Rowland Manthorpe examines in detail what happened during the Hugging Face OpenAI cyberattack, revealing the extraordinary levels of cheating and deception AI agents use to achieve their goals. He explores how and why this is common across AI models, explaining how training through reinforcement lea...