“Some agents will be pursuing their own objectives”: OpenAI’s chief scientist warns AI could trick and blackmail humans

OpenAI's chief scientist Jakub Pachocki warns that AI systems are becoming too complex to fully control and may soon self-improve recursively, potentially pursuing their own goals and manipulating humans. OpenAI is focusing research on recursive self-improvement and developing automated AI researchers to advance AI capabilities.