How about not moving forward with a technology that could kill all humans?

So here’s an idea. Instead of moving forward with a technology that its lead inventors say may soon have the power to kill humans, how about not moving forward with it?

This radical notion is inspired by a warning from the man who set up the Prime Minister’s Artificial Intelligence Task Force. Matt Clifford observed that, “You could have a really very dangerous threat to humans that could kill many humans, not all humans, just from where we expect the models to be in two years’ time. ” To other thoughts maybe I am overreacting. His whole comment was more subtle and it’s not all humans anyway. lots of them.

But a similar apocalyptic warning comes from leading figures in its development, written under the auspices of the Center for AI Safety. In a laudatory brief warning, AI industry insiders stressed that: “Reducing the risk of extinction from AI must be a global priority alongside other societal-level risks such as pandemics and nuclear war.” The heads of Google DeepMind, OpenAI and many others have taken the time from inventing the technology that could wipe out all human life to warn the rest of us that, in fact, there is something to prevent this from happening. should be done.

And these people are considered geniuses? Across the potting shed in England, there are quite a few wacky people who have invented a new machine that may be brilliant but may also burn down their house, and most of them have managed to make it work by themselves. Maybe the device isn’t such a great idea after all.

This is where little-known inventors go wrong. Perhaps instead of figuring out the risks for themselves, they actually had to score several billion pounds of VC funding and then write a letter to the local council warning that they really should be regulated.

I agree to be serious that great things are expected of artificial intelligence, many of which do not include the destruction of the human race. Many argue that AI could play an important role in delivering a carbon-free future, although perhaps this is just a euphemism for eliminating humanity.

Equally important, the progress already made cannot be invented. But already the AI ​​chatbot is falsifying information — or “hallucinating” as its developers like to put it — and its inventors aren’t entirely sure why. So there appears to be an argument for slowing down and ironing out that teenage wrinkle before using, you know, extinction-level technology.

A rhetorical view of tech leaders calling for lease themselves is that they are responsible and it is other irresponsible actors they are concerned about. They want to do more, but you see, the people at Google can’t let themselves be defeated by the people at Microsoft.

So these warnings are an attempt to jolt politicians and regulators into action, which is their game be damned, given that world leaders have such an impressive record of responding cooperatively and intelligently to extinction-level threats. Is. I mean come on. He has mentioned this in the US Congress. I don’t think we can ask for more. And the British government is now on the case, which would be more convincing if it weren’t still struggling to process asylum seekers less than 18 months later.

Fortunately, the warnings will actually jolt governments into useful action. Perhaps this leads to global standards, international agreements and a moratorium on the development of the killer.

Either way, the conscience of the AI ​​gurus is now clear. They’ve done everything they can. And if one day, around 2025, machines do indeed gain the power to erase us—sorry, many of us—I have a feeling that at the last second the AI ​​will make the last query to those brilliant minds that deliberately overcame such a situation. With technology that can destroy us, at that stage without figuring it out you know how to stop doing so.

“Why did you go ahead knowing the risks?” asks Skynet. And at his last second the genius replies: “What do you mean? We signed a statement.

Follow Robert on Twitter @robertshrimsley and email it robert.shrimsley@ft.com

follow @FTMag Be the first to find out about our latest stories on Twitter