Anthropic called Thursday for artificial intelligence companies worldwide to consider slowing AI development, warning the technology is approaching the ability to autonomously design its own successors.
The AI firm wrote on its website that “if it were possible to effectively slow the development of this technology to give ourselves more time to deal with its immense implications,” the company believes that “would likely be a good thing.”
“Without a global coordination mechanism, companies and governments will have to make difficult decisions about safety while under competitive and geopolitical pressures.”
The blog post acknowledged a dilemma: slowing AI development could allow “the least cautious actors” to catch up technologically, potentially leaving everyone less safe.
Anthropic stated it would support a slowdown or temporary pause in frontier AI development “if other developers at or near the frontier also did so in a verifiable manner.”
Any meaningful pause would require “multiple well-resourced labs at or near the frontier, in multiple countries, agreeing to stop under the same conditions,” according to the company.
The Anthropic Institute plans to conduct research and take actions “to help build the systems that a credible slowdown or pause would require,” including verification mechanisms to ensure bad actors couldn’t secretly advance AI capabilities during a coordinated pause.
The statement follows a Wednesday letter from Anthropic CEO Dario Amodei, Google DeepMind CEO Demis Hassabis, and OpenAI CEO Sam Altman urging Congress to implement safeguards when technology companies order synthetic DNA and RNA, The Wall Street Journal reported.
Risks associated with autonomous AI systems include unintended optimization through destructive means, lack of accountability in complex decision-making, and systemic instability, according to cybersecurity firm Secureleap.
