[1987]

The fact that the opening sounds appropriate to what's going on today is just wild.

> Third, one of the major benefits from AI research has little to do with AI directly; the programming environments and techniques developed for AI research can also be applied to conventional programs. In particular, the ability to do rapid prototyping or what some refer to as exploratory development. This is being seen as a way to limit risk in large programming projects. Rapid prototyping is also valuable in helping customers to more fully define and articulate their needs at the beginning of projects.

> Finally we're beginning to see what I think is a most telling sign for knowledge based systems they're ceasing to be just standalone applications they're more commonly becoming integrated with the installed base of conventional computing especially with databases and realtime applications. As knowledge-based systems become components embedded within larger systems some of the glamour may disappear but their usefulness will greatly increase.

Aside from a few popular terms that didn't transfer ("knowledge-based" and "rapid prototyping") this sounds like something that could have been said today.