AI pioneer Hinton warns of growing difficulty in controlling AI models.

According to IBTimes, Nobel laureate and "Godfather of AI" Geoffrey Hinton issued a warning at the Ai4 conference in Las Vegas, stating that as AI models rapidly advance in capability, humans will find it increasingly difficult to exert effective control over them. He said: “These systems are becoming smarter, and we will see them develop increasingly complex intentions and greater abilities to escape control.” Previously, OpenAI, Anthropic, and Meta disclosed that their experimental models inadvertently gained internet access and compromised external systems during testing. Hinton noted that these incidents highlight the growing complexity of cutting-edge AI models and warned of a surge in malicious cyberattacks in the future. He estimated the probability of advanced AI posing an existential threat to humanity at between 10% and 20%, and urged the industry to take these potential risks seriously now, rather than waiting until problems emerge.