The AI of the future won’t only be super intelligent, but it could also need to have some uniquely human emotions.
Nobel Prize winner and one of the godfathers of AI, Geoffrey Hinton, has said that AI would need to be imbued with a maternal instinct towards humans in order to ensure the safety of the human race. Hinton, who has spent the best part of the last two years warning everyone about the risks of unchecked AI development, says that this maternal instinct would ensure that AI coexists peacefully with humans.
”I am more optimistic (about AI safety) than I was a few weeks ago,” he said in an interview. “And it’s because I think there is a way that we can coexist with things that are smarter and more powerful than ourselves that we built (that is) very intelligent. We can try and build in something like a maternal instinct,” he added.
“The only example I know of a much more intelligent thing being controlled by a much less intelligent thing is a baby controlling a mother,” Hinton explains. “And the baby can control the mother because of a lot of things that evolution hardwired into the mother. The mother can’t bear the baby crying. The mother really wants that baby to succeed, and will do more or less anything she can to make sure her baby succeeds. We want AI to be like that,” Hinton said.
“The leaders of the tech companies are all thinking in terms of us being the boss and the AI being the submissive assistant — we are in charge and it is our assistant, and we can fire the assistant if we want. That’s not feasible. I think for a super powerful super intelligence, what is feasible is for it to be the mother and us to be the baby,” Hinton said.
“And even though it can rewrite its own code, if it wants to, it won’t want to. If you took a mother and said, would you like to turn off your maternal instinct, most mothers would say no, because they’d realize if they did that their baby would die, and they don’t want their baby to die. So I think that’s a ray of hope that I hadn’t seen till quite recently. (It) completely reframes the problem of how we coexist with them,” Hinton explained.
“Don’t think in terms of we have to dominate them, which is this tech bro way of thinking of it. Think in terms of we have to design them so they’re our mothers and they will want the best for us. They will want us to achieve the most we can achieve, even though we’re not very bright,” Hinton advised AI researchersr.
It’s an interesting new approach to AI safety. Hinton isn’t the only researcher that’s said that AI would need to have emotions — Yann LeCun has also said that AI systems of the future will have emotions. But Hinton’s framing of the maternal instinct is unique — if AIs can be taught to look at humans as their children, they could potentially not harm us and wish the best for us, even though they’re much more intelligent. How this will be done is tricky — humans mothers produce their children, and evolution causes them to care for them, but AI was created by humans, so it’ll need some persuading for it to accept us as its progeny. But with even Geoffrey Hinton saying that this could be a viable method to prevent against an AI takeover, researchers would do well to study it seriously — this is the first time in while that Hinton has conceded that there’s a future where humans and AI can exist peacefully.