This was a 180 seat room with 250 people in it, filling the aisles and spilling out into the hallway. Who would have ever thought that Artificial Intelligence could become so trendy! The attraction was an IBM speaker who talked about the so-called "expert systems" tools available from IBM, and their product direction.

IBM has 140 expert system projects ongoing at last count, and takes the technology very seriously. There is not much new to report in the basic product line; they still support three shells, and IBM is working to make them operate uniformly across the product line.

ESE (Expert System Environment) is intended for end-users, and is a quick way to get basic rule-based systems operational. It is not known for its high-performance aspects. Knowledge-Tool is a set of PL/I extensions intended for data processing types; it is a good performer, but requires programming skills to make work properly. KEE (Knowledge Engineering Environment) was originally a product of Intellicorp, but IBM has adapted it to the RT PC and MVS worlds. KEE is intended for persons with formal AI experience; a comprehensive knowledge of Lisp is recommended. It provides all the bells and whistles you can name - forward and backward chaining, frames, external functions, etc. - but is a performance hog and requires heavy hand holding.