Celestial Weasel (celestialweasel) wrote,
Celestial Weasel

The Weasel School of Failology

Another book I have finished is The Fifth Generation: Artificial Intelligence and Japan's Computer Challenge to the Worldby Edward A. Feigenbaum and Pamela McCorduck. This was a call to arms for the American computer industry written in 1983 back when we all believed that the Fifth Generation Project coordinated by the Japanese Ministry for International Trade and Industry was going to lead to Japanese dominance of the computer industry comparable to that of the car industry. Obviously this is a slightly bizarre historical artifact, since the fifth generation project was essentially a gigantic failure and as a result we aren't all using KIPS (knowledge information processing systems) and writing in KL1 (a parallel Prolog variant).
The book is actually a bit light on detail as to how KIPS and the Expert Systems implemented to run on them were going to change the world.

It could perhaps be argued that reading so many books about failed technologies and/or business failures and/or corporate malfeasance and/or the perfidy of NGOs such as the World Bank and IMF etc. may not help my state of mind, but failure is more interesting than success. I even have 3 books on Prestel, including one written by James Martin as in the James Martin 21st Century School at Oxford University. His book is rather turgid, certainly not nearly as good as the book co-written by Sam Fedida, its inventor.

However, there are some failures that are not adequately documented, so I bring you the list of research topics for the Weasel School of Failology.

1. Why did the Fifth Generation Project fail? The interweb is surprisingly silent on this. There is always the nagging doubt at the back of ones mind that Prolog might actually be very clever and that it might be possible to get it to work. Probably an illusion...
2. Back in the 70s people used to talk about a number of allegedly successful expert systems. Where are the expert systems of today? Was this all just hype, or is this one of the cases where the technique has become so mainstream that it's not thought of as 'AI' any more?
3. Why has so little of value come out of the Media Lab?
4. Why did nothing much apart from a bankrupt games company come out of Interval Research?
5. Taligent. Clearly the 'why did a joint venture between IBM and Apple to create a new OS fail?' question need not be asked. Hint - the answer is in the question. However, at one point it was rumoured that a good bitch and tell book was in the works, but nothing ever became of this. So, where are the sickening yet amusing stories?
6. ITV Digital. Why did the companies that failed so badly with BSB manage to fail in almost exactly the same way a second time?
7. Back in the days of the 5th generation project, the British government had the wind put up it sufficiently to fund computer science research. In the way of such things, the money was then, as far as I can see, trousered by people to do what they were going to do anyway. Apart from Expert Systems, covered above, the big ticket item was 'Formal Methods'. Whilst I don't doubt their use in some narrow areas, the idea seemed to be that end-users really would look at Z or VDM definitions of software. Surely this must have been a con-trick or wind-up, surely no-one who had been in the same room as an end-user, or been in a room with someone who had been in a room with one could have believed this?

Answers to The Celestial Weasel 21st Century Institute, in a 70s concrete office block, near the multi-storey carpark.
  • Post a new comment


    default userpic

    Your reply will be screened

    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.