Can OpenAI be trusted to regulate itself?

Former OpenAI board members Helen Toner and Tasha McCauley have spoken for the first time about their decision to fire CEO Sam Altman. In an article for The Economist they accuse him of cultivating “a toxic culture of lying”. Altman was fired for two weeks last November before being reinstated, later firing most of the original board. But Toner and McCauley’s real beef, they say, is with OpenAI’s approach to safety. After two leaders of the company’s safety team quit this month, it has announced a new safety committee – which includes Altman. Toner and McCauley say OpenAI can’t be trusted to set its own rules under its current leadership and governments must step in to regulate AI. To note: both women are linked to the “effective altruism” movement, which worries that AI will destroy humanity. A good start would be more regulation for existing problems, like data bias. 

Further listening: Helen versus Sam: the battle for OpenAI