Artificial general intelligence, the prospect of a synthetic entity meeting or exceeding the cognitive power of a human, is a polarizing topic.
For some, like OpenAI CEO Sam Altman, it's a mission statement — the lodestar guiding the company's $300 billion operations. For others, like activist and organizer Guido Reichstadter, it's an existential threat to be resisted at any cost.
Reichstader is now on day nine of a grueling anti-AGI hunger strike in front of the San Francisco headquarters of the AI giant Anthropic. He's previously been arrested for barricading OpenAI's doors, the legal fallout of which he says is ongoing.
"When you get into the systems that these companies are trying to build, that exceed human capabilities in every important respect, this is an entirely new class of danger," Reichstadter told Futurism from the sidewalk in front of Anthropic's office. " Whatever speculative benefits these systems could bring, the fact is that none of these companies have any idea, any practical roadmap for how to control that power once they build it."
"None of these companies have a right to do what they're doing, which is consciously endangering my life, my family's life, all of our lives," the activist continued. "The correct thing for them to do is stop the global race toward really dangerous AI that we're all involved in."
At first glance, the Anthropic might seem like an odd choice for an anti-AGI protest. Dario Amodei, Anthropic's CEO, has previously called AGI a "marketing term," positioning himself as something of a contrarian in a field of otherwise duplicitous billionaire tech founders.
Yet underneath Amodei's adult-in-the-room schtick is an optimism that AGI — or something like it — is on the horizon.
Anthropic's cofounder, Benjamin Mann, has previously predicted we will "hit" AGI as early as 2027 or 2028, which he says will usher in a baseline unemployment rate of at least 20 percent. Amodei has likewise proclaimed that human-level AI will somehow solve all manner of social ills, from alleviating mass poverty to "curing most mental illness."
And as Reichstadter notes, if Anthropic's CEOs were truly rattled by the prospects of AGI, you wouldn't know it by the way they've scaled up their own operations. The company has experienced massive growth since it launched its proprietary AI model, Claude, in 2023, and recently completed its sixth funding round to bring in some $13 billion in investments to further develop AI.
For Reichstadter, the hunger strike is both a way to mobilize those critical of the tech industry's posturing around AGI, and to embody the mortal threat to life Anthropic and its fellow AI companies represent. Until the CEO acknowledges him, Reichstadter says the strike will continue.
"If Dario Amodei isn't willing to [stop], he owes me an explanation, to my face," the activist said. "My goal is to either stop work, or receive an explanation."
At day nine, Reichstadter is well within the "ketosis phase" of the hunger strike, the point where the body starts to burn fat for energy in the absence of food. At the moment, he says he's "holding up okay," adding that he previously completed a two-week hunger strike as part of a campaign addressing climate change.
As for mobilizing fellow travelers, the activist's efforts are already bearing fruit. Days after he announced his protest from San Francisco, a pair of demonstrators took up hunger strikes outside of Google's DeepMind offices in London. Reichstadter says the protests weren't coordinated ahead of time, but he urgently welcomes all the help he can get.
"Why are we sitting here while these people are putting our lives at risk?" he challenges. "Let's go lock their f**cking doors shut."
More on AGI: Billion-Dollar AI Company Gives Up on AGI While Desperately Fighting to Stop Bleeding Money