The Doomers Who Shaped the AI Safety Freakout

The Wall Street Journal has published a report on a community of artificial intelligence researchers who work from a co-working space in Berkeley called Constellation. According to the article, the group has spent more than a decade focused on the topic of AI safety. The report describes members of this community as 'doomers,' a term it applies to people who hold concerns about the risks of advanced artificial intelligence. The article's title states that these doomers shaped what it calls the AI safety 'freakout.'

The article states that the resignation this month of an Anthropic researcher, Jacob Coxon, woke up the American public to concerns that runaway AI development could end humanity. The report presents the resignation as the event that brought those concerns to a wider public audience. It also states that such 'doomers' were among the earliest employees at OpenAI and Anthropic, the two companies named in the article.

The report situates the Berkeley co-working space and the community that uses it within the broader history of AI safety work, noting that its members have been focused on the subject for more than a decade. The article states that the concerns associated with this group centre on the possibility that runaway AI development could end humanity, and it presents the Coxon resignation as the moment those concerns reached the American public. It names OpenAI and Anthropic as the employers of some of the earliest members of this community.