Always higher. In recent years, China has made significant progress in artificial intelligence. So much so that it is now the first nation in the world in terms of patent applications in generative AI. This was revealed by the UN on Wednesday in a study that explains that the number of Chinese patents in this area has increased eightfold since 2017.
The acceleration is truly phenomenal: a quarter of the 54,000 patents filed over the past decade for innovations in generative AI were filed by China last year, indicatedWorld Intellectual Property Organization (WIPO), in its report.
Thus, between 2014 and 2023, more than 38,000 GenAI (generative AI) innovations came from China, out of a total of 54,000. This is six times more than the United States, which is in second place with 6,276. South Korea comes in third with 4,155, followed by Japan with 3,409. India, where 1,350 GenAI patents have been filed, shows the strongest growth.
Chinese companies and entities at the top of the ranking
Not surprisingly, it is Chinese entities and companies that tops the list of generative AI patent applications. Like Tencent, at the top of the list, followed by Ping An Insurance, Baidu and the Chinese Academy of Sciences.
On the American side, the IT giant IBM only appears in fifth place, followed by China’s Alibaba (6th), Samsung Electronics from South Korea is in seventh place and Google’s parent company, Alphabet, in eighth place.
Also note: the Chinese ByteDance (parent company of TikTok) and the American Microsoft occupy the last places in the top 10 companies, for GenAI-type patent applications. Furthermore, the image is the field that has seen the most deposits, but WIPO also reveals the very strong growth of everything related to molecules, proteins and genes.
Warning of certain hazards
The WIPO report also provides its analysis of the development of generative AI. This “has become a revolutionary technology”, emphasizes the head of the PMOI, Daren Tang, in the study. ” This is a thriving field », declares Christopher Harrison, Head of Patent Analysis at WIPO.
But Daren Tang also warns of the potential dangers of this technology, such as massive job losses or lack of respect for intellectual property. Another point of vigilance: if GenAI ” undermines human creativity and prevents a human creator from making a living, I think that’s going to be something we really need to be aware of “.
And the head of WIPO hopes that the designers of GenAI models and content creators find common ground. Such revolutionary technology must keep people” at the center of the innovation ecosystem “, he asserts.
Government initiatives to discuss the risks of artificial intelligence
In mid-May, China and the United States had their first discussion about the risks posed by artificial intelligence. No joint press release or commitment was planned at the end of this meeting.
But a senior US official, interviewed anonymously by AFP, expressed his concern just before this meeting:
” China has made the development of artificial intelligence a major national priority and is rapidly deploying capabilities in both civilian and military fields, often in ways that we believe harm both US national security and that of our allies. »
On the other side of the Atlantic, it is also moving. By 2023, the UK acquired a AI Safety Institute. This organization whose goal is to assess the risks of large language models “the most advanced”, submitted an initial report a few months ago expressing its concerns. On the EU side, the subject is also taken seriously. The recent vote on the AI Act testifies to this. This European law thus established a legal framework that enabled better evaluation of artificial intelligence.
The generative AI revolution is underway
For the record, the general public really discovered the power of generative AI with the ChatGPT application in late 2022. Today, GenAI operates a number of industrial and consumer products, including chatbots such as ChatGPT and Gemini from Google. AI can also contribute to the design of new molecules for drug development and enable the design and optimization of new products.
Since then, these computer programs have nurtured and “ Trained » using billions of data, you manage to create text, video, music and also computer codes in seconds, based on queries in everyday language. GenAI patents still represent only 6% of all AI patents worldwide, but the number of applications is growing rapidly.
An international association fighting for a break from artificial intelligence
“Stop the race to suicide”, “ Let’s stop artificial intelligence, because you wouldn’t board a plane with a 14% accident risk.” This is what we could read in mid-May on the signs of activists from the global association Pause AI, in Paris. At the same time, nine other demonstrations were held by the same organization in London, Berlin, San Francisco, New York, Rome, The Hague, Stockholm, Oslo and even Sydney.
Pause AI, which was founded by Joep Meindertsma, a Dutch engineer, calls, as the name suggests, for a pause in the development of this technology and calls for strengthening its security. How ? By the establishment of a binding international agreement ending the arms race in the most dangerous AI systems, the creation of an international agency to enforce this agreement and a global referendum on the development of superhuman artificial intelligence.
“There is total ignorance on this subject in France, which leads our leaders to make terrible decisions”worried about the gallery, Maxime Fournes, creator of the French division of Pause AI.
(With AFP)
