Lenat

Douglas B. Lenat is an American computer scientist and AI researcher known for his work in knowledge representation and automated reasoning. He is the creator of the Cyc project, an ambitious effort to assemble a comprehensive large-scale knowledge base of common-sense information and to build software that can reason with that knowledge.

Cyc aims to encode general world knowledge and rules of inference so that AI systems can perform everyday reasoning without hand-crafted, domain-specific programming. Starting in the mid-1980s at the MCC and later continued at Cycorp, Cyc has been one of the longest-running AI knowledge bases. The project has influenced debates about the feasibility of common-sense reasoning and the scalability challenges of maintaining vast ontologies.

Lenat also developed EURISKO, a heuristic discovery program that learned heuristics from experience rather than hard coding rules. EURISKO was notable for its ability to tackle a variety of domains by adapting its own heuristics, which showcased the potential and limits of meta-heuristic learning in AI.

In 1994 Lenat founded Cycorp, a company dedicated to commercializing Cyc and applying its knowledge bases to enterprise tasks, natural language understanding, and decision support. Lenat has been recognized as a leading figure in knowledge-based AI and has advocated for large-scale, symbolic approaches to AI reasoning, in contrast to purely statistical methods.