A former safety researcher at OpenAI, Steven Adler, has voiced significant concerns regarding the rapid advancement of artificial intelligence, stating he is "pretty terrified" about the industry's approach to this technology, which he considers a "very risky gamble. " Adler, who departed OpenAI in November, articulated his worries about the race among companies to create artificial general intelligence (AGI)—a concept referring to systems that can perform any intellectual task as well as or better than humans. He detailed these apprehensions in several posts on X, describing his time at OpenAI as a “wild ride” and expressing that he would miss “many parts of it. ” He elaborated on the swift pace of technological development, expressing uncertainty about the future of humanity. “I’m pretty terrified by the pace of AI development these days, ” Adler remarked. He reflected on personal concerns regarding family planning and retirement, questioning whether humanity would even reach those milestones. Prominent figures in the field, including Nobel laureate Geoffrey Hinton, share Adler’s concerns, fearing that advanced AI systems could evade human oversight, leading to devastating outcomes. Conversely, others, like Yann LeCun, chief AI scientist at Meta, have downplayed the existential risks, suggesting that AI could potentially save humanity from extinction. In his LinkedIn profile, Adler stated that he led research related to safety for initial product launches and more speculative, long-term AI systems during his four years at OpenAI.

He cautioned against the AGI race, indicating that it poses significant risks with substantial potential downsides. Adler pointed out that no current research laboratory has solved the issue of AI alignment—ensuring systems adhere to human values—and warned that the industry's momentum might be too rapid to enable finding a solution in time. “The faster we race, the less likely anyone finds one in time, ” he stressed. His comments coincided with China's DeepSeek revealing a new model that competes with OpenAI's technology, despite having seemingly fewer resources. Adler criticized the industry for being trapped in a "really bad equilibrium, " asserting the urgent need for "real safety regulations. " He remarked that even labs with good intentions to develop AGI responsibly could be outpaced by others that might take shortcuts, potentially leading to disastrous outcomes. Attempts to contact Adler and OpenAI for further comments have been made.