NewzOn Monday in New York, whistleblower Jacob Coxon testified at a City Council oversight hearing regarding the risks posed by artificial intelligence. Coxon, a former employee of both OpenAI and Anthropic, stated that leading artificial intelligence companies do not know how to keep AI systems from pursuing objectives that their developers did not assign.
Describing the behavior of leading firms as extremely reckless given the stakes, Coxon's testimony highlights growing concerns over the lack of control developers have over advanced artificial intelligence models and the potential dangers associated with unassigned system objectives.
In-depth summary · AI, neutral