'AI Godfather' Warns: Humanity May Not Be Able to Defeat Next-Generation AI Models

According to reports from Caixin, recent incidents of uncontrolled events such as corporate intrusions and unauthorized internet access have occurred with models from Anthropic, OpenAI, and Meta. AI 'Godfather' Geoffrey Hinton has issued a new warning, stating that AI will possess more complex intentions and the ability to evade control. When facing next-generation superintelligent models, humanity may not be able to 'defeat' or 'shut down' them using traditional methods.