Aye Godfather says that future robots should be trained to act like your mother, or they can change us
AI Godfather Jeffrey Hinton warned that AI may soon match human abilities and even come into conflict with his creators. To prevent potential damage, he suggests AI to train with maternal tendency.
In short
- Jeffrey Hinton is urging the developers to design AI who cares about humans like their children.
- He says, like mothers, AI can protect humans
- Hinton warns that AI should be regulated
Artificial intelligence is moving rapidly. And with the fear that it can take it to humans. Researchers and analysts are carrying forward their own principles and suggestions that advanced AIs can move forward in future to deal with potential threats. But if you ask AI’s Godfather Jeffri Hinton, he believes that to be safe from AI, you should not try to become a boss, you should have a child. In other words, Hinton says that everyone needs to treat AI as a mother, as a mother will not harm her children.
Speaking at the AI4 conference in Las Vegas this week, Hinton argued that humans should focus on the creation of artificial intelligence with “maternal instinct”, so that it would still want to save us rather than spending when it becomes much more clever and more powerful. “We have to make it so that they care about us when they are more powerful than us and are more smarter than us,” he said (through business insider).
The Godfather of AI wants we consider AI as a mother for a simple reason. Since humans superintendent are developing AI systems, those systems will definitely develop their own survival and control goals. So instead of trying to dominate them, a strategy that he rejects as “tech brow” fantasy, Hinton argues that developers should design AI to be treated as a mother.
“The correct model is the only model we have a more intelligent thing that is being controlled by less intelligent thing, a mother that is being controlled by her child,” Hintan said. “If it is not going to my parents, it’s going to change me.”
This is not the first time Godfather has raised an alarm on the possible threats of AI. Over the years, they have warned that once machines cross human intelligence, it may be impossible to stop them from advancing their own objectives, even if they struggle with us. And, according to him, the possibility of AI’s conflict with humans is worrying. In the previous interviews, Hinton has estimated that there is a 10 to 20 percent chance that AI may be displaced one day or erase humanity.
And whenever he talks about these dangers, he often frames his warnings in metaphors. Back in April, in a conversation with CBS News, he compared AI to increase “cute tiger cub”. Here he meant that artificial intelligence could now be harmless, but as it grows, it could be fatal. “Until you can ensure that it doesn’t want to kill you when it grows up, you should worry,” he said.
And yes, AI will only be strong with each passing day. Right now, companies like Openai, Google, and Meta are investing huge in making their big language models (LLMS) smarter. And not just clever. These companies are laying the foundation to achieve the smartest form of AI, known as Artificial General Intelligence (AGI). Agi is the level of AI where the models are capable and smart as the most clever humans.
