AI Winter Refugees
Danny Faught
a (small) section on careers in AI. It says, basically, not to do it,
that there isn't much of a market for AI professionals. It suggests
that people interested in AI should apply AI techniques in one of the
more common computer career areas.
Maybe the AI people are the hardest hit in the alleged decline in the
total population of scientists. Hal Hellman, in his 1976 book,
_Technophobia_, argues that the number of scientists is declining,
and the number of genuinely new fields of science are being exhausted.
"I once asked an editor of _Scientific American_ why their articles were
often so difficult to read. His answer, only half in jest, was that all
the simple subjects had been used up. I wonder if that doesn't reflect
the actual situation in scientific and technological research." (p. 12)
--
Danny Faught da...@ponder.csci.unt.edu
Save this sig - I'll be famous someday
Tom Fawcett
>AI should be viewed as science, rather than engineering.
>This way, there will be a lot more solid basic work, a lot less
>disappointment.
I would hope there's room for both science and engineering. Otherwise the job
prospects for people in AI will be pretty much like those for theoretical
physicists. (well, we may be at that point now anyway).
Funding aspects aside, science and engineering feed one another. Some
fundamental problems are uncovered and delineated when researchers try to
apply existing techniques to real-world problems.
>The disapointment results from the inflated hopes that AI can
>create magic with no need for a fundamental understanding of intelligence.
The problem is with the "inflated hopes that AI can create magic" -- with or
without a fundamental understanding of intelligence.
-Tom
John Nagle
>The ACM's guide to computing careers (can't remember the exact name) has
>a (small) section on careers in AI. It says, basically, not to do it,
>that there isn't much of a market for AI professionals. It suggests
>that people interested in AI should apply AI techniques in one of the
>more common computer career areas.
I wonder what happened to those people who studied for degrees in
"Knowledge Engineering" at Stanford in the mid-80s. This was
back when Feigenbaum was appearing on national TV saying the US would
become an agrarian nation if the government didn't finance his proposed
national AI center, and people were really worried about the Japanese
Fifth Generation effort.
"Will write rules for food", maybe?
John Nagle
Tom Fawcett
>Finally, let me quibble a bit with your use of the term "refugees",
>which brings to mind innocent victims of a disaster not of their
>making. My perception is that the people being justifiably pruned
>from AI overlap substantially with those responsible for the hype that
>surrounded the field in less responsible days.
His original message mentioned under-employment among AI professionals. It
sounds like anyone graduating with a PhD in AI within the last few years would
qualify as a "victim" of this. I doubt they are the ones responsible for the
AI hype of the 70's.
-Tom
Roberto Desimone
>
>o what were the causes of AI Winter?
>o why has AI gone out of fashion in the US industrial sector?
I'm not sure that this is true. Certainly there is a greater trend for
larger US companies to reduce the amount of fundamental AI research
they are pursuing, and, instead, pursue more applied (shorter-term) AI
research and development.
>
> Countervaling views are welcome, but please be prepared to explain
>the following phenomena: near-zero employment opportunities advertised
>in AI Magazine and similar forums; near-zero recruiting at recent AAAI
>conferences; major declines in attendance at AAAI over the past
>several years.
Possible (partial) explanations for these phenomena are that there are
many more people around with AI experience (albeit at widely varying levels
of expertize), and so considerably more people than before going for a
slightly larger number of jobs. Whereas in the past (only a few years ago)
companies had to advertize to find good people to fill their staff positions,
now they are getting plenty of good resumes and don't feel they have to
advertize so much. Likewise for recruiting at AAAI.
Regarding attendance at AAAIs. Again there are many more AI conferences now.
For applied AI, here are a few that are well attended:
IAAI - Innovative Applications of AI
CAIA - Conf. on AI Applications (IEEE)
IEA-AIE - Industrial and Engineering Applications of AI & Expert Systems
Then there are the more specific (sub-field) conferences and workshops.
I'm not stating that the above totally explains the phenomena you describe.
I'm sure that some companies are thinking twice about expanding their AI
staff, others are even shrinking. But, on the whole, I don't see a widespread
reduction in AI staff. I tend to keep in touch with the more applied side of
AI, rather than the pure research end of AI, so maybe I'm getting a biased
view.
--
Roberto Desimone rob...@erg.sri.com Tel: (415) 859-4038
SRI International (EK335), 333 Ravenswood Avenue, Menlo Park, CA 94025
Pat Prosser
exception it has come from "the good old US of A".
Europe, and UK in particular, had its AI winter in the late 70's due to
the Lighthill Report. Lighthill, an advisor to Her Majesty's Government
concluded that AI was pissing into the wind. All funding dried up.
AI research then picked up due to the scaremongering tactics of E Fiegenbaum
and the 5th Generation, courtesy of the Alvey Directive and ESPRIT in
Europe. So we have a lot to thank the Japanese and E.F. for.
Mathew Ginsberg made the following observations:
>The result, of course, was that many relatively incompetent
>individuals started doing AI. As their incompetence has become
>recognized, the field has arrived at a point where it needs to prune
>these people and retrench. This process is in many ways complete in
>the academic arena; individuals have failed to get tenure and an
>increasing number of job openings are appearing in AI as time passes.
Again, this is a phenomenon peculiar to the goUSoA. And it backfires.
The "publish or die" rule in the academic world (US) results in a low
signal/noise ratio, and it may just have taken too long for the
funding bodies to work that one out.
>In the industrial sector, things are more complex because industrial
>jobs are somewhat more stable than academic ones. Often individuals
>cannot be dismissed; entire corporations must disappear instead.
Same comment as above. In UK academic positions are relatively stable,
whereas industrial jobs are unstable.
So ... is there an AI winter over here (you might ask)? There is a research
winter over here. It is affecting all scientific research. AI is no different than any
other endeavour in that respect (I believe).
However ... my research is funded by a large national telecomunications
company (who will remain anonymous). They no longer use the term AI within their
organisation. It is tabu, and is considered as a 2 letter acronym for "pipe dreams".
They now use the 3 letter acronym AIP (Advanced Information Processing).
Patrick
Bill Park
> I am writing an article on AI Winter Refugees: un(der)employment
> among AI professionals.
--Right on, brother!
> I would like to solicit information and
> opinions from the AI community on the following points:
>
> o what were the causes of AI Winter?
Chuck Williams of Inference Corp. places the beginning of the AI
Winter in the first quarter of 1987 (AI Expert Magazine, January 1992,
p. 34). So it has lasted over five years --so far.
The reasons for AI Winter were frequently discussed at AAAI-92 and
IAAI-4 in San Jose, July 12-16, 1992. Some of them most popular were:
* Too many AI vendors chasing too few customers.
* Hype, on a scale rivaling that of the claims for machine translation
back in the 1950s.
* Greed. See "hype," above.
* "Low-aspiration" "tactical" expert systems (i.e., anything that you
could do on a PC in a few weeks) were nothing to crow about because:
-- They often didn't work out (making YOUR boss look like a putz to
HIS boss and forever after making AI a four-letter word in the
whole company).
-- If they did, you were lucky to see a return on investment
of 2:1 or 3:1 from them
-- They usually didn't go any further than the computer-friendly
environment of the AI lab
-- If they did, they could usually only be used at a few places
in the organization (you only had one pretzel-bending
machine in the whole factory that needed a diagnostic expert
system for pretzel bending machines)
-- Once it worked, the guy who developed it got transferred to
the Oshkosh plant and no one could fix it when it screwed
up ever again
-- He didn't get the transfer, but they replaced the model 50
electric pretzel-bender with the new gas-fired model 65,
making the expert system totally useless because it always
wanted you to change the fuses.
* "High-aspiration" "strategic" AI systems would occasionally
manage to get built, despite:
-- The best efforts of hardware and software vendors
everywhere to sell your company a complete line of
incompatible products,
-- The retirement -- or coronary, even -- of your "champion,"
who was the one guy in the whole organization with the
clout to keep pouring money down a rathole for three years
of missed deadlines and incomprehensible dog-and-pony shows
for the Board (mainly because this was going to be his
last, flashy coup before retiring with a bang).
-- Having to rewrite the AI team's mish mosh of Uncommon Lisp,
Prolog in Forth, and object-disoriented C+- into nice,
clean, self-documenting COBOL so that the MIS people would
let you near their precious mainframe with it and you could
finally get at the real gigarecord corporate database
instead of the 150 "representative" records that you'ld
been debugging with for the last three years.
Six months later, having brought the entire MIS department down
several times --once requiring a restore from the full backup
tapes on Christmas Eve-- and with a contract out on your entire
family from the MIS director, you finally get the damn thing
on-line with real customer data pouring through it. This has
actually occurred enough times in the last 5-10 years for an
important conclusion to have been drawn around the
highly-polished tables in the boardrooms of America: When they
work, strategic AI systems give you ROIs in the 10,000:1 range.
Higher, some people claimed at the AAAI conference! Why?
Because they are the kind of system that is used at hundreds or
even thousands of "seats" in an organization, not just on one
pretzel-bender. To field them, someone has to actually walk
around the plant and talk to employees and discover how your
company actually works. They are thus a path to achievement of
Total Quality, the corporate Zen Enlightenment of our age. Just
ask your local CEO if he'd like to be handed a Baldridge prize
by the President of the United States.
* Word got around the MIS departments of the land, and quite a few
decided that they wanted in on this kind of high-visibility
bacon-saving. Pretty soon, high-priced AI consultants were
finding they weren't so indispensable; the boys in the back room
could handle everything now. Client-server architectures were
coming in, too, so you didn't have to make a $50,000 Lisp
Machine pretend to be someone typing on a TSO terminal (and
"reading" the screen, error messages included!!). In fact,
those 680X0 and 80X86 and RISC chips got so fast, you could
actually run 20- or 30-megabyte AI programs quite nicely in
them. In fact, the MIS folks would rather you did, and let them
get on with processing 1,000 tiny transactions per second. And
the rout of other GUIs by the free X Window software didn't
hurt, either. By then, you didn't need to advertise for Lisp
Wizards any more; any old C/Unix hacker fresh out of school
would do. And many of them wore normal clothes.
* In an effort to keep selling $50,000 software packages to large
clients, the AI vendors pretty much all put up new signs saying
"Software Engineering Systems Sold Here," and took the tattered
old "AI, Take It or Leave It" sign out back and burned it. They
learned to say "Verification and Validation," though they still
haven't figured out what those strange words mean. They are
probably French. Of course, many AI shops tobogganed into
Chapter 11 bankruptcy, or got bought by a chain of Indonesian
Yak meat restaurants before they could clean up their acts.
This further decreased the need for people who knew a car from a
cdr, in favor of those who could relate to a data base or work a
net. People in the street began to nod wisely to each other and
say, "A Little AI goes a Long Way."
* Just when we were starting to get the boys in the banks and
brokerage houses in New York to buy a few expert system shells
and get used to searching around a screen the size of a
cafeteria tray with twenty windows all scrolling madly at once,
some damn fool had to spill the beans to them about neural
networks. Once they realized that, of all SIC codes, the money
manipulation industry has the best historical data to train a
net with, it was all over for knowledge-based systems. Who
wants to answer a lot of crazy embarassing questions about the
possibility and plausibility values you estimated for that
$10-million-dollar trade you took the wrong side of last week?
Especially when the kid doesn't know a butterfly from a straddle
and thinks soybeans are things you buy in the organic food coop
in the Village. Put a $10,000 neurocomputer board in a PC, hire
some nerdy little guy to feed it all the numbers in the house,
and you can retire to your own Greek island in a year.
* By the end of the 1980s, realization began to grow that the
Japanese 5th Generation Project was not going to be all that
fatal a death blow to he occidental computing industry, after
all. In 1982, the U.S. was stampeded into setting up the MCC in
Austin, Texas under Bobby Inman with a budget of $65
million/year (of course, it took another two years for President
Reagan to sign the bill allowing IBM to participate in saving
the nation's bacon without being sued for antitrust violation!).
Likewise, the Brits spun their science policy around 180 degrees
and spent 350 million pounds on the Alvey Projects, after they
had nearly strangled all their typically first-rate computer
scientists in response to the Lighthill report at the end of the
1960s. The Europeans, in turn, spent about half as much on
their ESPRIT projects. By the end of the 1980's people were
beginning to wonder when the promised breakthroughs from all
this expensive hustle-bustle would begin to appear on store
shelves. At least we here in the land of the free (or at least,
the marked down) can look forward to 1995, when the Cyc system
will start reading by itself. It will then rapidly complete its
education about common-sense things without further human input
(said Doug Lenat, during a panel session at the conference). The
question is, if Cyc really does develop common sense, will it be
allowed to vote?
> o why has AI gone out of fashion in the US industrial sector?
* Earnings are down, thanks to (1) the recession, and (2) the
untimely departure of those awful commies that the Pentagon found
so very useful for prying our tax dollars out of congress to
scatter around their favorite military contactors for DIVADs,
B-1s, Aguilas, and --oh boy, oh boy, oh boy!-- Star Warts!! Used
to be we'd have either a recession or an outbreak of peace that
had to be put down, but not both at once, so we were always able
to tighten our belts and squeeze through. This time, we've got
'em both at once and certain nonlinear effects are begining to be
felt in our politico-econo-industrial institutions. Like, a
million layoffs is bound to get noticed sooner or later.
* A manager's first duty when earnings are down is to fire someone.
Not himself, that's unsanitary; someone under him. A couple of
months ago, during a panel at the expert systems SIG of the
Software Entrepreneurs' Forum, some folks who used to sell expert
systems to financial organizations said that the only thing the
executives of those companies wanted to know was "How many people
can I fire if I buy your software?"
* In tough times, the best people to fire are people who don't do
anything important and whom no one else likes. Research staff
fit these criteria. Can 'em and nobody even knows they're gone,
right? In fact, a lot of people are glad to see those overpaid
snotnose draft dodgers having to get out and find a real job for
a change. After they're gone, then you can cancel the
ridiculously expensive support contracts for all those wierd
computers they insisted on buying instead of using the company
mainframe like the guys in payroll do. Jeez, aren't PCs good
enough for them? That'll help stanch the haemorrage of cash in
your department and maybe the V.P. will decide to close down old
man Carpfutzer's area instead of yours. Maybe. With mortgage
rates so low, it might be a good idea to refinance the house
now, while you still have a job and can qualify.
> o stories of AI professionals or recent graduates who are or have been
> unemployed or underemployed due to AI Winter;
I'm a 48-year-old Ph.D in control systems from U. of Penn, Dean's list
several times at MIT after skipping freshman math, physics and
chemistry --in other words, not as dumb as I look. 17 years of
experience in robots, expert systems, neural networks, and waving of
vu-graphs in clients' faces. I've had one interview in 14 months of
unemployment, and I'm filling a scrapbook with shaft-O-grams from the
finest companies in Silicon Valley.
I've seen a lab full of talented roboticists in a local aerospace
company having to bootleg "overhead" time to work on their robot
system because there is no money for such work any more.
A top-notch roboticist in another aerospace company here is only
working half time and is just waiting for his hearty handshake.
And of course, you, Steve Vere, are the fellow who wrote the DEVISER
program, a landmark in automatic planning, which is going to be such a
big industry that the expert systems boom will look like the great
Edsel buying frenzy. Are you working?
> o alternative career paths followed by AI Winter refugees.
A friend of mine of similar advanced years who knows object-oriented
data bases inside out has had to take a job as a sales clerk in a
hardware store.
> Countervaling views are welcome, but please be prepared to explain
> the following phenomena: near-zero employment opportunities advertised
> in AI Magazine and similar forums; near-zero recruiting at recent AAAI
> conferences; major declines in attendance at AAAI over the past
> several years.
Another friend of mine has started a business that arranges for good
programmers in India to write your software at a far lower cost than
U.S. programmers would work for. Like the auto manufacturers, we are
faced with unbeatable price competition from Asia. Protectionism is
doubly ludicrous when your are talking about bits entering the country
through a satellite dish on someone's roof.
A triple ludicrousy: The engineers we do have are wandering the
streets in a daze, out of work, yet the NSF is predicting a shortage
of engineers (yes, they're doing it again!!), and both political
parties want to improve the education system so we can turn out more
educated people. I ask you, friends, is the invisible hand of
economics playing pocket pool?
The capital formation system in the U.S. dictates that managers ignore
long-term strategic planning for short-term stabilization of earnings
growth. Otherwise the stockholders sue them. Big Japanese companies
don't have to operate under this constaint as much, since each
Zaibatsu has its own bank, resorting to a lawyer is a disgraceful sign
that you have failed to manage your own affairs well, and Japan has a
critical shortage of lawyers (can't we help them out by sharing a few
boatloads of our bountiful surplus with them?).
> Steven Vere 7057...@compuserve.com
> Boulder Creek, California ve...@netcom.com
Bill Park, Ph. D., President, CEO, CFO, COO, CIO
William T. Park & Associates (me and the cat)
Computer-Aided Quality (me and the Macintosh)
Park Research Institute (me and the cat and the Macintosh)
An equal oportunity employer. We do not discriminate on species.
--
Grandpaw Bill's High Technology Consulting & Live Bait, Inc.
Jacob L. Cybulski
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: ja...@latcs1.lat.oz.au. ... G'Day
=============================================================================
Steven Vere
> slightly larger number of jobs. Whereas in the past (only a few years ago)
> companies had to advertize to find good people to fill their staff positions,
> now they are getting plenty of good resumes and don't feel they have to
> advertize so much. Likewise for recruiting at AAAI.
>
I see. A lot of companies are still hiring AI types, but now they just don't
have to advertise anymore. To document your assertion for the benefit of
graduating students and other uninformed refugees, please post a list of
organizations (with contact phone numbers) actively seeking to hire AI
engineers and researchers. Alternatively, admit you are verbalizing
nonsense.
> [. . .]
> Roberto Desimone rob...@erg.sri.com Tel: (415) 859-4038
> SRI International (EK335), 333 Ravenswood Avenue, Menlo Park, CA 94025
--
___________________________________________________________________________
Steven Vere 7057...@compuserve.com
Boulder Creek, California ve...@netcom.com
___________________________________________________________________________
Richard Wojcik,snake
>o what were the causes of AI Winter?
The recession, corporate downsizing, and the lack of government support
for scientific and technological progress. Expensive, high risk ventures
have been replaced by "safe" investments. Without government incentive
and direction, American industry has tended to drift, whereas foreign
competitors have continued to invest heavily in R&D, driven by generous
subsidies.
>o why has AI gone out of fashion in the US industrial sector?
I think that it is mainly the cost and risk that has kept industry away.
And the problems have not been entirely outside the AI community.
The chief internal problem has been the technology transfer issue. AI researchers
tend to be more academic than applications-oriented. AI solutions often
involve a very big change in the way work is done, and there is a natural
resistance to taking risks with unproven technology. Worse yet, AI
researchers don't always understand the work processes that they set out
to support. So they sometimes build solutions that must go looking for
problems. Within a company, you might work your brains out on a customer's
problem, receive high praise for your efforts, and find your project
finished without any hope of going into production. The reason is that a
flirtation was going on, rather than a serious proposal of marriage. With-
out some concrete successes, you are sure to feel the budget axe at some
point. This is not to say that there haven't been successes, only that they
have been too few to give AI a reputation as producing a good return on
investment.
-----
Disclaimer: Opinions expressed above are not those of my employer.
Rick Wojcik (rwo...@atc.boeing.com) Seattle, WA
Ron Sun
>> 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
>neaties (theorists) rather than by scruffies (software engineers).
>
According to what you said, it seems that AI Ph.D programs should
be renamed Ph.D programs in LISP programmming?
Unfortunately, as one of the earlier respondents points out,
AI applications nowadays do not need much of specialized AI programming
expertise. Any competent software engineer can do it.
Why pay more for a Ph.D?
No physicist has to be a "bloody good" car mechanic.
why do AI Ph.Ds have to be "bloody good" programmers?
Can we view AI more as science, and less as engineering?
--Ron
Steven Vere
The following is an interesting response I received by email from a
person who would like to remain anonymous to avoid jeopardizing
his/her job. In this text, the name of the company has been replaced
by <MC> for Major Corporation.
[begin email text]
>o what were the causes of AI Winter?
I feel it's a combination of factors.
Here at <MC> (and this is only my opinion not <MC>'s)
what I've seen happen is that lots of money was
spent on several prototype systems, of which nearly
all worked. The problem arose when it actually came down
to using the systems. The management here has no concept
of how it runs its processes, because they're held
responsible for money, schedules and people. Therefor they
see very little need to "fix what ain't broke." They
have also been conditioned not to take a risk. With this
kind of corporate culture, AI expert systems just couldn't
be implemented. They wouldn't take the risk, change the work,
or spend the money to implement the prototypes. Therefore
time after time the systems died on the vine. SO that
after a while it looked like a lot of money was being wasted
on a technology that didn't work. When I looked at the data
I found out the technology was working they just weren't
using it. And because of the corporate capacity for self-
delusion, the technology got the blame when it was more
the lack of any ability to implement new technologies in
the information management area.
This brings up another problem. At least here, the corporate
culture really hampers technology transfer to occur in areas
that one can't "touch and feel" the result. ie. It's much
easier to bring in new technologies to the factory floor
that to the MIS department. Our managemenmt just doesn't
understand computers and how to handle information/knowledge.
They can't feel it, or touch it, and computers are much
too archane. In talking with my colleagues, this is a big
problem in any corporation that has an older, white male
management, like <MC>. In this kind of culture, AI systems
just don't flourish, even when used for applications they are
suited for. It's been very frustrating to see this process.
>o alternative career paths followed by AI Winter refugees.
Currently I'm working at a more conventional computing. It's
the pits.
The problem with the corporate culture I think is a much more
important factor than most professionals realize. Once a technology
gets the label of "useless" whether it's deserved or not, it's
very hard to get funding. Even for reasonable projects. In US
corporations the cultures are very similar to <MC>'s - no risk,
short term, quick profit, "don't fix it if it ain't broke",
crisis management that will talk "quality control" but has no
idea how to implement it. This kind of environment will kill
AI type projects every time. It's like putting an African Violet
into a desert and expecting it to grow. Therefor the projects
don't work even though lots of money gets spent. Therefor, to the
US corporate way of thinking, the technology was over hyped and
useless. The "word" gets out to the rest of the US corporate
community, and that is that.
I find it interesting that Europe is not having the same
problem. I've been talking to colleagues who are finding
opportunities in Frace, England, Spain, and Geramny. They
systems are being used there, and jobs are more numerous.
I also find it interesting which companies are still using
this technology here in the states. They're keeping very
quiet about it. They are also branching out their expertise
to include Fuzzy logic, neural nets, and object-oriented
computing -- often starting to combine 1 or more of these
with the expert system paradigms.
[end email text]
Kenneth I. Laws
> Countervaling views are welcome, but please be prepared to explain
> the following phenomena: near-zero employment opportunities advertised
> in AI Magazine and similar forums; near-zero recruiting at recent AAAI
> conferences; major declines in attendance at AAAI over the past
> several years.
Hi, Steve!
I second comments by Roberto Desimone, and stand in awe
of Bill Park's essay. My own take on this is that we don't
have an AI Winter. The original AI Winter, after the Lighthill
report, shut down AI research in most of the UK (and drove
Prolog researchers to the U.S.) When AI sprang back to life,
there was a "gold rush" of schools, corporations, military
services, and individuals trying to enter the field early.
What we have now is merely a shakeout, or perhaps even the
beginnings of a steady state. NSF's AI budget, although small,
continues to increase. There have been no massive cuts at
ONR, AFOSR, or NASA, so far as I know. The famous pioneers,
or publishing scientists, of a decade ago are mostly still
employed (and calling each other AAAI Fellows). AI journals
and books are published at an incredible rate, and there are
too many conferences, workshops, and symposia for us to even
read the calendar notices. Universities everywhere now have
AI professors and turn out AI-interested graduates in great
numbers. (That wasn't true ten years ago.) AI tools are
easily available -- your choice: public domain (by FTP or
CD ROM), shareware, or commerically supported. Applications
conferences are hot, and are well reported in the press. PBS
just showed a series on computers and AI, and AI stories are
common in newspapers and business magazines. This is not an
AI Winter. It feels more like late Spring.
So why aren't there jobs? Well, the recession is certainly
a factor. So is the defense drawdown. (Defense R&D is predicted
to be little effected, but even so the industry is waiting for
new fiscal-year money.) DARPA cut back in neural-network
research (although other agencies took up the slack); perhaps
there have been other government cutbacks that I haven't tracked.
There is certainly more emphasis on near-term results, but I'm
not sure that has reduced total R&D spending.
The real change has been in corporate AI. First, we've exceeded
the three-year time frame that new labs were permitted. Second,
we've already sold to anyone who was interested in buying. And
third, business has learned that it doesn't need what we're
selling. Companies are desperate for competitive advantage and
survival, and only a few financial company see AI as offering
immediate payback comparable to other management options.
What we have here is a simple case of oversupply. We have
produced more graduates and more consultants than the saturated
market can support. Those who are employed are hanging on
tight, so there are no job openings due to churning. Consultants
and tools are plentiful, so new job slots can be filled with
cheap labor -- recent graduates, or even AI wannabees -- and an
occasional training session or consulting contract. A few more
slots will open up as new vertical applications are attempted,
but that happens less often than labs close down or jobs are
eliminated.
One begins to appreciate the genius of the American Medical
Association in limiting the number of U.S. doctors.
I'm afraid we're going to have to live with it. If jobs are
attractive, supply will be plentiful and wages for most will
be low. The laws of economics are the same for AI researchers
as for athletes, actors, artists, professors, and librarians.
-- Ken
--
Dr. Kenneth I. Laws, (415) 493-7390, la...@ai.sri.com.
Moderator of the Computists International AI/IS/CS mutual-aid association.
Ask about my weekly online career newsletter, The Computists' Communique.
Amos Omondi
>
> I am writing an article on AI Winter Refugees: un(der)employment
>opinions from the AI community on the following points:
>
> unemployed or underemployed due to AI Winter;
>
>the following phenomena: near-zero employment opportunities advertised
>in AI Magazine and similar forums; near-zero recruiting at recent AAAI
>conferences; major declines in attendance at AAAI over the past
>several years.
Here is my 2 cents worth: There might be a feeling that, relative
to what AIers promised to deliver and the quantity of $s that they
received, there has been little coming out of AI. Whereas in the
good times employers might be willing to let a few $$$ go down the
tube, in hard economic times they prefer to have a greater
probability of return for their investment. Alternative career
path might be to retrain in something like Networking, Distributed
Systems, UNIX hacking, etc.
Jacob L. Cybulski
> According to what you said, it seems that AI Ph.D programs should
> be renamed Ph.D programs in LISP programmming?
> Unfortunately, as one of the earlier respondents points out,
> AI applications nowadays do not need much of specialized AI programming
> expertise. Any competent software engineer can do it.
> Why pay more for a Ph.D?
> Can we view AI more as science, and less as engineering?
>
> --Ron
Well, you've pin-pointed exactly what I was trying to say... Lots of
companies nowadays ask the questions why should they employ an AI wiz
with a PhD but no programming skills when they could do away with a
Software Engineer some limited AI knowledge (say a few degree level AI
courses and some experience with symbolic programming techniques)?
Certainly if they go for the first solution, they'll have to employ
the second guy anyway, so what's the choice?
I am not saying that a good Software Engineer is going to replace an
AI expert in every single job, but with a slow death of Defense Industry
the number of real AI jobs shrunk dramatically!
I think the problem is that in the early years of AI, most of the AI
PhDs came from the Computer Science background in the first place.
Whereas now there exist many different ways of getting your PhD in
AI through Cognitive Science, Philosophy, Maths, Engineering, etc.
They are simply not equivalent degrees!
Ehud Reiter
AI is the level of academic dishonesty, i.e., situations where someone gets
a large research grant to do something which he or she knows is impossible,
on the theory that `money is money, and once I've got it I can put it to
good use'. Anyone in the physics community who submitted a grant to build a
perpetual motion machine would be laughed at, but similar things in AI seem
to happen all the time, and my fear is that this will lead the outside world
to regard us as a bunch of charlatans, who don't deserve any funding at all.
--
Ehud Reiter
(e.re...@edinburgh.ac.uk)
Tim Menzies
Practical Prolog conference in London. There, literally dozens of
fielded Prolog applications were reported. Were they AI? Since I don't
know what AI means, I can't say. Would those applications have
benefited/ did benefit from being developed by persons with exposure to
the AI literature? Definitely, yes.
Maybe there is no "AI Winter". Perhaps, instead, there is a rude
awakenning and a return to reality for a collection of people who got
off on the wrong track and did not realise it till the banks declared
them bankrupt. That is, I endorse the sentiments of the Jacob L.
Cybulski posting. The so-called AI winter is a result of a
mis-direction by the academic AI community. Please, let us leave our
ivory towers and go out and get out hands dirty with the real problems
faced by real users out there in the real software world. I recommend
such activity. Everytime I have faced a real-live problem, my
understanding of the theory has matured.
Oh, and small note on marketting. If you want to make money, do not
call what you do "AI". Read the AI literature, certainly, but call
yourself "Advanced Technologies" or "Interactive Software Engineering"
or "The User-Modelling Prototyping Group" or some such but not AI. The
term has a bad smell.
--
^ Tim Menzies (a 1/3 doctor) | "More than iron, more than
/"\__/\ AI Lab, Computer Science, Uni. NSW | lead, more than gold I
* \ P.O. Box 1, Kensington, Australia, 2033 | need electricity. I need
\ _ _/ +61-2-663-4576 (fax) +61-2-697-3940 (w) | it more than I need lamb
\| "" ti...@spectrum.cs.unsw.oz.au | or pork or lettuce or
cucumber. I need it for
my dreams." -- RACTOR
Orville R. Weyrich
>
>o what were the causes of AI Winter?
General economic downturn and in some cases general-purpose hardware/
software improving faster than AI software due to mass-market funding.
That's what happened to LISP machines, for example.
>o why has AI gone out of fashion in the US industrial sector?
I don't know that it has -- but it is like one leg of a stool -- pretty
useless without a couple more. For example, I am migrating a database
for a client from an IBM mainframe to a UNIX platform. Part of the migration
involves extracting as much information as possible out of a free-text
field in one of the input files and stuffing it into half a dozen other
fields in the output files. Guess what? It is a job for natural language
processing, and guess what? People learn how to do that in AI classes.
But I would not have the job if I couldn't ALSO do C and UNIX.
I think that there is a lot of AI being done in industry -- but it is just
a part of an application. People don't say "let's develop something
foxy like, for example an AI application." They say "hey look! some AI
techniques are useful in solving this problem that we have."
>o stories of AI professionals or recent graduates who are or have been
> unemployed or underemployed due to AI Winter;
One company I worked for had a R&D project to develop a reverse engineering
tool using Prolog, theorem proving, and expert system technology. Since
their bread and butter was an IBM mainframe 4GL they were hit hard by
marketplace disenchantment with IBM mainframes and they were
unable to sustain the AI R&D project to fruition. They considered the
AI-oriented reverse engineering project to be the future hope of the
company. When the company went under, I was one of the last to go, and the
AI project had not had time or adequate funding to become a product. The
Japanese bought some rights to the technology, so it may still be alive
somewhere in Japan. But the point is, the AI project was not dropped because
it went out of fashion, it was held up as the future generation as far
above the water as possible by its drowning father, but the father succumbed
before the child could be saved.
>o alternative career paths followed by AI Winter refugees.
While looking for my next position I focused on improving my C/UNIX skills
because they are becoming the lingua-franca of industry. I am now doing
AI using those C/UNIX skills.
>
> Countervaling views are welcome, but please be prepared to explain
>the following phenomena: near-zero employment opportunities advertised
>in AI Magazine and similar forums; near-zero recruiting at recent AAAI
>conferences; major declines in attendance at AAAI over the past
>several years.
Declines in attendance -- personally I have not gone to ANY conference
that required me to travel for the past several years -- travel budgets
have suffered major declines. If AAAI has suffered more than others, then
which others?
Near-zero recruiting? AI is still leading edge stuff and companies are
not launching new leading edge (i.e. highly R&D oriented) projects in the
middle of a depression (err... 'economic downturn' :-) They don't need
new people with AI as their only skill -- they are struggling to keep
funding for existing projects.
I checked and don't see much in the way of employment opportunities advertised
in Dr. Dobb's either. Last time I looked in Communications of the ACM
and in IEEE Computer Magazine, there were jobs advertised in all disiplines,
including AI, but the quantity was much less than I remember seeing in the
past. Perhaps you might give some comparative statistics for the
percentage and absolute number of jobs advertised in AI in such a general
forum. Personally, I consider myself to be an AI person, but if I am looking
for a job, I think CACM and IEEE Computer, not AI Magazine. Perhaps advertisers
think likewise.
orville
-------------------------------------- **********************************
Orville R. Weyrich, Jr. Weyrich Computer Consulting
Certified Data Processor POB 5782, Scottsdale, AZ 85261
Certified Systems Professional Voice: (602) 391-0821
Certified Computer Programmer Internet: orville%wey...@tnet.com
-------------------------------------- **********************************
Bill Park
you liked my "essay" on AI Winter (article <3pbn3!b.p...@netcom.com>):
Thanks.
Bill Park
=========
Lee Erman
From: gins...@t.Stanford.EDU (Matthew L. Ginsberg)
Newsgroups: comp.ai
Date: 25 Aug 92 16:29:54 GMT
References: <x+an!9a....@netcom.com>
Organization: Computer Science Department, Stanford University
[...]
In the industrial sector, things are more complex because industrial
jobs are somewhat more stable than academic ones. Often individuals
cannot be dismissed; entire corporations must disappear instead.
Witness the demise of Teknowledge.
[...]
Matt Ginsberg
Without engaging too deeply in this interesting thread, I would like to
correct a factual mistatement in that article:
"Witness the demise of Teknowledge."
Teknowledge merged with American Cimflex Corp a few years back. The
organization that was Teknowledge became the Knowledge Systems Division of the
new Cimflex Teknowledge Corp. There are about of 35 of us here, many after 8
years or more (including me, Rick Hayes-Roth, Allan Terry, Bob London, Terry
Barnes, and Jim Davidson). We are selling AI products (M.4 (nee M.1),
ProductTrack, Enterprise/DX), doing commercial AI work (GM, Motorola, etc),
and working on govt (DARPA and other) AI and architecture contracts (DSSA,
DICE, MMACE, Focus: HOPE, etc). We've certainly not had the success of
Microsoft, but I wouldn't call it demise either.
I guess I can't resist an additional opinion: Although many of the articles
in this thread put forth interesting hypotheses, almost all include no data to
back them up. Here's a bit of data: Of the "original 4" AI companies:
Teknowledge, Intellicorp, Inference, and Carnegie Group, ALL of them are still
around.
--Lee
Lee D. Erman
Cimflex Teknowledge Corp.
Knowledge Systems Division
1810 Embarcadero Road phone: 415/424-0500 ext.422
P.O. Box 10119 fax: 415/493-2645
Palo Alto, CA 94303 email: ler...@teknowledge.com
Matthew L. Ginsberg
takes exceptions to my remark, "Witness the demise of Teknowledge."
Webster gives one definition of demise as "a loss of position or
status." During the AI heyday, Teknowledge stock traded in the low
20's. The last time I checked, Cimflex/Teknowledge was trading at
13/16. Teknowledge in its heyday employed well over 100 individuals
(I think it was closer to 200, but I'm not sure). Now they employ
some 35.
I have a lot of respect for the 35 people left at Teknowledge; they
do good work. But there really shouldn't be much disagreement that
the company is only a shade of its former self.
Matt Ginsberg
Bill Park
> I am writing an article on AI Winter Refugees: un(der)employment
> among AI professionals. I would like to solicit information and
> opinions from the AI community on the following points:
...
> o alternative career paths followed by AI Winter refugees.
Here's a timely cautionary tale for us frostbitten AI types, from
"Combatting defense cutbacks," by Tekla Perry in <The Institute> (news
supplement to <IEEE Spectrum>), September/October 1992, pp. 12, 11.
(The article is about the ongoing horrorshow among their aerospace
members in Southern California) ...
"'In the early '70s, the IEEE came up with a program to retrain
aerospace engineers into ocean engineers,' IEEE Past President
John Guarrera said. 'We retrained 20 people and ended up with 20
unemployed ocean engineers.'"
The article also notes that the IEEE, in addition to loss of members,
loss of conference revenue, etc., is also suffering a loss of
volunteer labor. Their members are either out of work (therefore too
busy scrambling to pay the mortgage to do any extra volunteer work),
or are still employed but working "long overtime hours to fill in for
those who are gone and to prove that they should survive the next
round of layoffs." Is this double-duty also a (supply-side) factor in
AI Winter?
Easy prediction: We'll soon be seeing megabuck lawsuits against such
employers for a U.S. version of the current Japanese "pop" disease of
"karooshi" (death from overwork). Ambulance-chasers, take note!!
Spouses, take notes!! You'll need them for the deposition.
Constructive comments:
One fly in the ointment with retraining is that, after you complete
your training, you still have no experience. Look at the job ads on
Usenet or in the classified section of the newspaper and notice how
many require two or three years experience, commercialized products,
or even say flat out "no entry-level people." I think the best
retraining opportunities would be for fields that are so new that
almost no one has had the opportunity to build up any experience yet,
but have growth potential. Here are some straw people to stimulate
discussion:
* Programming Apple's Personal Digital Assistants in their new
"dynamic language," Dylan. This is going to be one of those
"content-driven" markets, so if you think up new, useful things to
do with Newton(tm?) and his little brothers from Sharp, you'll get to
drive to Safeway for a loaf of bread in your new Testarossa.
* Computerized aids for the molecular biology/genetic engineering
labs. See Larry Hunter's new book on this almost-virgin application
area (reminds me of my old girl friend). Good for a Jensen, anyway.
* Throw in with one of those road-show management consultants /
consulting groups that charge huge fees to scold CEOs of big
companies about what bad boys they have been and tell them they have
to shape up, Restructure, Compete Globally, implement Total Quality
Control, achieve Total Customer Satisfaction, Rationalize their
Factories, Re-Engineer their Organizations, quant. suff. ad naus.
Develop some simple little program for the guru to sell them for an
$11,999 site licence -- plus support and training fees, of course.
The CEOs would then get to take these systems home and stick them on
their unfortunate managers' desks, where, for the next few quarters,
they would guide those managers step-by-step through your chosen
guru's magic formula for success. And of course (nyuk, nyuk!), they
would report back to the CEO's leather-bound touch-screen Executive
Information System ($10,000 in teak commode with integral
temperature-regulated wine storage compartment) on each of those
manager's performance-weighted achievement/compliance metric.
... what? Well, of COURSE it would run on a PC!
...
> Steven Vere 7057...@compuserve.com
> Boulder Creek, California ve...@netcom.com
Bill Park
Roberto Desimone
> Teknowledge in its heyday employed well over 100 individuals
>(I think it was closer to 200, but I'm not sure). Now they employ
>some 35.
Am I wrong? Or didn't Teknowledge Federal become ISX Corporation,
Westlake Village, CA. ISX seems to be doing pretty well.
Roberto
-----------
--
Richard Wojcik,snake
>person who would like to remain anonymous to avoid jeopardizing
>his/her job. In this text, the name of the company has been replaced
>by <MC> for Major Corporation.
>>o what were the causes of AI Winter?
>...
> This brings up another problem. At least here, the corporate
> culture really hampers technology transfer to occur in areas
> that one can't "touch and feel" the result. ie. It's much
> easier to bring in new technologies to the factory floor
> that to the MIS department. Our managemenmt just doesn't
> understand computers and how to handle information/knowledge.
> They can't feel it, or touch it, and computers are much
> too archane. In talking with my colleagues, this is a big
> problem in any corporation that has an older, white male
> management, like <MC>. In this kind of culture, AI systems
> just don't flourish, even when used for applications they are
> suited for. It's been very frustrating to see this process.
I agree with the writer on a lot of things. In fact, the lament isn't all
that different from the remarks I have made. I emphatically disagree with
the implication that the attitude has anything to do with the racial or
gender makeup of the power structure in the corporation. In fact, you find
exactly the same attitudes in women and other minority managers. What you
really have to understand is that AI solutions don't work if you think of
the problems as only technical. When you set out to change a work process,
you are heading straight into a briar patch of politics. You can run away
from it, or you can look for people who can help you sell the vision. Every
corporate environment has its liberals and conservatives when it comes to
change. There are tremendous opportunities in every business environment
that needs to manage information (i.e. *all* of them). If you want to
succeed, you must search out leverage for change. You had better do your
homework in understanding how that change will affect work processes and
listen to those who voice the most opposition. They may even have a point.
>I find it interesting that Europe is not having the same
>problem. I've been talking to colleagues who are finding
>opportunities in Frace, England, Spain, and Geramny. They
>systems are being used there, and jobs are more numerous.
As I said in an earlier post, this may have something to do with the
government's role. Europe and Japan have no problem with establishing
R&D funding agencies to direct and support the private sector. The
US is stuck in the pig-headed position that only military subsidies
are valid. Everything else is "socialism."
Frank Adams
>What you
>really have to understand is that AI solutions don't work if you think of
>the problems as only technical. When you set out to change a work process,
>you are heading straight into a briar patch of politics. You can run away
>from it, or you can look for people who can help you sell the vision.
There is nothing special about AI here. *All* software projects run into
this.
>Every
>corporate environment has its liberals and conservatives when it comes to
>change. There are tremendous opportunities in every business environment
>that needs to manage information (i.e. *all* of them). If you want to
>succeed, you must search out leverage for change. You had better do your
>homework in understanding how that change will affect work processes and
>listen to those who voice the most opposition. They may even have a point.
They almost certainly have a point. There are bound to be problems you
haven't thought of. The best solution is almost always a compromise.
Chris Malcolm
>With regard to the AI Winter debate, one thing that really bothers me about
>AI is the level of academic dishonesty, i.e., situations where someone gets
>a large research grant to do something which he or she knows is impossible,
>on the theory that `money is money, and once I've got it I can put it to
>good use'.
Supposing this is true, there could be a number of reasons for it:
1. AI researchers are dishonest charlatans.
2. The funding agencies have ridiculous goals, and the
researchers are trying to get some decent work done despite
this.
3. The AI funding agencies operate much sloppier controls on how
their money gets spent than do physics etc.
Charlatanry seems to follow big money, and if you're motivated by big
money and research empire building then AI is not at the moment a
fruitful field. For example, I'd expect genetic engineering and AIDS
research to be more attractive to charlatans. I can't think of a reason
why the AI funding agencies should be sloppier. The funding agencies
operate general policies, and their staff move around. So, if you
allegation is true, I would suspect 2. above. Which seems quite a good
reason to be dishonest. Remember von Braun being dishonest with Hitler
about his wonderful new V2 missile? It was a dreadful missile, cost a
fortune to develop and to make, and buried itself so deep in the ground
before exploding that it did very little damage. Being supersonic and
coming straight down it didn't even terrify people the way the V1
("buzzbomb") did. But it made a _lovely_ spaceship!
--
Chris Malcolm c...@uk.ac.ed.aifh +44 (0)31 650 3085
Department of Artificial Intelligence, Edinburgh University
5 Forrest Hill, Edinburgh, EH1 2QL, UK DoD #205
Shad Tomlinson
Z-Mail, Z-Fax - Z-Code Software
**********************************
Z-Code Software Corporation in San Rafael, CA is seeking strong,
experienced -senior- software engineers in the following categories:
Z-Fax: UNIX, X must. Knowledge of fax modems a plus.
Z-Mail: Microsoft Windows. Just a great GUI programmer.
The best candidate would be the type of person that has done a lot
of personal programming -- a wiz at Windows or X programming. Someone
that has such a desire to program and build apps, that you spend a
lot of personal time building cool utilities and things. A dedicated
hacker.
Please don't phone initial correspondence unless you were told
about this and literally have no email access. (415) 499-8649
I'd rather see code samples than a resume (at first). Send an ftp
site if code is too large for email. If you must send a resume,
please send text only -- no formatting code like troff. Also, when
you send mail, let us know what you're doing now, what you've done
recently with X and/or Windows, and what other sorts of things you
think might be pertinent.
About Z-Mail:
Z-Mail is the most portable electronic mail package on the market
today ported to virtually all UNIX platforms and DOS. (We need
someone to take this port over to Windows.) UNIX World magazine
awarded Z-Mail Product of the Year in 1991. Z-Mail supports Motif,
Open Look and character based (tty) interfaces.
About Z-Code Software:
Z-Code is located in Marin County, about a 15 minutes north of the
Golden Gate Bridge. There are now 12 people in the company, soon to
be at least 15 by the end of the year. Stock options available, etc..
Very casual environment.
Mail all correspondence to pers...@z-code.com
matteo vaccari
>In article <1992Aug28....@aisb.ed.ac.uk> e.re...@ed.ac.uk (Ehud Reiter) writes:
>>With regard to the AI Winter debate, one thing that really bothers me about
>>AI is the level of academic dishonesty, i.e., situations where someone gets
>>a large research grant to do something which he or she knows is impossible,
>>on the theory that `money is money, and once I've got it I can put it to
>>good use'.
>Supposing this is true, there could be a number of reasons for it:
> 1. AI researchers are dishonest charlatans.
> 2. The funding agencies have ridiculous goals, and the
> researchers are trying to get some decent work done despite
> this.
> 3. The AI funding agencies operate much sloppier controls on how
> their money gets spent than do physics etc.
>Charlatanry seems to follow big money, and if you're motivated by big
>money and research empire building then AI is not at the moment a
>fruitful field. For example, I'd expect genetic engineering and AIDS
>research to be more attractive to charlatans. I can't think of a reason
>why the AI funding agencies should be sloppier.
Well, I am convinced that it is far easier to be an AI charlatan than
a biology charlatan. This is because there is very little agreement
on what the nature of AI is, what are its goals, its means, its most
significant achievements. Look at the debates in comp.ai.philosophy!
While biology is a much older and established science, where you can't
fool a biologist into believing you are competent when you are not.
--
Matteo Vaccari, vac...@ghost.dsi.unimi.it
Richard Wojcik,snake
>In article <81...@bcsaic.boeing.com> rwo...@atc.boeing.com writes:
>>What you
>>really have to understand is that AI solutions don't work if you think of
>>the problems as only technical. When you set out to change a work process,
>>you are heading straight into a briar patch of politics. You can run away
>>from it, or you can look for people who can help you sell the vision.
>There is nothing special about AI here. *All* software projects run into
>this.
That is true, but not to the same degree. If your project is considered more
or less conventional technology, then its affects on the work process are
relatively better known. Managers can look at their resources, the competence
of their personnel, and the likely cost. Implementation is straightforward.
And you may be doing it just to stay competitive. If it fails, you will pro-
bably be faulted for other things than having taken the risk in the first place.
With AI, you don't know whether it will scale up, what it will take to maintain
it, how the users will take to it, etc. Generally, there are no standards that
apply to it. And does your competitor have such a product? What ever made you
think that such a hare-brained idea would work? Bigger risks. Sweat beading up
on your forehead. A headache in the making. And your little budget won't allow
it. You've got other things to worry about--more conventional software projects.
Safer ones. Better to be prudent. ;-)
Kenneth I. Laws
> I can't think of a reason
> why the AI funding agencies should be sloppier. The funding agencies
> operate general policies, and their staff move around.
This doesn't seem to be the way that it works in the U.S.
NSF program directors do not move around, except for a
few bureaucrats who keep paperwork moving in vacant positions.
Directors move in and out, not around. :-)
The amount of "slop" in the system depends on the disparity
between funds available and funds required to support the
best research. Funding agencies that get their money from
Congress typically request as much budget as they think
they can get, regardless of the need in each category.
Congress then tweaks the amounts based on perceived payoff,
media interest, photo-op potential, competing programs,
politics, etc. -- again without much relationship to the
number of really good proposals available.
In biology, it's not uncommon for a proposal to be rejected
because it fails to state what hypothesis is to be investigated,
what methods will be used, or how the results will support or
disprove the hypothesis. In AI it's unusual to see any of those
addressed.
Marie Bienkowski
that there's been some discussion about the AI winter manifesting itself
as a slowing down of hiring people for AI jobs. My impression is that, while
there are certainly as many people graduating with AI degress, people who
already have AI jobs are staying put because of the recession. Consequently,
employers may not be advertising as much, since new graduates are hired via
campus recruiters. This is what I've observed from resumes we've received.
We get a lot from fresh-outs, almost none from people with 3-5 years
experience.
We also get a lot from people who I'd call thinkers, and not many from people
who I'd call doers. And given the requirement to show results and show them
soon, doers are what *we* need.
I'd be interested in hearing from people who are job hunting in AI to
see what kinds of qualifications are out there.
Thanks,
Marie B
bi...@erg.sri.com
Steve Chenoweth
>...
>We also get a lot [of job applications] from people who I'd call thinkers,
>show results and show them soon, doers are what *we* need.
You could start a big time debate in comp.ai about the value of "thinkers"
versus "doers." Generally, a major tenet of AI is that there's value in
having general vs. specific problem-solving skills and knowledge. Another
is that we need to look at creating both immediate and also long-term
solutions to problems in data processing.
Thus, AI people may be seen by recruiters as "thinkers," particularly if those
recruiters are focused in on a much more specific level, like, "How fast
can they code in C?" or even "How many expert systems have they built?"
My experience has been that AI people usually are both doers
and thinkers, but become quickly disenchanted when they discover
that the folks they work for, or might work for, have only short-term
goals.
You'd expect to see more of these people disenchanted, especially during
a recession time when everyone's desperate and long-term ideas are out of
fashion.
I'll admit Bienkowski has a point when it comes to people who have focused in
on theoretical issues to the exclusion of pragmatics. You want to see things
in a resume that show they can deliver quality products to a real user,
when the position will include this kind of work. (Probably we could
initiate another great debate by asking whether or not the latter experience
as well as the former ought to be required for one to do useful research...)
There is a growing feeling in business that individual workers need to pay
real attention to the long-term and general effects of their work, to trends
in their business, and to opportunities they might create. It's all part of
everyone's becoming "empowered," with associated reductions in the number
of managers needed to run the business. This feeling about individual
contributors suggests that foresight and general problem-solving skills will be
increasingly valued in business as time goes on. Perhaps the "AI Winter"
will be seasonal after all.
Craig Lindley
--
--------------------------------------------------------------------
Craig Lindley ( lin...@syd.dit.csiro.au )
CSIRO Division of Information Technology, Sydney, Australia.
PO Box 1599, Macquarie Centre, North Ryde, NSW, Australia, 2113