"I Like Money, Bro": Ex-OpenAI, Anthropic Researcher On "Blood Money" Stock

"I Like Money, Bro": Ex-OpenAI, Anthropic Researcher On "Blood Money" Stock

"I like money, bro."That's how former OpenAI and Anthropic researcher Jacob Coxon responded when asked why he hadn't given up his financial stake in OpenAI, despite believing that the technology being developed by AI companies could eventually threaten humanity.Coxon was speaking to Stanford political science professor Robert Reich during a conversation hosted by Stanford University's Institute for Human-Centered Artificial Intelligence (HAI) last week.Reich pressed Coxon on whether holding on to OpenAI stock could undermine his credibility when speaking about the dangers of AI development. Coxon acknowledged that giving up his shares could help him take the moral high ground, but made little effort to disguise his preference."I like money, bro. I agree, it's blood money."Watch our quick take:Reich's response:"But you could take the moral high ground."Coxon conceded that he could, before Reich pointed out that he wasn't doing so.When Reich asked why he wouldn't give up the shares, Coxon said that while it might be easier to make his arguments about AI safety without a financial interest in OpenAI, he liked the money.He then offered a more qualified explanation for why he wasn't considering donating the proceeds."If I thought for a moment that donating it would change anything, then sure, I'd do it."Reich then suggested that giving away the money might give Coxon greater moral credibility.But Coxon pushed back with an answer that drew another sharp response from the professor."I agree, but you could also stop asking the question."Reich called it a dodge, to which Coxon responded: "Yeah."Coxon has not disclosed how much OpenAI equity he holds or what it is worth. Also, retaining OpenAI shares is different from retaining Anthropic equity.Coxon reportedly left Anthropic before any of his equity there had vested. He had spent four months at the company, while its employees needed to stay for six months for their stock to vest. He had clarified then too that he still held equity in his previous employer, OpenAI.The Researcher Who Said AI Could Kill Us AllCoxon made global headlines in September after publicly announcing his resignation from Anthropic, warning that the race to develop increasingly powerful AI systems was putting humanity at risk.He had spent approximately three years working on pretraining research at OpenAI and Anthropic, and accused both companies of failing to act responsibly."The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt," he had said."They are racing straight to self-improving superintelligence and gambling with our lives."Self-improving AI refers to systems that could potentially help develop more capable versions of themselves, accelerating technological progress beyond the ability of humans to adequately monitor or control it. While a part of the AI industry believes the technology has already reached that point, others put it around a year away.Coxon's concern was that such systems could become superhuman, gain access to resources and capabilities, and pose risks that existing safety measures might not be equipped to handle.His resignation was particularly notable because he had worked inside two of the world's leading AI labs, rather than simply commenting on the industry from outside.Now He Isn't Sure Quitting Was The Right DecisionThe conversation at Stanford also offered a glimpse into Coxon's own uncertainty about walking away from the AI industry.Despite his public warning about the dangers of advanced AI, the former researcher admitted that he was still unsure whether resigning from Anthropic had been the right decision."I'm not sure I made the right decision in leaving. I have not been convinced that I would not be having more of an impact by staying and doing safety research."He then left open the possibility of returning to an AI lab."Who knows? Maybe I'll go back. It's not completely out of the question."Coxon also described himself as confused about the entire situation.To be sure, Coxon isn't necessarily saying that he has changed his mind about AI risks, he's questioning whether leaving the industry is the most effective way to address them.The Bigger AI Safety ArgumentThere is clearly a contradiction between the financial incentives driving AI development and the warnings coming from people inside the industry.Coxon argued that frontier AI companies have financial incentives to develop increasingly powerful systems quickly, even as they publicly talk about slowing down."As financial entities, the frontier labs want to go as fast as possible. They only give lip service to slowing down in order to appease their employees so they don't have mass exodus."Coxon is still trying to work out what comes next for him, but the irony of someone who calls his OpenAI equity blood money but isn't prepared to give it up is not lost on anyone.

Original Source

Read the full article at Ndtv →

KhanList aggregates and links to publicly available news content. We do not host full articles from third-party sources. Always verify important information with original sources.