Many years ago I wrote some articles and a long book chapter to explain -- and argue against -- proposals to develop artificial intelligence for military weapons. They are collected here out of personal and historical interest.
The "Star Wars" Defense Won't Compute, The Atlantic, Vol. 255 No. 6, June 1985, pages 18 -- 30. Illustrations by Jose Cruz.
I believe this was the first article in a popular publication to explain that computing and software would be a serious problem for the Strategic Defense Initiative (SDI, or "Star Wars"). In addition to reporting on SDI, it also explains artificial intelligence and the problems of software reliability to non-specialists.
(SDI was a proposal for space-based missile defense that received enormous attention in the 1980s. It was never built.)
Debate on Computing Requirements of Star Wars, The CPSR Newsletter, Vol. 3 No. 4, Fall 1985, pages 1, 10 -- 11.
I attended this debate on SDI computing at MIT, that featured four prominent computer scientists. It was reported in the national press, including The New York Times. My report includes more of what the participants said, and how the audience responded.
The Strategic Computing Program, in Computers in Battle: Will They Work edited by David Bellin and Gary Chapman, Harcourt Brace Jovanovich, 1987, pages 171 -- 208, 333 - 337.
The Strategic Computing Program was another military project active at the same time as SDI, devoted to applying artificial intelligence and other speculative technologies to weapons. This book chapter provides a short history of its sponsor the Defense Advanced Research Projects Agency (DARPA) and explains its central role in launching American computer science research. The chapter describes the political background and content of the program, and discusses the tensions that arose when the formerly benevolent sponsor of generic research became militant and politicized. It reports on criticism of the program, the blustering responses of its advocates in the Defense Department, and the rationalizations, evasions, and misdirections of its apologists among the computer scientists.
This chapter (and the Atlantic article) also report on the wide-eyed claims and predictions for artificial intelligence that were being made in the 1980s.
Revised May 2023