Look to your left for a chunky, sand-colored office block with big square window grids wrapping the corner-more “serious paperwork” than “storybook castle.”
This unassuming building once hosted one of Glasgow’s most ambitious brains-and-bolts experiments: the Turing Institute, an artificial intelligence lab that ran from 1983 to 1994. And yes, it’s named for Alan Turing-the wartime codebreaker and computing pioneer-because one of its founders, Donald Michie, had actually worked alongside him at Bletchley Park. That’s a pretty direct line from cracking ciphers to teaching machines to learn.
The Institute was set up in 1983 by Michie, Peter Mowforth, and Tim Niblett, growing out of the Machine Intelligence Research Unit at the University of Edinburgh. The big idea was bold and very Scottish in its practicality: do serious AI research, but also get it out the door and into real companies. With backing from the Scottish Development Agency and a push from Sir Graham Hills, it moved to Glasgow and built a strong partnership with the University of Strathclyde. The board and staff mixed high-level names with hands-on builders-like Stephen Muggleton, who became a key figure in an approach called inductive logic programming, basically getting computers to form rules from examples instead of being spoon-fed instructions.
In 1984, the Institute became an official training center under the United Kingdom’s Alvey programme, which tried to give British AI a fighting chance. Under Judith Richards, major companies seconded people here-IBM, British Airways, Shell, Unilever-sending their own staff to learn how to turn AI into something that actually worked on a Monday morning.
And then there was the library-quietly radical. Starting in 1983, it built an electronic database pulling material from AI labs worldwide. Subscribers would dial in (yes, with the old squeal-and-hiss modem sound) and get weekly updates of new items, which they could order or download as abstracts. It was like a brainy newsletter service for the pre-web world-proof that “information superhighway” began as more of a narrow, noisy lane.
The research itself was a mix of daring and peculiar. They helped Westinghouse improve nuclear plant efficiency using machine-learned rules-then used that momentum to launch “Freddy 3,” an advanced robotics project exploring robot learning and even robot social interaction. Elsewhere, they worked on everything from credit card scoring to seed sorting, and even wrote code-using a rule-learning system called Rulemaster-for a Space Shuttle auto-lander project based on NASA simulator training examples. If that sounds like a lot to ask of 1980s computers, well… it was. But they did it anyway.
They also partnered with Sun Microsystems on new interface tools, and one researcher, Arthur van Hoff, later helped create core early Java tooling. The Institute even ran the first Robot Olympics in 1990-Glasgow hosting robots awkwardly competing, which honestly feels like the most believable sports event ever invented.
Then came the hard part: funding dried up, the broader United Kingdom mood cooled on AI, and by 1994 the Institute closed-amid political questions and plenty of bitterness. A place built to prove the future arrived a bit too early, then got told to pay rent.
Ready for Glasgow City Chambers? Just walk east for about 2 minutes.
