Early AI History according to RobotWisdom
These are Jorn’s memoirs from the Institute of Learning Sciences at Northwestern University, from which he was fired in 1992. For “AI” in RobotWisdom read Knowledge Modelling, with a human / behavioural slant. Reminded to re-read by the Arthur Andersen / Roger Schank Virtual Learning reference in the New York Metro on the Andersen / Enron fallout. (The story, by JJCramer of TheStreet.com, also for the Auditor / Consultant scam content.)
Interesting too, ironic actually, given my empathy for Jorn’s line of thinking, and his falling-out with Schank / ILS, that Schank’s company should be called “Socratic Arts“. It’s yet another classical vs non-classical reasoning battle. No problem with Schank’s Virtual Learning approach – Heuristic, with low-risk “space to fail” simulator environment – basic common sense as used for decades in high-health-and-safety-risk industries. His success stories in selling simulator packages is like any application software success – Provided the application domain is reasonably clearly defined, a classical-rational information model will suffice. This appears to be more a matter of (engagingly / efficiently / effectively) “teaching” something that is already known, rather than learning new knowledge. Excellent training strategy, but pretty limited for “learning organisation” aspirations. (Interesting independant summary / commentary on Virtual Learing from JJJKasvi, Helsinki University of Technology – interesting source of other KM material.)
Same issue with Chris Gray’s stuff (JIMS Cambridge) on organisational learning ?
Same confusion between “education” and “knowledge” ?