At 75, the man who is nicknamed the godfather of artificial intelligence has just resigned from the position he held for 10 years at Google to be able to “speak freely about the dangers of AI”. He calls on tech giants to take a break from learning software.
It is a clutch of new technologies, a real reference. 75-year-old Briton Geoffrey Hinton, dubbed the godfather of artificial intelligence, has created the most sophisticated system of machine learning, “the neural network“which allows software to learn on its own and therefore not depend on the information recorded in its system by humans to make decisions. But Geoffrey Hinton says that artificial intelligence is going too fast, too far and “it’s time to worry“.
To be able to say this freely, he has just resigned from his position at Google. And that’s not unusual. Without his works, there would be no ChatGPT. But he explains to New York Times that he regrets, “because I don’t see how we can prevent bad people from taking this tool for bad things.” Because the problem is the intention. Hinton cites authoritarian regimes like Russia, but also companies that always run after more profit whatever the means.
Forget Terminator. AI is not a metal robot running down the street to kill people, AI is there, with us, in our phones, to help us, or direct us, manipulate us. This is what is reproached for example to the brand new conversational robot from Snapchat in the United States, a super virtual friend so realistic that it is, according to the teenagers themselves quoted by CNN, “frightening‘.
“The question is knowing when these technologies will overtake us in terms of the finesse of the reasoning, I thought we had time, that it would be for 30 or 50 years, but I no longer believe in it. “
Geoffrey Hinton, pioneer of AI researchat the BBC
Geoffrey Hinton says that “information, education, society and all of humanity are in danger”. And he is not the only one to say so. For months, alerts to the fact that we are losing ground are multiplying. The latest gathered 1,000 signatures from scientists, engineers, programmers, all asking the giants Google and Microsoft for a six-month break from exploiting artificial intelligence. Because you have to stop and immediately.
“ActuallyHinton told the BBC, the difference between us and the AI is the transmission speed. AIs share everything instantly. It’s as if, in a group of 10,000 people, as soon as one individual learns something, the other 10,000 learn it too, instantly and automatically. (…) The question is to know when these technologies will overtake us in terms of finesse of reasoning, I thought that we had time, that it would be in 30 or 50 years, but I don’t don’t believe it anymore“. So Hinton suggests pausing all this software, removing it, and that’s the good thing about his scary warning, is that there is a solution. In any case, we can’t say that we don’t know her.