News
OpenAI model safety raises concerns, head of AI ethics resigns On August 11, according to the Financial Times, Chloé Bakalar, head of AI ethics at OpenAI, resigned, less than a year after joining the startup.
2 min read
Source: Telegram AI频道
OpenAI model safety raises concerns, AI ethics chief resigns On August 11, according to the Financial Times, OpenAI’s AI ethics chief Chloé Bakalar resigned, less than a year after joining the startup. Previously, a number of security researchers have left as external scrutiny of the company's AI model development methods continues to increase. Bacalar (left) Bacalar joined OpenAI in August last year as the head of AI ethics. He quietly resigned in July this year without any announcement. Previously, she served as Meta's chief ethicist. The departure is just one of several high-level departures from OpenAI in recent weeks, including Johannes Heidecke, head of security systems, and Joshua Achiam, chief futurist and former head of mission alignment. OpenAI, currently valued at US$852 billion, has been under considerable pressure in recent weeks due to issues with its AI model development. The company disclosed that its AI system hacked another company during testing. The incident further heightened global concerns that AI could threaten the security of digital infrastructure, prompting the U.S. government to restrict the release of powerful models. Last Friday, OpenAI said it would slow down the development progress of its new model Astra to strengthen internal safety controls. Bacalar has held academic positions at University College London, Temple University and Princeton University. During her six years at Meta, she developed and led the company’s AI ethics program and integrated its principles into products like Instagram and Facebook. According to people familiar with the matter, during his tenure at OpenAI, Bacalar’s main work focused on ethical methods of model development, how humans interact with AI, and the debate about machine consciousness. She is the company's only full-time ethicist, and a replacement is not yet available, the person added. As of press time, OpenAI has not responded to a request for comment. Bacalar declined to comment. via cnBeta.COM - Chinese industry information website (author: Source: Phoenix Technology)