Pull the plug! Pull the plug! Stop the slip! Stop the slip! For a few hours this Saturday, February 28, I saw two hundred anti-AI protesters walk through London’s King’s Cross tech hub, home to the UK headquarters of OpenAI, Meta and Google DeepMind, chanting and waving signs. The march was organized by a coalition of two separate activist groups, Pause AI and Pull the Plug, who called it the largest protest of its kind to date.
The range of concerns on the show included everything from online slop and abusive images to killer robots and human extinction. A woman wore a large homemade billboard on her head that read “Kun Hoga Whose Tool?” (with Os in “tool” cut as eye-holes). There were signs that said “Pause before reason” and “EXTINCTION=BAD” and “Demis the Menace” (referring to Google DeepMind CEO Demis Hassabis). Another said simply: “Stop using AI.”
An old man wearing a sandwich board that read “AI? over my dead body” told me he was worried about AI’s negative impact on society: “It’s about unemployment risks,” he said. “Satan seeks work with idle hands.”
These are all familiar things. Researchers have been calling out the real and hypothetical harms caused by creative AI—most notably OpenAI’s ChatGPT and Google DeepMind’s Gemini—for years. What has changed is that these concerns are now being raised through protest movements that can gather significant masses of people to take to the streets and shout about it.
The first time I bumped into anti-AI protesters was outside a London lecture hall in May 2023 where Sam Altman was giving a speech. Two or three people stood in front of an audience of hundreds. In June last year Pause AI, a small but international organization founded in 2023 and funded by private donors, drew a crowd of a few dozen people to a protest outside Google DeepMind’s London office. It felt like an important addition.
“We want people to know that Pause AI exists,” Joseph Miller, who heads Pause AI’s UK branch and is co-hosting Saturday’s march, told me on a call the day before the protest: “We’re growing very fast. In fact, we seem to be matching the growth of AI ourselves.”
Miller is a PhD student at the University of Oxford, where he studies mechanistic interpretation, a new area of ​​research that involves trying to understand what happens inside LL.M. as they perform a task. His work convinced him that technology could be forever beyond our control and that this could have catastrophic consequences.
It doesn’t have to be a rogue superintelligence, he said. All you needed was someone to put the AI ​​in charge of nukes. “The more stupid decisions humanity makes, the less powerful AI has to be before things get worse,” he said.
After a week in which the US government tried to force Anthropic to allow Anthropic’s LLM Claude to use for any “legitimate” military purposes, such concerns are less likely. Anthropic stood its ground and OpenAI signed an agreement with the DoD instead. (OpenAI declined an invitation to comment on Saturday’s protest.)
For Pause AI member and protest co-organizer Matilda da Rui, AI is the last problem humans will face. She thinks that technology will either allow us to solve every other problem we have, or it will wipe us out and there will be no problems left. “It’s a mystery to me if anyone would really pay attention to anything else if they actually understood the problem,” he told me.
And yet despite this urgency, the atmosphere at the march was pleasant, even fun. There was no sense of anger and little sense that let alone life – the very survival of our species was at stake. This may be due to the broad coalition of interests and demands that the protesters brought with them.
A chemistry researcher I spoke to outlined a wide range of complaints, ranging from the conspiratorial (that data centers emit infrasound below the threshold of human hearing, causing insensitivity in those living around them) to the reasonable (that the proliferation of online AI slop was making it difficult to retrieve). The researcher’s solution was to make it illegal for companies to exploit the technology: “If you can’t make money from AI, it won’t be a problem.”
Most people I spoke to agreed that technology companies probably wouldn’t take notice of this kind of protest. “I don’t think pressure on companies will ever work,” Maxim Fornes, global head of Pause AI, told me when I caught up with him in March: “They’re apt to ignore the issue.”
But Fournes, who worked in the AI ​​industry for 12 years before joining Pause AI, thinks he can make things difficult for these companies. “We can reduce the race by creating protections for whistleblowers or showing the public that working in AI is not a sexy job, it’s actually a scary job — you can dry up the talent pipeline.”
In general, most protesters hoped to use the platform to educate more people about the issues and push for government regulation. Organizers presented the march as a social event, encouraging anyone who cared about the cause.
It looked like the job was done. I met a guy who worked in finance who tagged along with his roommate. I asked why he was there. “Sometimes you don’t have that much on a Saturday,” she said. “If you can see the logic of the argument, it kind of makes sense to you, then it’s like ‘yeah, sure, I’ll come along and see what it’s like.’
He believed that the concerns surrounding AI were difficult for anyone to fully oppose. “It’s not like a pro-Palestinian protest, where you’d have people who might disagree with the cause,” he said. “With that, I feel like it’s very difficult for anyone to be completely opposed to what you’re marching for.”
After passing through King’s Cross, the march ended at a church hall in Bloomsbury, where tables and chairs were arranged in rows. Protesters wrote their names on stickers, stuck them to their chests and awkwardly introduced themselves to their neighbors. They came here to find out how to save the world. But I had a train to catch and I left them to it.