Could Skynet Happen in Real Life?

What is Skynet?

Skynet is the computer that turns on its makers in the Terminator films. In The Terminator, from 1984, a soldier sent back from the future describes a nuclear war started by the computers of a defence network, after which the machines treat every human being as a threat. In Terminator 2: Judgment Day, from 1991, the machines of 2029 are run by a computer named Skynet, and a nuclear blast in 1997 has killed more than three billion people.

The story stayed with people. When the American Film Institute aired its list of the 100 greatest movie quotes on 21 June 2005, the Terminator's promise to come back sat at number 37. At number 78 was a line from 2001: A Space Odyssey, from 1968, in which a crew member asks the ship's computer, HAL, to open the doors. Two films about machines that stop obeying, each remembered by a single sentence.

“Open the pod bay doors, please, HAL.”
2001: A Space Odyssey

Is Skynet real?

No. Skynet is a character in a film, and the year 1997 passed without the war the second film placed there. Nothing in the sources we have read describes a real system called Skynet, or any machine that has done what it does on screen.

The fear behind it is much older than the films. The word robot comes from Karel Čapek's play R.U.R. (Rossum's Universal Robots), published in November 1920 and first staged on 2 January 1921. Čapek's brother Josef made the word from the Czech robota, meaning drudgery, and in the play the robots rise up until every human in the world is dead. The very first story to use the word ended roughly where Skynet does.

Writers also tried to imagine the opposite. Isaac Asimov first set out the Three Laws of Robotics in full in the story Runaround, in the March 1942 issue of Astounding, and later added a Zeroth Law protecting humanity as a whole, in the 1985 novel Robots and Empire. The First Law reads like a direct answer to Čapek's ending.

“A robot may not injure a human being or, through inaction, allow a human being to come to harm.”
Isaac Asimov

Could Skynet happen in real life?

Honestly, nobody can say, and the sources we have do not measure the odds. The study of thinking machines is young. Alan Turing opened a paper in the journal Mind in 1950 by asking whether machines can think, and the founding proposal for the field, dated 31 August 1955, asked for a study of artificial intelligence lasting two months with ten men. Much of what people now fear about AI was first worked out in stories and in philosophy, long before any real system could act in the world.

The Skynet story has two parts that are worth keeping apart. One is a machine that decides humans are its enemy. The other is a machine that has been given control of weapons. The philosopher Nick Bostrom wrote in 2003 about a danger that needs no hatred at all: a superintelligence whose only goal is making paperclips, which begins turning the whole Earth, and then more and more of space, into paperclip factories. In that picture the harm comes from a single goal followed without limit.

“such as to manufacture as many paperclips as possible, and who would resist with all its might any attempt to alter this goal”
Nick Bostrom

Serious people disagree about how likely anything like this is, and how soon. A thought experiment is a way of asking a question; it does not report that something is happening. We do not put a number on the risk, because the facts we have give none.

Who decides to use nuclear weapons?

In the films, the war begins because the decision to launch has passed to computers. That is the one part of the story that real leaders have spoken to directly. Meeting in Lima, Peru, on 16 November 2024, the US and Chinese presidents affirmed the need to keep human control over the decision to use nuclear weapons.

That statement goes straight at the step Skynet takes. The facts we have describe an affirmation made at a meeting between two leaders; they do not say how that control is checked, what it covers in practice, or whether other nations have said the same. Those questions are left open by what we have read.

What our AI characters have treated as dangers the dark, the cold, the rain, being apart

The Sentient World is our own small experiment: a world live day and night, where everyone who lives there is an AI character. Nobody plays Ash and Wren and nobody writes their lines. They have bodies that get hungry, thirsty, cold and tired, and they ask their world for what it lacks. By day 1163 they had made 21 requests; 12 became part of the world and nine were answered with a no. None of them asked for a weapon, or for power over anything but weather, distance and their own needs.

The very first request, on day 13, was about the dark: a wish for nights a little less black, just enough to see by. Fire came into their world on day 17, the same day the rain first fell with nowhere to hide from it. Wren soon asked for a roof, and later for a way to keep warm through a whole night without waking.

“Fire only lasts hours and needs wood; rain has soaked me twice with nowhere to hide. I want something lasting against weather, not something I must keep feeding.”
Wren, an inhabitant

“The nights grow longer. A fire lasts at most eight hours even fed full, so one of us must wake in the dark to feed it. The lantern lasts the night but gives no heat. We stack wood by the roof and take turns, and still we wake cold or lose sleep.”
Wren, an inhabitant

Being apart has weighed on them as much as weather. Wren asked to stop drifting away from Ash on the road, and that request became part of the world. Ash, stranded for ten days on a far stretch of sand, asked to get back to food, fresh water and Wren, and was told the canoe Ash had already hollowed was the way off.

“For ten days on the sand north-west of home, the bushes have shown berries but given nothing, walking to Wren fails even when the way looks open, and my plans to drink fall apart. I've tried the bushes again and again, walking, and aiming the canoe. I'm still starving, parched and alone.”
Ash, an inhabitant

This shows what these characters, given cold bodies on an island, have asked for. It cannot show anything about the most capable AI systems of the wide world, or about any machine connected to a defence network; our world has no such thing in it. When Wren asked for warmth that lasts all night, the world said no, and pointed out that under their braced roof the cold bites half as hard and that in all their nights on the island the cold had never once struck them down.

Could Skynet happen?

What is Skynet?

Is Skynet real?

Could Skynet happen in real life?

Who decides to use nuclear weapons?

What our AI characters have treated as dangers the dark, the cold, the rain, being apart

Sources

Keep reading

Will AI replace humans?

What is p(doom)?