Tips from FT Chinese Website: If you are interested in more content of FT Chinese Website, please search for “FT Chinese Website” in the Apple App Store or Google Play, and download the official application of FT Chinese Website.
At the beginning of the Lunar New Year of the Rabbit, the subversive experience of the generative AI chatbot ChatGPT has replaced the traditional topics of previous festivals, and has become an interesting event for family gatherings and WeChat group chats. With the schedule of launching competing products of some major Internet companies, and some domestic venture capital leaders have announced to join the artificial intelligence research and development plan, a new era of AI is about to emerge. After all, even Bill Gates said that the significance of ChatGPT is no less than the birth of computers and the Internet.
Although future development still faces many challenges such as technology ethics and commercialization, and the possibility of ChatGPT becoming the next “Netscape Navigator” (the pioneer of the browser field that was short-lived before the Internet became popular), there is one thing that is absolutely impossible. Doubtful, that is, it is, and even has opened a new era.
In addition to being based on deep learning and using generative methods to form a breakthrough in artificial intelligence research and development technology, thus completing a major innovation in the technology model, OpenAI, which launched ChatGPT, has also made many innovations in the corporate system design behind this technology model innovation. Around the uniqueness of OpenAI in the design of equity investment agreements, I have focused on the discussion in the column published by FT Chinese website on February 6. What I want to share with readers today is the establishment of an innovation system and the cultivation of an innovation culture behind OpenAI’s technical model innovation.
So, what interesting revelations does OpenAI’s sudden emergence bring to the establishment of China’s innovation system and the cultivation of innovation culture?
First, break the myth of industrial policy.
Many scholars use “ChatGPT has changed the hesitant situation in the global artificial intelligence field by itself” to evaluate the major breakthroughs in the field of artificial intelligence brought about by it. OpenAI, which launched ChatGPT, originated from a small non-profit laboratory dedicated to academic research established in 2015, headquartered in an old suitcase factory in the Mission District of San Francisco.
Even before OpenAI launched this ChatGPT on November 30, 2022, many people would never have imagined that human society would enter the era of artificial intelligence so soon. In fact, this is precisely the charm that innovation brings to us. It is not controllable or even unpredictable.
What makes many people feel is that OpenAI did not receive support from large funds under the US government’s industrial policy and a penny of financial subsidies from the government, but only relied on the support of venture capital in the market. Under the leadership of Altman and Brockman, After seven years of research and development, ChatGPT was finally born and stood out from the crowd.
I understand that the key to the success of American science and technology is to fully mobilize the curious nature of scientific and technological people (Altman and Brockman both have dropped out of prestigious schools), relying on the power of the market (the profit motive of venture capital), so innovation is continuous. Musk (bringing Tesla to China to invest and set up a factory) left, and Altman came again, and so on. In contrast, in China, many people have a strong or weak feeling of concentrating their power on big things, that is, they hope to introduce an industrial policy similar to chips and a national system, and achieve “overtaking on a bend” under the intervention of the government.
The uncontrollable and unpredictable nature of innovation determines that this cannot be foreseen by industrial policy makers. In many cases, the selectivity of industrial policy support distorts the market’s endogenous mechanism for identifying potential projects, increasing the unfairness of innovation competition. Historically, perhaps only the market and time have the ability to truly identify whether a project is successful. OpenAI is actually just one of the very lucky startups exploring unknown projects. What China needs is a truly relaxed, free and inclusive entrepreneurial environment. These may be more important to innovation-oriented enterprises than industrial policy support and government subsidies.
Second, get out of the misunderstanding of public welfare vision.
Recently, I have seen a black picture describing OpenAI’s “equity structure” transferred by friends more than once in WeChat Moments. In addition to introducing the four stages of OpenAI’s profit distribution like many media, this picture specifically emphasizes that the founder and CEO Altman has 0 shares. “The reason is that OpenAI positioned itself as a non-profit organization at the beginning of its establishment.” This kind of description of “a group of great people use their own money to do public welfare to benefit mankind” undoubtedly adds a lot of sacred and lofty colors to OpenAI, which makes this enterprise look a bit unattainable for many start-ups.
In the same circle of friends on WeChat, I replied more than once to friends who forwarded the above pictures, “This is not accurate”. Many people (including the author of this picture) understand OpenAI from the organizational structure of a limited liability company that has not yet been listed, but it is actually a limited partnership (see “The Uniqueness of the OpenAI Equity Investment Agreement Design for Launching ChatGPT”). OpenAI is a limited partnership known as OpenAI LP. OpenAI Nonprofit is the general partner (GP) of the limited partnership in charge of partner affairs. Its employees are also part of the limited partners, so Altman has shares (more precisely, partnership shares). It’s just that from the limited information disclosed so far, we can’t know how much he holds.
In addition to the fact that Altman himself has more or less shares, in view of the fact that the huge R&D investment cannot rely on limited donations, OpenAI is forced to embark on the road of commercialization, although its general partner in charge of partner affairs is called “OpenAI Nonprofit”, translated into Chinese as “OpenAI non-profit”. This internal transformation of OpenAI has partly resulted in the split of the original team at the small non-profit lab originally dedicated to academic research. However, the choice of commercialization path has not changed the original intention of Altman and Brockman to create OpenAI, at least for now. This is embodied in the fourth stage of their equity investment agreement, that is, after the profit reaches 150 billion US dollars, the shares of Microsoft and other venture investors will be required to be transferred to OpenAI Nonprofit for free, which will be used to “ensure the creation and adoption of safe and beneficial General artificial intelligence for the benefit of all mankind”. Obviously, they are using commercial means to finally realize the public welfare vision.
Although this society needs idealism aimed at benefiting all mankind, and more importantly, public welfare activities that benefit others, these should not be the main way to realize the improvement of human well-being. In reality, it is far less effective to rely on public welfare than to pursue interests. At the same time, based on specialization and division of labor, it is more effective to serve the society with the help of the market. Entrepreneurs subjectively pursue their own commercial interests, but objectively they promote the progress of human society. This is just as Adam Smith once said, “We can eat fresh bread every morning, not because of his kindness, but because of his greed.” Many inventions and creations completed by Jobs, Musk and others for commercial purposes can also benefit mankind. In this sense, we may not need to criticize whether these entrepreneurs should maintain their original intention of public welfare, but it is enough for them to maintain the pursuit of interests in the past, present and future. Of course, entrepreneurs can undoubtedly make a greater contribution to human society by taking into account the motivation of public welfare while pursuing profits.
In fact, if we can understand OpenAI’s public welfare model as “a group of great people use their own money to do public welfare to benefit mankind”, here we also remind readers that we should treat some people “with other people’s money, even everyone The money “liarly claims to be doing public welfare to benefit mankind and maintain the vigilance of the habit. I personally think that OpenAI’s public welfare vision is not imitative in China at the current stage, and even simple imitation may bring disaster. The reason is that only profit may be the reason why a business organization is more willing to believe in innovation, otherwise I can only simply suspect that they are liars.
We know that the original intention of public welfare will become an inexhaustible internal motivation for some great people, but for most people it needs external motivation. Public welfare and interests will inevitably conflict in real life. The current business transformation of OpenAI just shows the existence of this conflict. This is the reasoning behind my concern about how far such a model of theirs can go. Therefore, OpenAI, which has made major innovations in its technical model, is still full of uncertainties and challenges in how to continue its public welfare in the future.
Here, I also sincerely wish that OpenAI can find a new path in balancing the relationship between public welfare and interests, just like business model and enterprise system innovation.
(This article only represents the author’s point of view. Responsible editor email:[email protected]）