Knowledge is Power

[mmcgeach at csail.mit.edu (Michael McGeachie)]

As most of you know or have by now surmised, my research is on decision
support systems (not, in fact, hierarchical beer diffusion in spatio-social
networks, so quit asking me for a dissertation on that).  Decision support
techniques have been around for a while, but have only more recently gained
significant interest from the AI community.  And so earlier this week I was
perusing the shelves of the Dewey Social Sciences library looking at books
with the phrase "decision support" in the titles.  I began to realize I'd
stumbled into an important parallel branch of research on this area.  These
books, written in the 80s for students of management, took a different
approach to decision support.  Some authors described decision support
systems that could catalog all manners of data and statistics, and then
display this information later.  The decision support system could thus keep
track of transactions, profit projections, phone numbers, shipping
schedules, and even meeting times.  Further, when it came time to make a
decision, the system supported the decision-maker by recalling and
displaying this information, frequently in sorted lists!  

Excited by my newfound knowledge of data-storage and recollection techniques
applied to decision making, I brought this discovery to the attention of my
advisor.  Unimpressed, he told me the following story: 

"Back then we had a catch phrase in AI, something like you'd use to sell
laundry detergent to people, and it was 'Knowledge is Power.'  Now, not
everyone interpreted this as an advertising slogan.  Some interpreted it
quite differently. In fact, at European conferences, Germans would accost me
about how American researchers were pursuing American imperialism by
cataloguing knowledge.  If Americans could gather all the knowledge in the
world, that meant Americans had all the power in the world, and so AI was
merely another tool of American global hegemony."

Now that I knew "decision support" used to mean "spreadsheet" and "AI" used
to mean "tool of American hegemony," I was interested to see what other
computing terms used to mean.

Trevor Darrell told me that "computer vision" used apply to all techniques
for spotting computers in the dark.  "That's why we used to make them really
shiny.  We were quite proud of ourselves, making computers shiny and
reflective.  Well, until someone thought of putting an LED on them; that
really revolutionized the field.  People could pick out those little
blinking lights from clear across the room.  And a room with a computer in
it had to be pretty big back in those days."

"Software Engineering" used to mean the science of building mechanical
contraptions to load and carry huge arrays of punch-cards, according to
Michael Ernst.  "One of the landmark achievements was basically a big
rolodex.  Man, did that make flipping through punch-cards easier."

"You don't even want to know what used to pass as 'graphics' research," said
Seth Teller.  "Well, ok, fine.  Let's just say it was nice not worrying
about all that linear algebra when you were designing Pong.  You can pretty
much program a Pong-interface without having to do a matrix transform of
jack %$#@."


And, so, that is the story of the time I realized that spreadsheets used to
be called "decision support systems."  Come share your radical insights at
this week's ... 



            +-                                                  -+
              girl scout benefit -+-  5:30 pm  -+- 32-G9 lounge
            +-                                                  -+

               For those coming from elsewhere: Building 32 is
                <http://whereis.mit.edu/map-jpg?selection=32>
           Once you are in 32, just take the G-elevator to the 9th
       floor and we will be in the lounge that you will be looking at