There’s been a lot of talk in the media recently about Artificial Intelligence (AI). Facebook uses it for targeted advertising, photo tagging, and news feeds. Microsoft and Apple use it to power their digital assistants, Cortana and Siri, and Google’s search engine has utilised AI from the beginning. There appears to be something of a chase to create flexible, self-teaching AI that will mirror human learning and apparently transform our lives.
There have been some big-name doom-mongers on this subject, however. Elon Musk thinks AI is probably humanity’s “biggest existential threat”. Stephen Hawking fears that AI may “replace humans altogether”. Bill Gates agrees with both of them. Me, I’m not so sure; surely you can always turn a machine off?…(on the other hand, have you ever tried closing Skype?)
This concept of computers/machines gone bad is a well-worn theme in science fiction, with the Terminator series of films an obvious example, but it was back in 1968, in Stanley Kubrik and Arthur C Clarke’s seminal 2001: A Space Odyssey, that we were introduced to our first electronic wrong ‘un, HAL 9000. HAL (from Heuristically programmed ALgorithm, apparently, though some have conjectured an easily-decrypted code version of IBM) is a sentient computer controlling the systems of the Discovery One spacecraft on its mission to Jupiter.
HAL is initially regarded as another member of the crew, engaging genially with its human colleagues, playing chess with them and so on. However, he begins to malfunction in subtle ways. As the malfunctioning deteriorates, the crew members discuss the possibility of disconnecting HAL’s cognitive circuits. Unfortunately, HAL can read lips and discerns their plan, and his programmed directives to protect the mission lead him to reason that he must kill the astronauts. In this classic scene, crew member Dave Bowman is outside the main craft in a “pod” and is seeking re-entry, asking HAL to open the pod bay doors. HAL (voiced chillingly by Douglas Rain) isn’t playing ball…