Three men with deep concerns about the threat artificial intelligence could pose to human existence have embarked on hunger strikes this month in an attempt to get the attention of prominent companies working to develop super-powerful AI. 

On Sept. 1, activist Guido Reichstadter set up an A-frame and folding chair outside the Howard Street office building that’s home to the AI firm Anthropic, beginning a hunger strike during which he is intaking electrolytes and vitamins — but no calories. 

In an X post two days later, Reichstadter said he is “calling on Anthropic’s management, directors, and employees to immediately stop their reckless actions which are harming our society and to work to remediate the harm that has already been caused.” Reichstadter made the stakes as he saw them plain: “We are in an emergency. Let us act as if this emergency is real.”

Days later, a second activist and former AI researcher, Michaël Trazzi, set up similarly outside the headquarters for Alphabet’s AI research lab, Google DeepMind, in London’s Pancras Square.

In an X post announcing his strike, Trazzi expressed support for Reichstadter’s action and belief that the stakes with AI companies’ frontier models are too high to pursue even with technical attempts to set up guardrails. Like Reichstadter, he’s asking DeepMind to halt development if other companies similarly pause their efforts.

“Experts are warning us that this race to ever more powerful artificial general intelligence puts our lives and well-being at risk, as well as the lives and well-being of our loved ones,” Trazzi wrote in part.

On Friday afternoon, a Standard reporter went to Howard Street to meet with Reichstadter, who was surrounded by signs splashed with dire warnings about AI risks and was expecting to have a call with Trazzi in London. Reichstadter, however, said Trazzi had experienced a fainting episode and stopped fasting at the recommendation of two doctors. 

Another protester, Denys Sheremet, planned to continue the hunger strike he’d begun with Trazzi outside DeepMind’s offices six days prior.

“I want to thank everyone who has supported me in this journey, both in person and online,” Trazzi said Thursday on X. “My hunger strike ends now, but the movement continues, with Guido and Denys still on strike in front of Anthropic and DeepMind.”

A Standard reporter spoke with Sheremet Saturday morning, when he FaceTimed with Reichstadter from London. The two discussed the strategies they have been deploying to get the attention of AI executives, and consoled one another in their struggle.

In the middle of the call, Reichstadter took the time to coach Sheremet on his nutritional regimen. Since a meal of chicken and pasta for dinner on Aug. 31, Reichstadter, 45, has consumed only water, vitamins, and electrolytes. Then, he told Sheremet the story of Sodom and Gomorrah, in which Abraham negotiates with God to save the city of Sodom so long as Abraham can find ten righteous people.

“Let’s see if there are 10 people of integrity at Anthropic,” Reichstadter said Sheremet nodded affirmatively.

“Keep it up,” Sheremet encouraged before they hung up.

Reichstadter said he delivered a letter to Anthropic CEO Dario Amodei through the lobby’s security guard on the first day of the protest, requesting a meeting to discuss the technology’s potential dangers. The company has not publicly responded to the protest, and Reichstadter said he plans to continue the hunger strike indefinitely until Amodei agrees to meet with him.

“This company is building technology that their CEO acknowledges puts my life at risk, my children’s lives, everyone in this society,” Reichstadter said.

Reichstadter, who previously conducted a 14-day hunger strike in April 2022 demanding Miami-Dade Mayor Daniella Levine Cara declare a climate emergency, said he chose the extreme form of protest because AI companies are “running into a minefield” by racing to develop artificial general intelligence, or AGI.

He was joined on Friday by Phoebe Thomas Sorgen, 71, a longtime activist who said she may begin her own hunger strike to support the cause. Thomas-Sorgen said she became concerned about AI risks about three months ago after meeting Reichstadter and other activists at Travis Air Force Base in Fairfield during a protest against the Trump administration’s immigration policies.

“It’s a horrible threat that we’re facing,” Thomas-Sorgen said, citing concerns about job displacement, data privacy, and environmental impacts from energy-intensive data centers.

Both referenced statements from AI researchers, including Geoffrey Hinton, who won a Nobel Prize last year and has warned about potential extinction risks from artificial superintelligence. A statement on AI risk signed by numerous researchers, including Anthropic CEO Dario Amodei, warns that “mitigating the risk of extinction from AI should be a global priority.”

Reichstadter said his three demands to Anthropic are simple: stop endangering lives through AI development, use the company’s resources to halt the global AI race through international negotiations, and have Amodei explain why he believes he has the right to risk human lives. 

Anthropic, founded in 2021 by former OpenAI executives including Amodei, says it has emphasized AI safety research alongside its commercial products. 

“I’m here to get the word out as long as I can,” he said. “Everyone that understands the situation we’re in has some responsibility to wake them up to the danger.”