The successful path of Chinese AI can be summarized as “Fail fast, learn fast, improve fast”.
Recently, Zhou Wei, founding partner of Chuangshi Partner Capital, shared the investment strategy of Chuangshi Partners and his thoughts on the ecology of China’s venture capital at the TechCrunch International Innovation Summit Forum (Shenzhen).
Zhou Wei believes that in China’s venture capital market in the past 20 years, the pain points of education and medical care have not been fundamentally solved through the “Internet” because these two areas are not facing the problem of “connection”, but The problem of “unbalanced supply and demand” should be solved by new technologies such as AI.
In his view, the success path of Chinese-style AI can be summarized as “Fail fast, learn fast, improve fast”. Entrepreneurship is a high-risk thing, especially Internet and technology entrepreneurship, usually the market where winners are all-in-one. The failure fails early, don’t waste time, because the slower the failure, the more time and money is wasted.
The following are some of Zhou’s speeches, recorded and organized:
Hello everyone, I am Zhou Hao of the creation partner capital. Our fund is relatively young, just over a little over two years. “Fail fast, learn fast, improve fast” is our overall observation of the Chinese AI industry.
I have worked in Silicon Valley funds for ten years. Compared to Silicon Valley and Chinese startups, 2011 is a boundary. Before 2011, China was copying the US model. After 2011, there were many innovations in China. These innovations were innovations in the copied model. The business model and many changes of Chinese companies are much faster than the United States. Weibo was the same as Twitter at the earliest, but today’s Weibo and Twitter have no similarities.
Chinese entrepreneurs and ordinary people have a characteristic, “quick iteration”, but have not heard the word in Silicon Valley. The products that Silicon Valley entrepreneurs, investors and users want to produce are perfect and there are not many problems. However, in China, there are often entrepreneurs who have a version at 5 pm and another new version at 12 pm. Through the actual use of feedback to do iterations, Chinese users have higher tolerance and poor patience, so they keep Iteration.
We must admit that there is a gap between China’s AI underlying technology and the United States. China’s AI is equivalent to IQ 120 normal children, but the US AI may be an IQ 130 or 140 more extraordinary child. But the American AI is in the lab, it doesn’t come out, it’s overprotected, it’s 18 years old, it’s smart, it’s good, but it’s actuallyThe ability to live is very poor. In China, the technology that has not yet fully matured is used in a commercial environment very early, just like a child of three or four years old can help you to soy sauce. Because of the early exposure to the real environment, running with real data, and its IQ, China’s AI is likely to be ahead of the US in a period of time.
For example, we invested in a company that did one-on-one English teaching. When we were at KPCB, when we talked about AI education, our partners would say that we would like to invest in a pure virtual teacher, but this technology will take 5-10 years. But in China, we have a very Chinese way to solve this problem: use the “AI + people” approach to solve. Our product takes a 30-minute class, and any scene that the teacher encounters in the class is all taken in small pieces. The finer the better, the AI is only responsible for collecting the child’s reaction and pronunciation pronunciation from the camera. According to the score, select an appropriate teacher from the background to give this child, because the child does not feel too much. This teacher has something abnormal, which is a one-to-one process from the child’s experience.
China’s Internet and venture capital have been going on for 20 years. The theme of these 20 years is “connecting”: connecting people and people, connecting things and things, connecting business and business. However, there are two particularly important areas where the Internet has not worked, one is education and the other is medical. why? Because the problem in these two areas is not a “connection” problem.
Internet education and Internet healthcare have been invested for more than a decade. Before 2015, Internet education and Internet medical education invested a lot of money, but no real big company successfully proved its business model and went to everyone’s recognized success. why? Because these two areas are imbalanced supply and demand, good teachers, good schools, good hospitals, and good doctors are not enough.
Now, we can use technology to solve the problem of insufficient supply of quality resources. How to solve it? For example, we voted for “叮咚 classroom”, they are equivalent to cloning a good teacher, and can give countless children at the same time, which solves the problem of insufficient supply. We voted for “Diankun Technology” to do AI imaging diagnosis, so that the diagnostic accuracy of the inexperienced doctors and hospitals is as good as that of Peking Union Medical College Hospital. It is equivalent to giving an ordinary man an iron man’s armor and turning into Superman.
When you look at the AI+ medical field, you will find that almost all companies are doing lung nodules because AI requires real data for training, and lung nodules data many years ago. It’s already open, everyone can use it online. , everyone will take the lung nodule products for AI when they do the imaging products. China’s AI is very radical in business, we pass the realThe operation has accumulated barriers to competition.
Entrepreneurship is a high-risk thing, especially Internet and technology entrepreneurship, which is usually a market where winners are all-in-one. The failure fails early, don’t waste time, because the slower the failure, the waste is all People’s time and more money.
Our generation of entrepreneurs, on average, spent eight years from start-up to listing, and there are countless opportunities to make mistakes in the middle, and they can change. But today there are countless companies that have been listed for business in three years. Under this rhythm, entrepreneurs have no chance to make mistakes. You have no possibility of making strategic mistakes. If you want to win, you have to be all right. The ability of the team is much stronger than before.
This speed is also a huge challenge for VC. When I entered VC in 2012, it took no problem to see a field for a year. Every project in the field, there are no problems in the due diligence for several months, but today, look at a field within 3 months. If you don’t invest, your chances in this round will be gone. So, our pressure is really great, forcing us to learn quickly and be very focused. Ten years ago, I have seen any good company, but many good companies today have not seen me. Our team has limited people and limited numbers of people. We can only focus on operations. Today’s environment is also forcing VCs to become more and more focused and more vertical.
I have been in VC for 12 and a half years and have experienced 4 cycles. Looking back, many companies on the market have invested the best time in late 2008 and the first half of 2009. Many good companies were at that time. There was no money, and it was repeated several times later. It was a similar situation. If you have money, this is a very good investment opportunity, and we are very active in finding investment targets.
All small businesses today, such as independent small hotels and small clinics, are online, can be posted on the reviews, and hang on the US group for sales, but these have no use for them. Because they have to spend money in front of the search page is impossible. We call these small businesses “the forgotten half”, but now we have a good technology and platform to help them.
For example, many small clinics in China have only one doctor and a small room that cannot be tested. The company we voted for “Cloud Call Technology” passed the “Mei Tuan + Mei Cai” way, so that the clinic owner can manage all patient cases and can call with one click. The staff of Yunhu Technology will go to the clinic to obtain the patient specimens and send them to the nearest third-party inspection agency through the self-built cold chain logistics system. After the test, the results will be returned from the cloud to the clinic and the medicine will be prescribed directly. In this case, the patient does not need to go to the hospital, and the small clinic does not lose the patient. <