Once the sole province of arcane computer science researchers, artificial intelligence is quickly emerging as a mainstream product for some of the biggest computer companies in the world, including International Business Machines Corp., Digital Equipment Corp. and Hewlett-Packard Co.

Interest in computers that emulate human intelligence is surging, despite an industrywide sales slump in computers and software. Attendance at the International Joint Conference on Artificial Intelligence this week in Los Angeles has more than doubled from last year's attendance of 3,000, and the number of exhibitors also doubled to nearly 60.

"People are still very bullish on AI," said Woody Bledsoe, president of the American Association on Artificial Intelligence and director of artificial intelligence programming efforts at the Microelectronics and Computer Technology research consortium in Austin, Tex. "We'll probably have a downturn at some point but, right now, we seem to be headed in the right direction."

For the first time, major companies such as IBM and DEC are offering products using artificial intelligence for sale rather than simply presenting research papers for academic consumption.

The big computer companies are now selling artificial intelligence tools that enable computer programmers to design their own artificial intelligence systems. Most notable are computer programmer workstations that run the LISP computer language. LISP -- which stands for LISt Processing -- is the most common language used for artificial intelligence programming. Initially, young companies like Symbolics Inc. pioneered the LISP workstation market, but now, the larger companies are moving in.

DEC introduced a $50,000 AI workstation for its Microvax II computer that runs LISP. Hewlett-Packard introduced LISP workstations in the $20,000 to $50,000 price range, and Xerox Corp. revealed a new $10,000 artificial intelligence applications workstation.

Business computer specialists who attended the conference, however, were particularly interested in the slew of new "expert systems" software development programs now being offered by both venture firms and more established companies. Expert systems are computer programs designed to replicate the decision-making process of experts using special decision rules. For example, a medical diagnosis program might offer thousands of computer-based "if . . . then" rules that could aid a physician in making a diagnosis.

IBM, which last year demonstrated an internally developed expert system, introduced an expert system in Los Angeles designed to enable people to customize their own expert systems on IBM computers. IBM's Expert System Environment/VM will cost roughly $35,000.

Texas Instruments announced that it would buy a 10 percent stake in the Pittsburgh-based Carnegie Group, a major expert systems developer, as well as acquire a Carnegie license for AI software products and training. There appears to be a major emphasis in selling the tools to make expert systems for such applications as engineering design and financial analysis, rather than design the expert systems themselves.

One important aspect of the mainstream computer industry's interest in AI is that it may be easier for established companies to sell AI hardware and software to customers than the less established venture companies. Data processing managers in major companies may feel more comfortable in exploring artificial intelligence products if they are backed by well-known suppliers.