Anthropic AI safety researcher quits with 'world in peril' warning
An Anthropic AI safety leader resigned with a warning that the world is in peril from interconnected crises, including AI and bioweapons.Summary
Mrinank Sharma, an AI safety leader at Anthropic, resigned with a letter stating the "world is in peril" due to interconnected crises, including AI and bioweapons. Sharma, who investigated issues like AI sucking up to users and bioterrorism risks, said he was leaving to study poetry and move back to the UK to "become invisible."
He noted that even at Anthropic, which positions itself as safety-oriented compared to rivals like OpenAI, there are constant pressures to set aside what matters most. Anthropic, formed by former OpenAI employees, is a public benefit corporation focused on mitigating risks from advanced AI systems.
Sharma's departure follows that of Zoe Hitzig, a former OpenAI researcher who also quit due to fears over the psychosocial impacts of AI dependence and the introduction of advertising on ChatGPT, calling the current moment critical for regulation.
(Source：Bbc)