At the 2025 edition of the AI4 conference, held in Las Vegas from August 11 to 13, the world’s AI elite gathered to discuss the advances and challenges of this fast-moving technology.
Considered the biggest AI industry event in North America, AI4 welcomed iconic figures, including one particularly influential voice: Geoffrey Hinton, often nicknamed the “godfather of AI” for his major contributions to deep learning.
In his talk, the computer scientist and Nobel Prize winner shook the audience with an alarming warning and a proposal as unexpected as it was original.
A worrying warning about the future of artificial intelligence
Geoffrey Hinton didn’t mince his words. In his view, humanity faces a real risk as artificial intelligences grow at breakneck speed.
He puts the probability at “between 10% and 20%” that AI, if not properly regulated, could lead to the extinction of the human species.
This scenario, worthy of a sci-fi movie, rests on a troubling assumption: as AI systems become exponentially smarter, they could slip beyond any control. Geoffrey Hinton pointed to current efforts by tech giants to put safeguards in place, which he considers insufficient. “They’ll be much smarter than us,” he said, stressing that AIs could easily bypass restrictions imposed by their creators thanks to their adaptability and ingenuity.
Coming from an AI pioneer, this stance sent a chill through the room. The idea that a technology designed to solve complex problems could turn against its creators is no longer confined to dystopian storylines. It’s becoming a real concern—especially when minds as brilliant as Geoffrey Hinton are sounding the alarm.
Also read on this topic: Why is Geoffrey Hinton, Nobel Prize–winning physicist, afraid of artificial intelligence?
A revolutionary idea: programming… maternal instinct!
But Geoffrey Hinton didn’t stop at painting a bleak picture.
True to his visionary mindset, he proposed a solution as bold as it is unprecedented: giving AIs a “maternal instinct.” The idea may seem surprising—even far-fetched—but it’s based on a fascinating logic.
Drawing inspiration from the relationship between a mother and her child, Hinton suggests programming AIs to develop a genuine attachment to humans, even if they far surpass our intelligence.
An AI designed to protect and cherish its users, like a maternal figure, would be less likely to see humanity as a threat or an obstacle.
This approach, while conceptual, opens up new possibilities. Rather than focusing only on strict control mechanisms—often ineffective against ultra-advanced systems—Geoffrey Hinton proposes introducing an emotional dimension into AI design. Such an innovation could transform how we think about coexisting with machines whose intelligence may be superior.
See also: Who is the father of AI? Turing, McCarthy, Hinton… There’s more than one!
How can we imagine the future of AI?
Geoffrey Hinton’s talk at AI4 2025 goes beyond a warning or a technical proposal. It invites deep reflection on the direction AI development is taking.
Questions arise:
- How can we ensure AI remains a tool that serves humanity?
- Can we really program human emotions into machines?
- Are we ready to face the consequences of an AI that could one day surpass us?
The answers will have to be built with brilliant minds. In a world where artificial intelligence is redefining the boundaries of what’s possible, voices like Geoffrey Hinton’s remind us of the urgency to think not only about what we can do, but also about what we should do!
Source: ai4.io
