According to a New York Times report, OpenAI leadership ignored employee warnings about AI safety