Scientist Geoffrey Hinton, one of the pioneers in the field of artificial intelligence, made statements that warned against progress in this field.
Hinton, 75, recently resigned from his position at Google.
“Now I can talk about what the dangers are,” Hinton said in an interview with the BBC, adding that some of these dangers are “very frightening.”
The scientist, who has British and Canadian citizenship, says that chatbots, known as chatbots, may soon exceed the information capacity of the human brain.
Hinton played a role in reaching the present point of artificial intelligence with his research on deep learning and artificial neural networks.
Advanced systems like ChatGPT are at the center of warnings about the future as well as the possibilities they provide.
Hinton also draws attention to this point:
“Right now, GPT-4 can far outshine a human in the amount of general knowledge it holds. In terms of reasoning, they are not equally good, but they can make simple reasoning.
Considering the rate of progress, we know that these systems will become more effective very quickly. And that should worry us. At the moment, as far as I know, they are not smarter than us. But they may soon be smarter.”
In the article he wrote for the New York Times, Hinton emphasizes that “bad people” can do “bad things” with artificial intelligence.
Answering the BBC’s question about this point, Hinton replies, “This is one of those nightmare scenarios”:
“Suppose one of the villains, Putin, allows robots to create their own sub-purposes. This may eventually lead to sub-objectives such as ‘I need to get more power’.
Hinton continues:
“I came to the conclusion that the type of intelligence we create is very different from what we have.
“We are a biological system, but these are digital systems. The biggest difference with digital systems is that you can make multiple copies of the same size.
“All these copies can learn different things but instantly share their knowledge with each other. So it’s like you have 10 thousand people and one of them learns something and the others have this information at the same time. That’s why these chatbots have more information than any human.
Regarding his departure from Google, Geoffrey Hinton said, “I wanted to say good things about them. “If I had said these things while I was working there, they wouldn’t have been effective enough,” he said.