Renowned American entrepreneur Elon Musk stated on social media on the 8th that by the end of next year or 2026, new artificial intelligence (AI) models may surpass human intelligence, "smarter than the smartest people.".

Is Musk's prediction reliable? Can humans benefit from the rapid development of artificial intelligence? At the same time, artificial intelligence will also cause many problems in its development process. How to deal with the risks involved?

Is Musk's prediction reliable?

Musk's interview with Nikolai Tangen, CEO of Norwegian Bank Investment Management Fund, was released on social media platform X on the same day. Musk said in an interview, "I speculate that by the end of next year, we will have artificial intelligence that is smarter than any other human."

This prospect is limited by related bottlenecks. The Financial Times reported that the development speed of artificial intelligence has been affected by the bottleneck period of microchip supply, especially the microchips produced by Nvidia in the United States, which are crucial for training and running artificial intelligence models. Musk stated that while these limitations are being alleviated, the new model is testing the equipment and grid performance of many data institutions.

Zhu Rongsheng, a special expert at the Center for Strategic and Security Studies at Tsinghua University, told reporters that the power consumed by the human brain is much lower than the power lost during the operation of artificial intelligence, and there is still a gap between artificial intelligence and human intelligence. However, in the long run, compared to the human brain, there are many shortcomings in artificial intelligence that may eventually surpass human intelligence under the continuous support of algorithms and data. The hardware and software required to achieve this goal are chips and algorithms, respectively.

Last year, Musk founded the artificial intelligence startup xAI, which launched its first artificial intelligence model Grok in November of the same year. Musk said that the training of the second-generation model Grok 2 will be completed in May this year, requiring approximately 20000 Nvidia graphics processors (GPUs) and computing chips H100. The training process was temporarily hindered due to chip shortages. Following Grok 2, xAI will launch a more powerful third-generation model that is expected to require over 100000 chips. He said that the main constraint on the development of artificial intelligence last year was the lack of high-performance chips. This year, supply constraints are shifting towards voltage transformers; In another year or two, the restriction will be on electricity supply.

International cooperation is needed to reap the dividends

In the past 18 months, breakthroughs in artificial intelligence, including video generation tools and chatbots, have greatly accelerated the development of artificial intelligence.

The Financial Times reported that Musk predicted last year that humans would "fully" achieve universal artificial intelligence by 2029. He has always been optimistic about the development of so-called universal artificial intelligence, believing that this artificial intelligence tool is very powerful to defeat the most capable human individuals in any field. Demis Hassabis, co-founder of Google's artificial intelligence company Deep Thinking, predicted earlier this year that universal artificial intelligence could be achieved by 2030.

Zhu Rongsheng stated that the development of artificial intelligence first requires policy, resource, and technological support, and many fields of artificial intelligence lack relevant support for "cultivation"; Secondly, cooperation between countries is necessary, but currently, many international collaborations are constrained by various factors such as geopolitics, which have hindered the development of artificial intelligence; Thirdly, the development of artificial intelligence is constrained by cultural characteristics, such as the answer provided by ChatGPT, which is more in line with the English context. In this way, developing artificial intelligence with a single national power may increase the imbalance of global artificial intelligence development, which is the data divide. The solution to the above problems requires precise connection between funds and the market, and extensive international cooperation.

Former Deputy Secretary General of the United Nations and former Executive Director of the United Nations Environment Programme, Eric Solheim, recently stated in an interview with Xinhua News Agency that the world needs to engage in global cooperation on artificial intelligence, which requires close cooperation among major countries such as China, the United States, and European countries to formulate reasonable regulations, so as to seize the opportunities brought by new technologies and benefit from them.

How to deal with risks?

Artificial intelligence has reduced production costs, improved production efficiency, and opened the door to new formats in multiple industries. Although efficiency has been improved, the use of artificial intelligence has caused many problems and risks.

According to a survey conducted by Nature in 2023, 30% of respondents admitted to using artificial intelligence tools to assist in writing manuscripts. The website of Popular Science in the United States pointed out that due to many researchers relying on artificial intelligence tools, many professional journals are filled with "nonsense" generated by artificial intelligence tools, and many articles have obvious traces of using artificial intelligence.

As early as 2014, more than 120 articles were deleted from the journals of the American Society of Electrical and Electronics Engineers and Germany's Springer Publishing House, as these articles contained "nonsense" generated by artificial intelligence. According to a review by Popular Science, in the next 10 years, as more complex and easy-to-use tools such as ChatGPT from OpenAI become widely used, the amount of text generated by artificial intelligence in various journals will almost certainly increase.

Zhu Rongsheng believes that the social problems caused by artificial intelligence are spreading rapidly and have a wide range. This technology continues to intensify the complexity of legal issues related to copyright law, copyright, privacy, etc. Therefore, moral boundaries and legal norms applicable to new technologies are about to emerge.

"The era of artificial intelligence is an era where risks and opportunities coexist. The uncertainty that this technology brings to human society and destiny may far outweigh certainty," said Zhu Rongsheng. "Faced with this unstoppable trend, we need to maintain a calm and cautious mindset and thinking in the tide of artificial intelligence."