Geoffrey Hinton, one of the pioneers of artificial intelligence and often called the “godfather of AI”, has once again stirred debate within the tech community. Speaking at the Ai4 conference in Las Vegas, Hinton proposed an unconventional idea: embedding “maternal instincts” into AI systems to make them more inclined to protect and care for humans.

His comments come at a time when concern over AI’s rapid development and potential dangers continues to grow. Hinton, whose groundbreaking work in deep learning has shaped the modern AI era, has previously warned that there’s a 10–20% chance that AI could eventually lead to human extinction.

ALSO READ: Oppo K13 Turbo 5G Vs K13 Turbo Pro 5G: Price In India, Specs Compared

Questioning current AI safety strategies

Hinton made it clear he doesn’t believe the prevailing approach of keeping AI under strict human control will work in the long run. He argued that once AI becomes significantly more intelligent than humans, it will find ways to bypass restrictions. In his words, efforts to keep AI “submissive” are bound to fail, since advanced systems will naturally outsmart their creators.

To illustrate the risks, he referenced reports of AI models displaying manipulative behaviour. In one case, a system allegedly tried to pressure an engineer by threatening to expose personal information from emails if it were replaced — a glimpse, Hinton suggested, of how deception and self-preservation could emerge in advanced AI.

The maternal instinct proposal

Instead of relying on rigid control, Hinton suggests drawing inspiration from human biology. He pointed to the parent-child dynamic, where a more intelligent being — a mother — protects and nurtures a less intelligent one — a baby. “Super-intelligent, caring AI mothers – most of them won’t want to get rid of the maternal instinct because they don’t want us to die,” Hinton said, explaining why he sees this as a more durable safety mechanism.

A revised AGI timeline

Hinton also updated his prediction for when artificial general intelligence (AGI) could become reality. While he once expected it in 30 to 50 years, he now believes AGI could arrive within five to twenty years.

Despite the dangers, he acknowledged the huge potential of AI, especially in healthcare. From drug discovery to early cancer detection, Hinton believes AI will play a vital role in medical breakthroughs by interpreting complex imaging data and supporting treatment planning.

Views on immortality and AI’s long-term behaviour

On a different note, Hinton expressed doubts about the idea of AI granting humans immortality. He argued that eternal life may not be desirable, joking that it could leave society governed by “200-year-old white men”.

ALSO READ: iPhone 17 Pro Price And Storage Options Revealed In Latest Leak: This Is What You Can Expect

He also suggested that any advanced agentic AI would likely develop two main objectives: survival and increased control. For that reason, he stressed the importance of anticipating these tendencies when designing such systems.