At a New York City Council hearing on a package of AI bills from Speaker Julie Menin, Jacob Coxon testified voluntarily that it is "more likely than not that humanity loses control to these AIs." Daniel Kokotajlo, a former OpenAI researcher and now executive director of the AI Futures Project, testified under subpoena that "our ability to even notice misalignment problems is already quite poor," Fortune reported on October 5.
Fortune's headline quoted Kokotajlo describing safety fixes as potentially "duct tape that will fall off." The hearing was the first full-committee session of all 51 members since 2022. The proposed bills would bar selling or deploying AI in the city without an outside validator's check and a human shutdown capability, with $25,000 fines per violation.
Source: Fortune