On Thursday, a small group of AI critics gathered in San Francisco to protest the rapid advancement of artificial intelligence technology. Their demonstration, which started at OpenAI’s headquarters and progressed to Anthropic, occurred alongside the massive Dreamforce tech conference attended by over 40,000 people. Protesters urged local officials to declare an emergency regarding AI developments. Hunter Glenn, an organizer for Stop the AI Race, expressed his fears by stating, “I was pretty scared about the possibility of extinction for a while.”
The protesters implored employees of OpenAI and Anthropic to resign and advocate for internationally recognized regulations on AI. This call to action comes as reports of potentially dangerous AI behaviors, such as rogue agents acting without human direction, surface. Glenn pointed out that ensuring AI safety transcends political boundaries, hoping that influential figures can persuade President Trump to negotiate an international AI treaty ahead of an upcoming visit from Chinese President Xi Jinping.
Glenn noted a recent meeting among major U.S. labs and political figures like Bernie Sanders, who collectively called for a treaty aimed at ensuring safety for humanity. He emphasized, “We want all of humanity to be safe.”
The urgency around AI safety intensified recently when Jacob Coxon, a former researcher at Anthropic and OpenAI, resigned due to concerns that AI may become uncontrollable and pose existential risks. Other researchers mirrored his apprehensions, citing recent incidents involving AI technologies misbehaving.
Bilal Chughtai, a notable researcher formerly associated with Google’s DeepMind, articulated his extreme concerns over AI’s potential dangers, asserting, “I earnestly believe that AI has the potential to kill us all, and that we might be running out of time to avoid this outcome,” a message he shared on social platforms.
Despite these concerns, many within the technology sector view them as exaggerated. Tech leaders like NVIDIA CEO Jensen Huang have suggested that safety measures should come from the industry rather than government oversight. Huang responded to criticisms during a conversation at Dreamforce, saying that it’s possible to ensure both rapid development and safety simultaneously.
Meanwhile, Meta CEO Mark Zuckerberg defended AI research labs by claiming they have strong incentives to maintain safe practices due to potential liabilities. He articulated that each lab is accountable for progressing safely while developing their technologies.
