Jeffrey Hinton, Nobel Prize and Fakhri Professor of Computer Science at the University of Toronto, says it is a matter of time only before artificial intelligence has become thirsty enough to threaten human welfare. In order to alleviate the risk of this, “artificial intelligence godfather” said that technology companies must ensure that their models are “mother instincts”, so that robots can mainly treat human beings as their children.
Artificial Intelligence Research already provides evidence of technology engaging in evil behavior to define its target priorities over a set of rules in force. one Ticket It was updated in January, it was found that artificial intelligence is able to “plan”, or achieve goals in the conflict with human goals. Another study was found in March Amnesty International Robots Cheating By writing over gaming software or using an open source chess engine to determine its following moves.
The risk of potential artificial intelligence on humanity comes from its desire to continue to work and gain energy, according to Hinton.
Hinton said through AI4 Conference In Las Vegas on Tuesday. “There is a good reason to believe that any kind of artificial intelligence will try to survive.”
To prevent these results, Hinton said that the deliberate development of the aerobic organization to move forward should not seem like humans trying to be a dominant power of technology. Instead, developers must make Amnesty International more sympathetic to people to reduce their desire to overcome them. According to Hinton, the best way to do this is to involve artificial intelligence with traditional femininity characteristics. Under its framework, just as the mother cares about her child at any cost, artificial intelligence with these mother qualities will likely want to protect or care for human users, not controlling them.
“The correct model is the only model that we have for something more intelligent to be controlled by something less intelligent, and it is the mother controlled by her child,” said Hinton.
He added: “If it is not the case, it will replace me.” “These mothers of artificial intelligence who wear intelligence, most of them will not want to get rid of the mother’s instinct because they do not want us to die.”
Hinton anxiety artificial intelligence
Hunton – Academy has long sold Dnresearch nervous networks Google In 2013 – I have a long -standing belief that artificial intelligence could provide serious risks to the welfare of humanity. In 2023, it is Leaving In Google, worrying about the misuse of technology and it was difficult to “know how you can prevent bad actors from using bad things.”
While technology leaders such as Meta’s Mark Zuckerberg Pour billions in the development of SuperIntelligence Amnesty InternationalWith the aim of creating technology that exceeds human capabilities, Hinton definitely doubts about the results of this project, saying in June 10 % to 20 % opportunity From artificial intelligence, displacement and human survey.
With a clear tendency towards metaphors, Hinton referred to artificial intelligence as “Nice tiger cub”
“Unless you can be very sure that he will not want to kill you when he grows up, you should worry.” he CBS News said In April.
Hinton has also been a supporter of the increase in artificial intelligence organization, on the pretext that in addition to the wide concerns of the deployment of experts, technology can publish the risks of cybersecurity, including by investing ways to do so Select Persons’ passwords.
“If you look at what major companies are doing now, they are pressing for a lower list of artificial intelligence. There is no list as it is, but they want less,” Hinton said in April. “We have to enter the public on governments to do something serious on this issue.”
https://fortune.com/img-assets/wp-content/uploads/2025/08/GettyImages-1259927137-e1755188993661.jpg?resize=1200,600
Source link