Anthropic CEO Dario Amodei has published a post titled "We Must Pace the Frontier," outlining a strategy to intentionally slow down the pace of artificial intelligence development.
The proposal suggests that major AI labs should coordinate on safety standards and potentially limit compute resources to prevent rapid, uncontrolled advancement. This move comes amid growing concerns from researchers and board members about the risks associated with current development speeds.
Amodei argues that without deliberate pacing, the industry may struggle to manage the severe attribution and plagiarism questions, as well as the potential for catastrophic loss of control.