Nearly half of OpenAI’s staff focused on AGI (artificial general intelligence) safety have left the company in recent months, as reported by former researcher Daniel Kokotajlo. The departures include several key figures, such as chief scientist Ilya Sutskever and Jan Leike, who were responsible for the “superalignment” team working on controlling superintelligent AI. Kokotajlo attributes these resignations to OpenAI’s growing emphasis on product development and commercialization, which appears to be sidelining the crucial safety research needed to address the risks posed by advanced AI technologies.
This shift in focus has raised significant concerns among AI researchers and safety advocates about OpenAI’s commitment to mitigating potential risks associated with developing AGI. Kokotajlo describes a “chilling effect” at OpenAI, where those trying to publish research on AGI risks face increasing resistance from the company’s leadership. The company’s recent hires, including a new CFO and chief product officer, further highlight its pivot towards commercialization over safety and long-term risk assessment.
Despite these changes, OpenAI insists that it remains dedicated to AI safety, stating that it engages with various stakeholders to address potential risks. However, Kokotajlo and others remain skeptical, worrying that profit motives and the race to be first at achieving Artificial General intelligence (AGI) are overshadowing the original mission to develop AGI in a way that benefits all of humanity. This trend has broader implications, not only for OpenAI but also for the overall trajectory of AI development, as other major tech companies race to advance their own AI capabilities.
