OpenAI doubles down on decision to fire three AI safety researchers
OpenAI is standing firm on its decision to fire three safety researchers after an investigation found they committed “a significant breach of trust.”
Read the full story on The Verge.