If you are old enough you might remember an early MS-DOS video game called Theme Park, from a British company called Bullfrog Games, founded by the legendary game-master Peter Molyneux, whose other popular games included Dungeon Keeper and Populous. Theme Park was like SimCity but for theme parks, and it required you to manage all the various elements of building a park and then making it profitable. In order to become a theme park mogul you had to expand, but you couldn’t buy a new park until you managed to sell your current one, and if your rides were extreme enough, you could cause your little people to literally fly off the rails and directly at you into the screen.
One of the designers who worked on the game was a 16-year-old math and chess whiz named Demis Hassabis. He is now better known as the Nobel Prize-winning founder and CEO of Google’s DeepMind AI division:
“I was pretty young at the time. I started working at Bullfrog when I was 16, and was immediately working on Theme Park actually,” says Demis Hassabis, the brain behind Theme Park and the simulation that powered it. Demis had already done a couple of summer internships at the Guildford-based studio, where he had done some level designing on Syndicate and AI programming for Populous II. Considered to be something of a child prodigy, Demis had two main passions: chess and playing and programming videogames. While he had been playing chess internationally, it was ultimately his curiosity about artificial intelligence within gaming that has since gone on to be the defining part of his career. That all began in earnest with Theme Park. “Funnily enough I wasn’t sure I wanted to work on it,” he laughs, “because I was keen to make a game like Syndicate, Populous, Powermonger and the other sort of earlier games. I was more keen to do that kind of game, maybe a strategy game.”