Home About us

plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative Qizhi

China Business Strategy 2024/06/27 14:16

Huashang Taoluo, plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative wisdom


Picture丨Picture Worm Creative, Oriental IC, Internet

Source丨Market Value Observation (ID: shizhiguancha)

Huashang Taoluo, plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative wisdom

With the title of "the first stock of AI + manufacturing", the godfather of innovation Kai-Fu Lee as the chairman, and many strongmen in the science and technology industry to join, so many halo blessings, but the innovation of Qizhi has played a good hand of cards to a pulp.

Since its listing, the company's share price has plummeted by nearly 80%, and since its establishment, it has lost nearly 2.3 billion yuan (before adjustment). In the AI 1.0 era, the innovative Qizhi incubated by Sinovation Works is far inferior to SenseTime, Megvii, YITU, and Yuncong, and in the AI 2.0 era, the war of 100 models is intensifying, but the company still seems to be a little quiet.

One trend is that the AI industry will gradually replace the Internet and become a new battlefield in the technology industry, but in such a hot field, there is a sense of helplessness in the field of innovation and wisdom.


Not worthy of the name

AI is a track that never lacks stories, but the new story of AIGC (generative AI) hasn't made some businesses any sexier.

In the 1.0 era, due to the focus on R&D and the difficulty of commercialization, most of the top players have not found a clear business model, and there are few companies in the industry that can successfully achieve profitability.

Kai-Fu Lee also said very frankly that the industry went from competing with each other to leaving only "chicken feathers" in the end.

On November 30, 2022, ChatGPT, a chatbot program developed by OpenAI, was born. Under his leadership, the wave of generative AI (AIGC) has swept the world, and artificial intelligence has officially moved from the 1.0 era to the 2.0 era.

The AI companies that have walked to the edge of the cliff are once again full of enthusiasm, and the venture capital circle, which was originally silent, has become very lively. Traditional AI and Internet giants have turned their heads in an attempt to seize the commanding heights.

Of course, Qizhi didn't want to miss this epic super opportunity, and Kai-Fu Lee described AGI (Artificial General Intelligence) as a dream he had been waiting for for more than 40 years.

In April 2023, the company launched its first AIGC product, Qizhi Kongming AInnoGC, which is oriented to vertical scenarios based on manufacturing and meets the commercialization needs at the 2B level. On March 27 this year, the Qizhi Kongming industrial model was updated to version 2.0 (AInno-75B) and released a number of native applications for large models.

In order to stand out in the highly competitive track, Qizhi Innovation has focused on empowering the manufacturing industry with AI technology since its establishment, trying to create a differentiated competitive advantage.

Prior to its IPO, Qizhi had received at least 7 rounds of financing from investors including Sinovation Ventures, CICC Jiazi, Huaxing New Economy Fund, SoftBank Vision Fund, etc.

Kai-Fu Lee has served as global vice president at Google and Microsoft, and was named one of the 25 "AI Leaders" in the world by Time magazine in 2023. One of the major features of Innovation Qizhi is that under the call of Kai-Fu Lee, the company has recruited a group of world-class scientific and technological talents.

In an interview, Kai-Fu Lee made no secret of his ability: "I can be the world's biggest AI headhunter, bring together the best people, and tell them to do their own part." ”

For example, Xu Hui, CEO of Qizhi Innovation, has a very dazzling resume, graduated from Shanghai Jiaotong University with a major in electrical engineering, and has successively held important positions in technology giants such as IBM, SAP, and Microsoft. After serving as vice president of Wanda Network for less than a year, he gave up his annual salary of 8 million yuan and came to the sea of stars of AI with Innovation Qizhi and Kai-Fu Lee.

With so many strongmen to help, the company has never been on the right track. Even though the veteran of science and technology, Kai-Fu Lee, personally took the lead, the development of innovation and intelligence could not keep up with people's expectations.

In the AI track, Innovation Qizhi can not catch up with the first echelon such as Baidu, Huawei, and iFLYTEK, and it also lacks bright spots compared with companies such as SenseTime, and even start-ups such as KIMI and Baichuan Intelligence have more voice.

Compared with the enviable management team and the blessing of Kai-Fu Lee, the godfather of AI, the performance of Innovation Qizhi is somewhat unworthy of the name.


The profitability conundrum

After the baptism of the capital winter, the capital market attaches more importance to profitability and landing. Gone are the days of financing by dreams.

Realistic performance is a cruel touchstone, and the capital market will not give too much "preferential treatment" because of Lee Kai-fu's dazzling aura.

At the time of the IPO, Qizhi positioned itself as an "enterprise-level AI-driven solution provider", and especially emphasized its position as a "manufacturing AI leader". But its scale is still small, with a revenue of only 460 million in 2020, which is much worse than the tail of the "Four Little Dragons", Yuncong Technology.

After three years of development, the company's revenue growth is gratifying, achieving revenue of 1.751 billion yuan in 2023, but only half of SenseTime (3.406 billion yuan) and only about 40% of 4Paradigm (4.204 billion yuan).

Moreover, the company has continued to lose money since its inception and has not yet seen the light of day. From 2019 to 2023, the losses will be 240 million yuan, 360 million yuan, 640 million yuan, 360 million yuan and 580 million yuan respectively, and the adjusted losses will still be 160 million yuan, 144 million yuan, 142 million yuan, 138 million yuan and 154 million yuan.

Huashang Taoluo, plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative wisdom

At the "Qizhi Kongming AInnoGC" press conference, Xu Hui emphasized that he would never be an artificial intelligence company that would only burn money.

Of course, this is a big problem for all AI companies. It's just that the operation of innovation and wisdom makes investors quite puzzling. Especially before going public, the company raided to "optimize" the R&D team and outsource more business.

Moreover, the equity incentives of R&D personnel are significantly less than those of administrative staff. In the first three quarters of 2021, Xu Hui took 57.72 million in equity incentives, exceeding the 47.24 million of all R&D personnel combined.

In 2023, the sales cost of Qizhi Innovation will reach 1.163 billion yuan, a year-on-year increase of 10.7%; Although R&D expenditure increased by 8% year-on-year to RMB450 million, the ratio of R&D to revenue decreased from 26.7% in the previous year to 25.7% in 2023. Administrative expenses were basically equivalent to R&D expenditures, which amounted to RMB407 million, up 23.4% year-on-year.

In the second half of 2023, the company's revenue will be 827 million yuan, with negative growth both year-on-year and month-on-month. But to the surprise of investors, the company still gave CEO Xu Hui 43.562 million yuan and independent non-executive director Xie Deren 8.074 million yuan in share incentives.

Xu Hui is arguably the highest-paid CEO of China's AI innovation companies, with Xu Li, chairman of the board of directors of SenseTime, receiving only 3.305 million yuan, Zhou Xi, chairman of Cloudwalk Technology, less than 2 million yuan, and Dai Wenyuan, chairman and CEO of 4Paradigm, less than 1.2 million yuan. But their market capitalization is much higher than that of Ingenuity.

In addition, the company has changed its CFO twice in the two years since it went public. On January 30 this year, the company announced that Xiao Lei resigned as chief financial officer, and Yu Jin served as the chief financial officer. The CFO of the IPO listing is Cao Jun, who was born in an investment bank.

Executives can receive more than 50 million yuan in compensation every year, CFOs are frequently replaced, and the value creation ability is so weak, which raises investors' questions, and some shareholders ask: What is the standard and fairness of the company's compensation? What is the significance of ultra-high equity incentives?


Win vertically?

In the competition of large models, players are hoping to create differentiated competitiveness.

SenseTime has shifted from visual recognition to generative AI, and its generative AI business revenue reached 1.2 billion yuan last year. iFLYTEK is known for its voice recognition and has a deep accumulation in the field of education. Baidu is taking the lead in the field of smart transportation.

In April 2023, the company launched its first AIGC product, Qizhi Kongming AInnoGC, which is different from other general-purpose large models, and focuses on private deployment of AIGC engineering algorithm engine for vertical industry applications.

However, whether it is technology independence, version iteration or running through the commercialization path, it is a bit slow.

For example, the SenseTime Ririxin large model launched at the same time has been iterated to version 5.0, and its comprehensive performance is fully benchmarked against GPT-4 Turbo, with 600 billion parameters.

On January 30 this year, iFLYTEK released the first Xinghuo V3.5 based on the training of national computing power. The company expects to release Xinghuo V4.0 on June 27, which fully benchmarks the current level of GPT-4 Turbo. iFLYTEK also worked with Huawei to build the first domestic computing platform "Feixing-1" that supports the training of trillions of floating point parameter large models......

In comparison, it was not until March 27 this year that the Qizhi Kongming industrial model was updated to version 2.0 (AInno-75B), and the number of parameters exceeded 75 billion.

Although the number of parameters is not the fundamental criterion for measuring the ease of use of large models, the update speed of Qizhi Kongming's industrial large models is really slow, and there is still a big gap with the leading enterprises.

"No computing power, no model", computing power is the foundation of AIGC, and the wisdom of a large model depends on the number of parameters, just like humans learn a language, only massive data learning and imitation can build a rich and efficient model.

However, the accumulation of computing power is relatively weak, and in the context of the control of Nvidia chips, the competition of Qizhi Kongming in the future will gradually decline.

At the same time, Alibaba, iFLYTEK, SenseTime, and Huawei have all begun to lay out small models in vertical fields and have received orders from all walks of life. The revenue of SenseTime's generative AI business alone is already close to the overall revenue of Qizhi, which is already in a disadvantageous position in terms of commercialization, resource integration, branding, technology research and development, computing power accumulation, etc., and it is even more difficult to catch up with the leading enterprises.

In 2023, iFLYTEK's automotive intelligent network business will contribute nearly 700 million yuan in revenue, a year-on-year increase of 52%. The revenue of the innovative Qizhi automotive equipment business is less than 400 million, which is still its largest single manufacturing business category.

In other words, the self-proclaimed single champion of "AI + manufacturing", but the actual performance is not as good as the large model with a general computing power base.

Huashang Taoluo, plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative wisdom

In addition to facing competition from the outside world, Kai-Fu Lee also personally founded Zero One Thousand Things and released a large model Yi-Large, which will inevitably compete with Qizhi Kong Ming in the future.

Under the influence of a variety of unfavorable factors, the backward innovation and wisdom in the 1.0 era may continue to be disappointed in the 2.0 era.

At present, the total market value is only 3.3 billion Hong Kong dollars (closed on June 26), while the company's valuation in the C+ round of financing is 8.5 billion, and the C round valuation is 1 billion US dollars. In other words, the current market capitalization is not even half of the Series C series. It is less than 1/3 of Yuncong Technology, and only about 1/16 of SenseTime.

Huashang Taoluo, plummeted by 80%, and it was difficult for Kai-Fu Lee to save the innovative wisdom

Back in 2018, when the company was founded, Xu Hui emphasized: "It is necessary to combine technology and business, which is the attitude that a responsible AI innovation company should have." It is not only about burning money, but also about generating commercial value. After generating business value, you can naturally make money, which is a good positive cycle. ”

Contrary to expectations, today's innovation and wisdom seem to be entering a negative cycle.

This article is from Xinzhi self-media and does not represent the views and positions of Business Xinzhi.If there is any suspicion of infringement, please contact the administrator of the Business News Platform.Contact: system@shangyexinzhi.com