According to BeInCrypto, David Robinson, who oversaw safety reports for 12 of OpenAI’s frontier launches, has resigned. He cited the company’s trial-and-error approach as a source of growing risk as systems become more capable.
Robinson argued that this culture increases the likelihood of failures that escalate with greater system capability. In an Atlantic essay, he compared AI development to nuclear power plant operations, recommending layered redundancy and careful planning to contain human error.
