Anthropic researcher quits over fears of ‘out of control’ AI systems

A researcher at artificial intelligence company Anthropic has quit, claiming that he is concerned that the firm and the industry are racing to build AI systems that he believes they will not be able to control.

In a series of social media posts on Wednesday, Jacob Coxon said that he had done pre-training research at United States-based firms Anthropic and OpenAI for the past three years, and that “neither company is acting responsibly”. The firms were “racing straight to self-improving superintelligence and gambling with our lives”, he said.

In AI, “pre-training” refers to the initial phase where a machine learning model learns patterns, grammar, facts and structures from a dataset before it can be trained for specific tasks.