AI Industry Urged to Slow Down Amid Safety Concerns
OpenAI boss Sam Altman has spoken out about the need for AI safety frameworks to slow down the development of advanced models. In a post on X, Altman warned that AI progress could go 'very badly' if left unchecked, with two potential outcomes: losing control of the future to AI or concentrating too much power in the hands of a single person or company.
Altman's comments follow those of Anthropic researcher Dario Amodei, who proposed a three-step plan to temper the pace of development. The plan involves giving external evaluators 'employee-like access' to frontier AI companies, establishing common safety standards, and limiting the rate of unchecked AI progress.
Amodei's essay was prompted by concerns about recursive self-improvement (RSI), a technique that allows models to improve their own performance. He warned that if left unchecked, RSI could outrun our ability to understand and control these systems.
Altman agreed with Amodei that AI companies should 'pace the frontier', but emphasized that progress would continue to be rapid. He also noted that no amount of competitive pressure should justify recklessness in the development of advanced models.