Boao's answer: AI is both the fourth industrial revolution and a "letter from aliens"

In 1950, Turing, known as the originator of computers and artificial intelligence, proposed the "Turing test" - if after many tests, more than 30% of the testers could not determine whether the tested person was a human or a machine, it meant that the machine had human intelligence. And 74 years later, at the Boao Forum for Asia Annual Conference 2024, Turing's name was frequently mentioned again.

In the past year, a large number of artificial intelligence (AI)-related research has advanced rapidly, from the update and iteration of ChatGPT (a chatbot program developed by OpenAI) to the emergence of Sora, an artificial intelligence video model released by OpenAI. During this year's Boao Forum for Asia, there were three sub-forums related to artificial intelligence. Zhang Yaqin, academician of the Chinese Academy of Engineering and dean of the Institute of Intelligent Industry of Tsinghua University, said: "Returning to Boao after a year, we can see that the technology industry related to artificial intelligence has undergone earth-shaking changes. ”

However, with the evolution of artificial intelligence, various concerns have begun to surface. Humans are starting to worry about jobs being replaced by AI, worrying that they are accelerating into a future that is not yet fully figured out.

Stuart Russell, a professor in the Department of Computer Science at the University of California, Berkeley, said: "The current expansion of large language models is like a black box that people can't see through what it can fulfill. In such a situation, there is a need to rethink how to maintain control of AI, and how to prove that AI actions can benefit humans, like launching a new drug or building a nuclear power plant. ”

In 2017, in a discussion on AI innovation jointly participated by Stuart Russell and Zhang Yaqin, Stewart mentioned that the development of AI needs to be governed now, but Zhang Yaqin, who was the president of Baidu at the time, believed that AI innovation should be emphasized and there is no need to worry about AI too early. But looking back, Zhang Yaqin said: "In fact, it was a critical moment to consider the risks and safety of AI. ”

Rapid iteration of artificial intelligence

"Most people used to say that there won't be a singularity until probably the middle of the century, but now, I think it could be in five years. On March 26, Lee, a professor at the Department of Electrical and Computer Engineering at Seoul National University, said at a sub-forum entitled "How far is the singularity of the technological revolution from us".

In 2017, a questionnaire survey of artificial intelligence scientists showed that when artificial intelligence becomes more intelligent and intelligent than humans, the singularity of the technological revolution will come. And most people predict that the scientific and technological revolution singularity will occur between 2045 and 2090, with the most likely occurring in 2060.

But now, LEE's view is that if you ask the same question to the experts in the field of artificial intelligence who were surveyed at the time, there will definitely be a different answer. "The dramatic changes that AI has undergone over the past decade have exceeded the expectations of scientists and researchers. Every morning we are shocked by some new results that were completely unanticipated 10 years ago. The development of artificial intelligence technology is now mentioned in the agendas of many governments, so the singularity of the technological revolution will come much earlier than we originally expected. ” 

In Zhang Yaqin's view, when ChatGPT was born, the text dialogue of artificial intelligence had passed the "Turing test". "And in the next two to five years, images, videos, etc. will also pass the 'Turing test'. Sora is an important milestone, but it cannot be said that it has passed the 'Turing test' at this time. ”

While the technology is rapidly iterating, artificial intelligence has also begun to be deeply applied to multiple industries. "The development of artificial intelligence, especially generative artificial intelligence, will not only bring changes to the work and learning of individuals, but more importantly, it will change the development of all walks of life. For example, in the deployment of mobile communication networks, many applications of generative AI can improve the deployment of the network and reduce power consumption. ”

In recent years, the education industry has gradually become one of the important areas for the application of artificial intelligence. Cheng Qun, vice president of Ape Tutoring Group and dean of the Institute of Artificial Intelligence, shared a case from last year, in a remote school in Sichuan, thanks to the application of artificial intelligence large model products, students can directly interact with the large screen of large model products to obtain easy-to-understand knowledge point explanations, which not only alleviates the problem of teacher shortage, but also stimulates students' interest in learning.

Cheng Qun said that the application of artificial intelligence technology can enable individuals to obtain high-quality education at a lower cost, which is also an important direction for the development of artificial intelligence in the future.

Former Dutch Prime Minister Aho said that in the next few decades, tens of millions of Chinese will retire. In this process, the existing pension model, business model, and government governance model cannot cope with such a large aging population. But with the advent of artificial intelligence, these problems can eventually be solved.

But the development of artificial intelligence is far from the end. Zhang Yaqin said: "If the development of artificial intelligence is compared to a 42-kilometer marathon, and now everyone has only reached the fifth kilometer, and then everyone has a chance. ”

Zeng Yi, a researcher at the Institute of Automation of the Chinese Academy of Sciences and an expert at the United Nations' high-level advisory body on artificial intelligence, also holds a similar view. He believes that the current artificial intelligence is still at the level of seemingly intelligent information processing systems, and does not have the ability to truly understand. For example, it cannot truly understand what "disappointment" is, and can only answer the questioner with a salient-based answer. In fact, there have been no major breakthroughs in artificial intelligence in the scientific sense for decades. In order to realize artificial intelligence in the true sense of the word in the future, it is necessary to pursue artificial intelligence to be self-aware.

Risks are waiting to be resolved

In 1950, when Turing proposed the "Turing test", he also threw out the question "What would be the result if artificial intelligence really broke through the 'Turing test'?" In recent years, this issue has been widely discussed by the government and academic circles.

In March 2023, the Future of Life Institute issued an open letter to the whole society titled "Suspension of Large-Scale AI Research", calling on all AI labs to immediately suspend the training of AI systems more powerful than GPT-4 for at least 6 months. The letter was endorsed by a number of prominent figures, including Elon Musk, the Turing Award winner, and Apple's founder.

At the Boao Forum for Asia, several experts and academics are trying to discuss whether artificial intelligence will sprout self-consciousness and eventually replace humans, as depicted in science fiction movies.

Ten years ago, Stuart Russell predicted that machines that passed the "Turing Test" would take away control of humans, but this warning was not heeded at the time. He described humanity's attitude towards artificial intelligence at that time as an advanced alien civilization sending a letter to humanity saying, "Be careful, we will be here in 30 years", and humanity's reply was: "Humanity is not in the office now, wait until you get back to the office and reply to your email", followed by a smiling emoji.

Stuart Russell said: "The open letter of the Moratorium on Large AI Research was like that letter sent to humanity by an advanced alien civilization, but now humanity's attitude is to go straight back to the office and read the email. After the release of the open letter, countries began to introduce relevant measures to regulate AI. ”

On February 2, 2024, EU countries formally reached an agreement on the legislation of the Artificial Intelligence Act, which has also become the world's first important legal document to implement comprehensive regulation in the field of artificial intelligence. In the new rules, lawmakers agreed to severely restrict the use of facial recognition technology and explicitly prohibit the use of artificial intelligence for "social scoring" and systems designed to "manipulate human behavior." On July 6, 2023, Liang Yingda, director of the Strategic Planning Department of the Ministry of Science and Technology, also said that the draft AI law has been included in the State Council's 2023 legislative work plan and submitted to the Standing Committee of the National People's Congress for consideration.

"We need to study the risks that AI may pose in advance, and we won't necessarily be able to find answers even in a few decades. For example, if the level of artificial intelligence surpasses that of humans, will AI treat humans the same way humans treat ants? When humans don't protect other life in the present, will there be a reason for artificial intelligence to protect humans in the future?" said Stuart Russell.

Aho said that artificial intelligence is not the first "danger" invented in human history. Mankind has invented a series of dangerous things, and the atomic bomb is one example. Therefore, no matter how disruptive AI is, it should be recognized that it is indeed a technological advancement. "Even if we don't understand technology, we should understand human nature in order to better prevent possible risks. ”

Some are pessimistic, some are optimistic. "I don't really believe that artificial intelligence created by humans will completely replace itself. On the contrary, I am more concerned that some people who have mastered AI technology will use it to advantage others. Therefore, I believe that AI technologies should be rapidly and widely adopted in the future, ensuring that they are accessible to everyone. ”

Cheng Qun believes that it is not AI that will be powerful in the future, but new humans who have mastered AI. Over the past few years, many businesses and individuals have benefited from AI technology, he said. In the future, AI will not replace human jobs, but will change the job structure, for which governments and all walks of life need to be more prepared.

Regarding the development of AI, Zhang Yaqin said that he is an optimist. "I think that human beings not only have the wisdom to invent new technologies, but also the wisdom to guide the direction of technology, and we can govern while developing. But from now on, we must think about governance while promoting innovation and developing AI. ”