Mccorduck’s-machines who-think after-twenty-five years-revisiting the-origins of ai
pecially ones not so intent on por-traying the genesis of AI as occurring
McCorduck’s
in splendid isolation from other dis-ciplinary innovations growing out ofWWII and, in particular, one more
Machines Who
closely bound with certain specificapplied concerns found in that era. McCorduck conceded in her bookthat her account was unabashedly
Think after
personal and impressionistic, but shedid not reveal the extent to which ithad been colored by her own close
Twenty-Five Years personal relations with some of the
early members of the Carnegie Mel-lon University Computer Science De-
Revisiting the
partment. In its execution, Machinestended to be dominated by the view-point of Herbert Simon in ways both
Origins of AI
big and small. In everything from itselevation of the symbol-processingapproach to center stage (alreadysomewhat outdated by the later
1970s), to its expressions of disdainfor philosophers, to its treatment ofJohn von Neumann’s later positionon computer intelligence as some-how perversely misguided, to the
■ Machines Who Think: A Personal Inquiryinto the History and Prospect of ArtificialIntelligence, Pamela McCorduck, San
AI, the text is redolent of Simon’s ex-
realized” (p. xii) and that this innate
century, a number of bookshave sought to explain AI to
duck’s Machines Who Think,1 now ap-
cal problems…. I like to think of arti-
Copyright 2003, American Association for Artificial Intelligence. All rights reserved. 0738-4602-2002 / $2.00
ly rejected McCorduck’s frame tale: “I
tractions of the term artificial intelli-gence had more than a little to do
notoriously elusive virtue of “intelli-
his celebrated lectures The Sciences ofthe Artificial. This side of Simon be-
Corduck’s version of the history. It is
are that no resistance has beenrecorded because no one has had theresources to attempt a large-scale
The roots of AI in operations research not only shedlight on the capacity of the early discipline to pro-ductively straddle the science- engineering divide
side of the history of AI that hadbeen the most successful in maintain-
but also go some distance in explaining other keyaspects of the history that McCorduck elides.
well as absorbing the vast bulk ofprogramming effort. This segment ofthe science adopted as its manifesto
sis.” It was also the version of intelli-
The term artificial intelligence encap-
fact that AI had to find its initial uni-
the differential status of operations re-
Bartree, T., ed. 1989. Expert Systems and AI.
Bowker, G. 1993. How to Be Universal. So-cial Studies of Science 23(2): 107–127.
Fortun, M., and Schweber, S. 1993. Scien-
tists and the Legacy of WWII. Social Stud-
tributed to Simon and Newell earlier. ies of Science 23(6): 595–642.
Heims, S. 1991. The Cybernetics Group.
Johnson, S. 1997. Three Approaches to Big
biosis. IRE Transactions on Human Factors
Mirowski, P. 2002. Machine Dreams. New
Nilsson, N. 1998. Artificial Intelligence: ANew Synthesis. San Francisco: Morgan
Norberg, A., and O’Neill, J. 1996. Trans-forming Computer Technology. Baltimore:
1. The other serious history that aimed to
Mangle. Social Studies of Science 32(5):
Crevier, D. 1993. AI. New York: Basic.
Rau, E. 1999. Combat Scientists, Ph.D.
thesis, Department of History, University
Guice, Paul Edwards, and a host of others.
2. Internalist is a term of historiography. It
and OR. Interfaces 17(4): 8–15.
means that the trajectory of an intellectual
Simon, H., and Newell, A. 1958. Heuristic
discussion is driven exclusively by contribu-
within a discipline or discourse community.
and Intelligence. Mind 59:433–460.
duck with J. C. Shaw, June 16, 1975, p. 31. Philip Mir- owski is the
thesis of assertions of failure of cybernet-
7. See Newell’s admission of this fact in
Reilly Center for the History and Philoso-
phy of Science at the University of Notre
Dame. He is the author of Machine Dreams
(Cambridge, 2002), Science Bought and Sold
(Chicago, 2002), and Effortless Economy ofScience? (Duke 2003). His recent work
9. See Denicoff in Bartree (1989) and Nor-
ranges from a history of early AI to docu-
commercialization of science policy. His e-
Abstract Much of what medical researchers conclude in their studies is misleading, exaggerated, orflat-out wrong. So why are doctors – to a striking extent – still drawing upon misinformationin their everyday practice? Dr. John Ioannidis has spent his career challenging his peers byexposing their bad science. In 2001, rumors were circulating in Greek hospitals that surgery residents, eage
Case Scenario 1 A 53 year old white female presented to her primary care physician with post-menopausal vaginal bleeding. The patient is not a smoker and does not use alcohol. She has no family history of malignancy. She had an endometrial biopsy that was positive for endometrial adenocarcinoma. She was sent to have a CT of the abdomen and pelvis and was found to have thickening of the uterus
the differential status of operations re-
Bartree, T., ed. 1989. Expert Systems and AI.
Bowker, G. 1993. How to Be Universal. So-
cial Studies of Science 23(2): 107–127.