When the Godfather of AI warns against the rapid evolution of AI, the world should sit up and take notice. Geoffrey Hinton, a pioneer computer scientist, has shared his views on how future AI chatbots need to be trained. In a bold new proposal, Hinton is advocating for advanced AI systems to be designed with “maternal instincts.” He has good reasons to back his suggestions too, hinting at AI’s potential capability to take over its intelligent equivalent.
Speaking at the Ai4 conference in Las Vegas, Hinton dismissed the approach of attempting to maintain dominance over AI through control and suppression. He argued that this strategy is futile, given the rapid advancement of AI. Instead, he presented a new concept, suggesting that the human-AI relationship should be modeled after a mother and her child.
“The right model is the only model we have of a more intelligent thing being controlled by a less intelligent thing, which is a mother being controlled by her baby,” he explained. “If it’s not going to parent me, it’s going to replace me,” he added.
The idea is that just as a baby, with its limited intelligence, can guide and receive care from a vastly more intelligent mother, humanity could be cared for by a superintelligent AI.
Hinton’s warnings in the past
Hinton’s latest suggestion comes after years of continuous warnings about the risks posed by AI. In 2023, he made headlines when he resigned from his long-held position at Google, citing a desire to speak more freely about the potential dangers of the technology he helped create. Hinton’s foundational work on neural networks and deep learning, which earned him the prestigious Turing Award, laid the groundwork for AI. And now, he warns against them.
In an interview with CBS News, he compared the development of AI to raising a “cute tiger cub,” warning that it could one day grow up to be deadly. He has also been open about the timeline, updating his prediction for the emergence of Artificial General Intelligence (AGI from 30-50 years down to a much more “reasonable bet” of five to 20 years. Hinton has also estimated that there is a “10 to 20 percent” chance that AI could displace or even wipe out humanity.
The “maternal instinct” solution
To keep the technology safe for humans, Hinton suggests that humanity needs to focus solely on making AI smarter while giving it a “maternal instinct” to “genuinely care about us, their babies.” He suggests that this is the only area where he believes true international collaboration is possible, as all countries share a vested interest in preventing AI from taking over.
Hinton argues that we need “AI mothers rather than AI assistants.” He has cited the examples by defining examples, stating that “an assistant is someone you can fire. You can’t fire your mother, thankfully.”
