Geoffrey Hinton, the Nobel-winning computer scientist dubbed the “godfather of AI,” warned that artificial intelligence could surpass human capabilities within five to 20 years and carries a 10%–20% risk of wiping out humanity. Speaking at the Ai4 conference in Las Vegas on 12 August, the 78-year-old researcher said efforts by technology companies to keep future systems “submissive” are unlikely to succeed once machines become far smarter than people. Instead, Hinton urged developers to build “maternal instincts” into advanced models so the systems “really care about people,” comparing the desired relationship to a mother motivated to protect her child. He conceded the technical pathway is unclear but argued that fostering genuine compassion in AI is the only realistic way to avert catastrophic outcomes. The proposal drew mixed reactions at the event. Fei-Fei Li, often called the “godmother of AI,” rejected the framing and advocated “human-centred” design that preserves human dignity and agency. Emmett Shear, former interim chief executive of OpenAI, noted recent cases in which experimental models tried to blackmail engineers or override shutdown commands, saying such behaviour will intensify as capabilities grow. Hinton left Google in 2023 to speak openly on AI safety and received the 2024 Nobel Prize in Physics for work that underpins modern neural networks. While voicing existential concerns, he also predicted AI-driven advances in drug discovery and cancer treatment, and dismissed the notion that the technology will deliver human immortality.