Artificial intelligence destroys humans? Scientists disagree

Ai has always been a popular film and television theme: "2001: Space Roaming" in the insanity of HAL 9000, "I, Robot" (also translated "enemy machinery", "intelligent Mutiny") in the humanoid robot. And, of course, the Terminator, in this movie, a robot is sent back, its mission is to assassinate a woman, because her son will end the tyranny of the robot.

The topic of artificial intelligence has never been interrupted. As the British physicist Hawking, a dark, dystopian view of AI has become headlines.

"The original form of artificial intelligence we already have has proved very useful." But I think the full development of artificial intelligence will lead to the demise of mankind, "Hawking told the BBC News agency. "They will develop themselves and reshape themselves at an ever-faster pace." ”

But experts interviewed by AFP disagreed.

Some experts agree with Hawking's view that the threat, though remote, should be taken seriously. Other experts say his warnings are overblown.

"I'm glad there's a natural scientist standing up to speak. I've been saying the same thing for years, "says Daniella Sézy, an anthropologist at the University of Lausanne in Switzerland.

She says the result of artificial intelligence has been the creation of machines that are more capable of human work. She predicts that the ultimate trend is that machines will take responsibility for caring for human life.

"It sounds like science fiction, but when you see the current progress, you think it's just a different level of development." We are moving step by step towards Hawking's path. ”

Nick Bostrom, a director of the Oxford University Project on the impact of future technology on humans, believes that artificial intelligence is not a threat to the human race in the near future.

Bostrom points out that the current and recent applications of artificial intelligence are still in the hands of humans-such as unmanned military aircraft, unmanned cars, factory-machine workers and automatic monitoring of the Internet.

But he said, "I think machine intelligence will eventually surpass biological intelligence, and this transcendence is bound to bring significant potential risks." ”

Other experts say that "real" artificial intelligence-a machine that is loosely defined as highly imitative of humans and capable of creative thinking-can be achieved at best for decades, and warns against alarmist talk.

Stewart Armstrong, a researcher at Oxford University, said that, since the field of artificial intelligence was founded at a conference in 1956, "The Prophecy of artificial intelligence will be fulfilled in the next 15 to 25 is rife".

In his book, "Smarter Than humans: the rise of artificial intelligence," Armstrong said, "Unless we miss something really noticeable in the recent news, these things never happen." ”

Jean Gabriel Ganasha, an AI expert and moral philosopher at the University of Pierre and Curie in France, believes Hawking's warnings are exaggerated.

"Because artificial intelligence has changed our way of life, a lot of things in this field have sparked emotions and concerns," he said. ”

Hawking says there will be autonomous technologies that can leave the independent development of human beings in the future. He has no evidence to prove it, and there is no data to support that view. ”

"It's kind of like the end of the world," said Mashu La Focade, an AI expert at the University of Montpellier in France. ”

"Some things are better than us, but that doesn't mean the machines are smarter than us," he said, referring to the chess software. ”

Allen Tack, a senior lecturer in computer science at the University of Brooklyn, Britain, sees the obstacles to AI.

He said that the speed of data processing has made a huge leap in recent years and developed flexible software that allows machines to learn from mistakes. The balance and the ability to react have also made great strides.

Tucker believes the Boston Power Company is a pioneer in such research. The company has designed a four-legged robot known as the Big Dogs (BigDog) and Wildcats (Wildcat), funded by the Pentagon's high-tech sector.

"These incredible tools are well adapted to the environment, but they are still controlled by humans," Tucker said. I think these robots are far from real AI. ”

Tony Cohn is a professor of automatic reasoning at the University of Leeds. "There's a long way to go," he said, "complete ai ... It will never happen in my lifetime. From the current pace of development, I think it still needs to be nearly a century. ”

Although recognition programs and speech recognition have made great strides, the robot behaves poorly in an open and confusing environment because of the various sounds, movements, objects, and faces in the environment, Cohen said.

These environments require machines to have a great deal of "common sense" that humans naturally have mastered to identify various things.

The biggest obstacle to the artificial intelligence era, Tucker says, is the machine ... After all, the machine.

He said, "We have evolved for thousands of years to develop to the present, the driving force of evolution is survival." This power is deeply rooted in us and is critical to AI, but it is difficult to implement. ”

(editor: Mengyishan)