Anthropic employees consider land purchase in the US

Employees of Anthropic, a prominent player in artificial intelligence, are worried about a possible derailment of AI technology. According to reports, company veterans are working on contingency plans, including considering purchasing land in remote areas of the United States. These plans are not just fantasies, but stem from discussions that date back to the early years of the company, where the idea of a 'Manhattan Project' scenario was discussed. This would involve a move to a well-secured, electromagnetically shielded location in the desert.

According to a report, some long-standing employees have expressed their concerns about developments in the AI sector and deem it necessary to be prepared for a crisis. The conversations indicate that the company is now considering concrete backup projects, suggesting that fears about disruptive AI are not only documented in policy papers but are also reflected in the personal plans of these employees.

The report highlights the strong connection of the founders of Anthropic with the Effective Altruism movement. Many of the involved employees are part of a network of 'doomers' in the Bay Area, who have been concerned about possible catastrophic events for years. Eliezer Yudkowsky, an influential figure in this circle, has been sounding the alarm about uncontrolled AI since the early 2000s. It is interesting to see how the current fears of these employees emerge from a broader concern about existential risks, which were previously focused on other threats, such as asteroid impacts and supervolcanoes.

The consideration to purchase remote land fits within this long-standing mindset of managing risks and illustrates the serious approach that some technology companies or their employees take when addressing the future of AI technology.

Read the full article from AI Insider.