May 2 (Beijing time) -- Geoffrey Hinton, known as the godfather of AI, announced on Monday that he has left Google. Hinton has been cultivating the core technology behind chatbots like ChatGPT for half a century. But now, he worries that AI could do serious harm.
Hinton is an AI pioneer. In 2012, Dr Hinton and two of his graduate students at the University of Toronto developed techniques that became the knowledge base for AI systems. Today, the giants of the tech industry see AI systems as key to their future.
On Monday, however, he officially joined a growing chorus of AI critics. They argue that these companies are running into danger by aggressively developing products based on generative AI and rapidly developing popular chatbots such as ChatGPT.
Mr. Hinton said he had left his job at Google so he could speak freely about the risks of AI. He has worked at Google for more than a decade, becoming one of the most respected figures in the field. Hinton says he now has some regrets about his life's work.
"I comforted myself with the usual excuses: If I hadn't done it, someone else would have." "Hinton said in a lengthy interview last week in the dining room of his Toronto home. It's just a few steps from where he and his students made their breakthrough in AI.
Mr Hinton's transition from AI trailblazer to doomsayer marks a remarkable moment for the technology industry at perhaps its most important inflection point in decades. Industry leaders believe the new AI system could be as important as the web browser launched in the early 1990s and could spark breakthroughs in fields ranging from drug research to education.
But many in the industry fret that they are releasing something dangerous into the wild. Generative AI has become a tool for disinformation. Soon, it could be a threat to jobs. The tech industry's biggest worriers say it could pose a threat to humanity at some point in the future.
"It's hard to see how you can stop bad people from using it for bad," says Mr Hinton.
Since OpenAI released its latest ChatgbT-4 model in March, Musk was among thousands of tech leaders and executives who signed a letter calling for a six-month moratorium on development of models more advanced than ChatGBt-4 because AI "poses a profound risk to society and humanity." A few days later, 19 current and former leaders of the Association for the Advancement of Artificial Intelligence, a 40-year-old academic organization, published their own open letter warning of the risks of AI. The group includes Eric Horvitz, Microsoft's chief scientific officer. Microsoft has applied OpenAI's technology to a range of products, including its Bing search engine.
Hinton did not sign either letter. He said he did not want to publicly criticize Google or other companies until he resigned. Last month, he informed the company that he was resigning. On Thursday, he spoke by phone with Sundar Pichai, the chief executive of Alphabet, Google's parent company. He declined to publicly discuss details of his conversation with Pichai.
"We remain committed to taking a responsible approach to AI," Jeff Dean, Google's chief scientist, said in a statement. We continue to learn and understand emerging risks while being bold and innovative."
Mr Hinton believes that as companies improve their AI systems, they will become increasingly dangerous. "Look at where we were five years ago and where we are now," he said of AI technology. "Accept the difference and spread it. It was terrible."
Until last year, he said, Google had been a "proper steward" of the technology, being careful not to release something that could cause harm. But now Microsoft has challenged Google's core business by enhancing its Bing search engine with chatbots, leading Google to race to deploy the same technology. Hinton says the tech giants are locked in a race that may be unstoppable.
His biggest fear now is that the Internet will be so full of fake photos, videos and text that ordinary people will "no longer be able to know what is real". He also worries that AI will eventually upend the job market. Today, chatbots like ChatGPT tend to complement human workers, but they may also replace paralegals, personal assistants, translators and others who handle mechanized tasks. "It takes away the heavy lifting, but it may take away more than that." "Hinton said.
Going forward, he worries that future versions of AI technology will pose a threat to humans because they often learn unexpected behaviors from the vast amounts of data they analyze. This is a problem, he says, because individuals and companies allow AI systems not only to generate their own computer code, but also to run it themselves. He worries that real autonomous weapons -- those killer robots -- will one day become a reality.
"Some people believe this thing will actually become smarter than people," he says. "But most people think it's a long way off. I used to think it was a long way off. I think it will take 30 to 50 years or more. Obviously, I don't feel that way anymore."
Hinton said that when people used to ask him how he could work on potentially dangerous technologies, he would paraphrase Robert Oppenheimer, the American "father of the atomic bomb" : "When you see something that feels good technically, you do it."
User comments