news

August 13 news, Titanium MediaAGIExclusive statistics from the National Enterprise Credit Information Publicity System:sinceChatGPTFrom the date of publication (November 30, 2022) to July 29 this year (latest), the number of newly registered artificial intelligence (AI)-related companies in China that are now in an abnormal state of cancellation, revocation or suspension of business reached 78,612, and nearly 80,000 companies have disappeared, accounting for 8.9% of the total number of 878,000 newly registered companies in the same period.

Titanium Media AGI also exclusively calculated that in the past three years, China has cancelled or revoked the registration of more than 200,000 AI-related companies, and a total of 353,000 AI-related companies have disappeared in China in the past 10 years.(Including cancellation, revocation, revocation, liquidation, suspension of business, delisting, order to close, etc.) As of August 7, 300,700 new AI-related companies have been registered this year.

(Note: Statistics include companies whose names, brand products, and business scope include "artificial intelligence")

As of now, there are 1.8043 million AI-related companies in China. Among them, the Ministry of Industry and Information Technology has counted more than 4,500 companies related to China's AI industry system, and China has completed the filing and launched more than 180 generative AI service models that can provide services to the public, with more than 564 million registered users.

On the one hand, influenced by the AI ​​entrepreneurship boom, hundreds of thousands of new AI-related companies have been added in China; but on the other hand, affected by multiple factors such as the dilemma of "burning money" crazily but not making money, and the downward trend of the venture capital industry leading to a cold financing situation, nearly 80,000 AI companies have been eliminated, forming a new "rotation" phenomenon in the industry.

Not just in the country.

In early August, AI unicorn company Character.AI announced that it was sold to Google for $2.5 billion, and its two co-founders joined Google’s AI team. Earlier, Inflection was acquired by Microsoft, Adept was acquired by Amazon, and Anthropic was facing an annual revenue and expenditure gap of up to $1.8 billion.OpenAICore members frequently change jobs or resign. The US AI field is also facing the crisis of being eliminated.

According to Pitchbook data, in the first quarter of this year, seed round (and earlier) financing for US AI startups dropped sharply to US$123 million, a year-on-year decline of 70%, almost returning to the level before ChatGPT came out in 2022. At the same time, the number of financing transactions for these US AI startups in the first quarter also decreased by 7.8% month-on-month and 31.2% year-on-year.

The cooling of the financing enthusiasm in China's AI industry is also particularly obvious. According to the "AI Industry Status Report" released by research institution CB Insights, in 2023, the number of investment and financing in China's AI field will be about 232, a year-on-year decrease of 38%; the total amount of financing will be about US$2 billion, a year-on-year decrease of 70%.

Zero2IPO Research Center believes that in the first half of 2024, China's equity investment market will continue to decline as a whole, the number of active institutions in the market will decrease, and institutions will generally reduce their investments. According to incomplete statistics, the number of early-stage investment, VC, and PE institutions in domestic outbound investment activities in the first half of this year decreased by 23.9%, 19.2%, and 25.2% year-on-year, respectively; at the same time, the average number of investment cases by a single institution decreased by 17.6%, 30.4%, and 19.3% year-on-year, respectively.

According to the data from the "2021 Research on the Current Operation Status of China's Small and Medium-sized Enterprises" released by the team of Ren Zeping, vice president of the China Private Economy Research Association and chief economist of Zhongyuan Bank,The average life cycle of Chinese companies is about 3-5 years, among which the average life cycle of small, medium and micro enterprises is only 2.5 years. This means that China's small, medium and micro enterprises not only have a short life cycle, but also very few of them can grow big and strong.

Currently, the high computing cost, extremely high R&D investment, and difficulty in profitability in the field of generative AI have become major challenges facing AI companies.

Founder of SogouBaichuan IntelligenceFounder and CEO Wang Xiaochuan once told Titanium Media AGI that the domestic AI large model market will be in a knockout stage in the future. "The first echelon may not exceed five companies in the end, for example, there will be three 'big companies'; and then two startups, one of which will be Baichuan."

Wang Xiaochuan emphasized that as the domestic large-scale model price war dropped to free in May this year, the entire To B model market will prosper faster. There is value space, but the tail companies will exit this track."Pearls will appear at the end of the tide, but there must be a bubble here, which will accelerate the bubble cycle and make it more prosperous. This is an inevitable thing in business."

Wang Guanchun, chairman and CEO of Laiye Technology, recently told Titanium Media AGI that there is definitely a bubble in the AI ​​industry, and companies like OpenAI may not win in the end. Even the product ChatGPT may not be used by anyone in five years, so in this process, the company still needs to find its own positioning.

Baidu founder, chairman and CEO Robin Li said that many large models are "reinventing the wheel" and the "100-model war" is a waste of social resources. Moreover, the more than 200 models in China are not used much, and no company can "occupy" all opportunities in the era of large models.

"'Volume' is normal. When anything new or a big opportunity comes, everyone will definitely rush in, and then there will be a process of survival of the fittest. This is very normal," said Robin Li.

Talking about the gap between the Chinese and American big models,Li Yanhong emphasized that from the perspective of technology quality, there is not much difference between China and the United States in the level of leading large models. However, when it comes to enterprise-level model applications or models in small languages, Chinese companies do not have the energy to do special optimization, so we (China) may indeed lag behind.

In the next step, China will strengthen independent control in the generative AI industry, stimulate innovation, guide application implementation, empower industry development, maintain the bottom line of security, prevent abuse risks, adhere to collaborative governance, and create a good ecology.