Textbook of AI

Also known as: FGCS, 5th Generation

The Fifth Generation Computer Systems Project (FGCS) was a Japanese national initiative running from 1982 to 1992, with a USD 400 million government budget and substantial industry participation. The project, led by ICOT (Institute for New Generation Computer Technology), aimed to develop massively parallel computers running concurrent logic programming (a Prolog derivative called KL1) as the substrate for an "intelligent" generation of computers, surpassing American symbolic AI by combining hardware, software and AI into an integrated whole.

The announcement triggered defensive responses internationally: the US established the Strategic Computing Initiative at DARPA, the UK launched the Alvey Programme, and Europe started ESPRIT. The US-Japan AI rivalry of the 1980s drove much of the era's funding and rhetoric.

The project produced significant technical results, parallel inference machines (PIM systems), the KL1 language, parallel Prolog implementations , but failed in its broader ambition. Concurrent logic programming did not scale to the complex applications hoped for, and the project's commitment to symbolic-logic-based AI missed the connectionist revival happening in parallel. By the time the project formally ended in 1992, the AI hardware market had collapsed and most of the project's specific technologies were already obsolete.

The FGCS is sometimes blamed for catalysing the second AI winter by being a high-profile failure that damaged the credibility of AI generally , though this read is contested. Its lasting contribution may be the model of large-scale, government-coordinated AI research that has resurfaced in modern Chinese, European and US AI strategies.