History

You won't understand AI unless you know something about its history - including the part played by rule-based systems. Why, for example, did symbolic models become popular in the late 50s?

For a brief run-through, see Chapter 8 of The Guide to Expert Systems, by Alex Goodall (Learned Information, 1985; RSL Comp BD 36; PSY KH:G 061). This was written in 1985, near the height of the expert systems hype. Indeed, the book was a result of this. A new expert systems journal had just started up, intended for a commercial audience. The book was to be given away (or sold cheaply) to new subscribers with no knowledge of AI, as a way of easing them into the subject.

For a more recent perspective, see AI: the tumultuous history of the search for artificial intelligence by Daniel Crevier (Basic Books 1993; PSY KH:C 068). I've just got this into the library, and will be recommending it as an introduction to AI in future years. Chapters 6 and 8 deal particularly with expert systems and their part in the commercialisation of AI. In a few years, it will be time to do the same for neural nets.

Some points in early AI history:

• Early focus on cybernetic models and simple neural nets. Belief in the ability of self-organising but unorganised systems to learn. Goodall pp 163-164; Crevier pp 26-37, 102-107.

• Naming the subject! Start of symbolic AI. Closely related to the development of digital computers. Goodall p 165 (see first quote); Crevier 37-50. See also GPS, Goodall p 166.

• Start of production-system models of mind. Crevier p 150. See where Newell and Simon got the idea from?

• Focus on domain-specific knowledge. Goodall pp 167-170. Crevier pp 146-156. See Feigenbaum quote on p 167 of Goodall and p 156 of Crevier. However, Crevier pp 146-147 also cites the limitations of early computers as a reason for lack of work on knowledge.

• Expert systems hype: See quote on Goodall p 173.

• Logic programming and Prolog. Crevier pp 190-196.

• Japanese Fifth Generation Project, and Western reactions. Goodall pp 175-176. Crevier chapter 8.

You can fill in the rest yourself. If you're not interested, you shouldn't be doing the course anyway! Be aware that different people have different interpretations of history, and that I've only mentioned a very small part of even the early work above.