OpenAI bets on new Superalignment team to stop AI going rogue

OpenAI is building a new Superalignment team to develop methods to train and steer artificial intelligence that surpasses human intelligence. It worries existing training methods will be redundant. Read more…
Neowin