb.edu.au!lugb!latcs1!jacob
- From: jacob@latcs1.lat.oz.au (Jacob L. Cybulski)
- Newsgroups: comp.ai
- Subject: Re: AI Winter Refugees
- Message-ID: <1992Aug27.010520.17749@latcs1.lat.oz.au>
- Date: 27 Aug 92 01:05:20 GMT
- References: <x+an!9a.vere@netcom.com>
- Organization: Comp Sci, La Trobe Uni, Australia
- Lines: 41
-
- > o what were the causes of AI Winter?
-
- There used to be the time that AI PhDs were regarded not only as great
- researchers but also as "bloody good programmers", today's AI PhD
- graduates became virtually unemployable outside academia as the major
- thrust of their research work shifted into the theoretical foundations
- of AI, so they simply became "not so good logicians". In my opinion,
- the current AI Winter is a direct result of practices by IJCAI and
- AAAI conference (and journal) committees who promote the work by
- neaties (theorists) rather than by scruffies (software engineers).
-
- > o why has AI gone out of fashion in the US industrial sector?
-
- Most of the hot-shot AI research is based on small, unscalable
- problems. Lots of promise, lots of funds, and no delivery. It seems
- that people from outside of AI but who use some of the AI techniques
- are much more successful, e.g. in databases, software engineering,
- programming languages, etc. They develop applications which start big
- and then the clever AI tricks are used to address very small, well
- defined and narrow problems (embedded AI).
-
- > o stories of AI professionals or recent graduates who are or have been
- > unemployed or underemployed due to AI Winter;
-
- I have met a number of AI wizards who specialised to the extent of
- total unemployability. Some people who were experts in non-monotonic
- logic, belief revisions, or default reasoning but who could see no
- practical application of their own work to solve any real problem,
- even when using Expert System shells or an AI language. I met
- graduates who believed that a 5000 line lisp program is big and many
- more whose PhD was based on work amounting to mere 100s lines of code,
- the so called demonstrations of concepts.
-
- Jacob
-
- =============================================================================
- Jacob L. Cybulski, Deputy Director, Amdahl Australian Intelligent Tools
- Programme, Dept of Comp Sci & Comp Eng, La Trobe University, Bundoora,
- Vic. 3083, Australia. Ph: +613 479 1270, Fax: +613 470 4915, Telex: AA 33143,
- E-Mail: jacob@latcs1.lat.oz.au. ... G'Day
- =============================================================================
-