Lu Qi is obviously so close to OpenAI, and so far away

Author|Qin Anna Editor|Yuanye

01

After the Spring Festival, Lu Qi's itinerary crosses along the east and west coasts of the United States, flowing between Harvard, Berkeley, Stanford and other universities. He discussed with students who were born two or three decades later than him about technology trends that could be commercialized: AI, robotics, autonomous driving, and projected technology reports from the annual meeting of the Qiji Achievement Arena into the auditoriums of universities.

It was a scene he was quite familiar with. After founding the Qiji Creation Forum, Lu Qi frequently appeared in major universities in China as an entrepreneurial mentor, and in order to meet him, there was no lack of people from other places. Compared with the money that Qiji can give, they value Lu Qi more, "If you can invite Lu Qi to be a co-founder, it will be more convenient to raise money later."

Lu Qi's fame continued in the United States, but employment replaced entrepreneurship and became the hottest topic in the question session.

This is also normal. At present, Silicon Valley is in the turmoil of large-scale layoffs, Microsoft, Amazon, Meta... Companies that can be called and cannot be named are busy reducing costs and increasing efficiency. Chinese engineers seem to be at the center of the layoff storm, at least in terms of the social networking crusade of the layoffs — they criticize the concentration of layoffs in the Chinese community, with few Indians and whites.

Under the new situation, employment is more urgent and important. The popular question in the first two years was to get an offer from factory A and factory B at the same time, how to choose? The popular question has become: Return home or stay in Silicon Valley? Where to go in Silicon Valley? When the question fell to Lu Qi's microphone, his answer was: There are two best jobs now, one is to start your own business, and the other is to join OpenAI.

Even for domestic users, the name is not unfamiliar. OpenAI's ChatGPT is sweeping the world, setting off a new artificial intelligence boom. Some even assert that a new industrial revolution has arrived.

The "revolution" is still in doubt, but artificial intelligence is becoming one of the few tracks that can break boundaries and build consensus. Bill. Gates said ChatGPT allowed him to see the most important technological advancement since graphical interfaces. "Generative AI is a new type of computer where we can program in human language and anyone can command a computer to solve a problem," Huang said. He even distilled a very popular subtitle: AI ushered in the iPhone moment.

Ma Yun, who has not appeared for a long time, also recently talked about ChatGPT, although he still stands on the side of humans, believing that machines only have "cores" and people have "hearts", but his focus on artificial intelligence has fully proved its popularity. As for Lu Qi's attention, it couldn't be more normal. The miracle he founded alone has invested in 95 artificial intelligence projects in 3 years.

However, in the classroom of American universities, Lu Qi advised students to go to OpenAI, but did not recommend them to Microsoft, his old club.

With its investment in Open AI, Microsoft has changed its weakness in the mobile ecosystem and stood up in the new era of artificial intelligence. The new Bing, which integrates OpenAI technology, has seen an 8-fold increase in downloads in the global app market in the past month. This is still when the new Bing is in closed beta and is not open to all users.

Aggressive Bing has made other search-based businesses nervous. Google, which is in a defensive position, followed by a presentation of conversational AI Bard at the Paris conference, unfortunately, because of the wrong answer, the market value fell by hundreds of billions of dollars. After all, Bard is a blockbuster product personally developed by Google CEO Sundar Pichai, who also called on 80,000 Google employees to participate in feedback in an internal letter, but unexpectedly suffered a mild rollover in his debut.

Baidu, Lu Qi's previous former owner, also launched a product "Wen Xin Yiyan" that benchmarks ChatGPT. Although it does not replicate the effect of the opponent's debut and peak, judging from the team configuration developed by Robin Li himself, Baidu's determination is firm enough.

Unfortunately, there was no Lu Qi at Wen Xin's press conference, and the glory of the new Bing was not directly related to him. Whether it is from technical ability, vision or network resources, Lu Qi should be able to become a pivotal figure in the AI process. But the height of achievement that a person can achieve has always been related to luck. If you don't step on the rhythm, even a genius can only silently endure the fool of fate.

02

Lu Qi was once called "the most powerful Chinese executive in Silicon Valley." His 18-year career in Silicon Valley can be summed up simply: competing with Google in different companies. The first decade was Yahoo, and the last eight years were Microsoft. Bing, which has recently skyrocketed in the search market, is a product made by Lu Qi and his team. In 2008, Lu Qi joined Microsoft and reshaped Microsoft's search business, launching Bing Search to replace the original Live Serch.

In July 2013, then-CEO Ballmer led Microsoft's largest business restructuring in the company's history, merging the original eight product divisions into four: operating system (OSG), application software (ASG), server (C+E) and hardware (WDG).

Lu Qi became one of Microsoft's "four kings" in this restructuring, responsible for the entire application software department, in charge of a series of important products such as Office, Bing, Skype, etc., and then he sat as the global executive vice president, becoming the highest management position in the US technology industry.

Position is just one of many elements of proven competence, and ideas are more cyclical than ideas. For example, Lu Qi proposed the concept of "dialogue as a platform" during his time at Microsoft, that is, there is no APP in the mobile phone, only services and assistants. Users can use voice as a more natural way of interacting with each other. Even at the moment, this is still an accurate judgment of technology trends, at least in the GPT model, out-of-the-loop products and application scenarios are presented in the form of dialogue.

But back to 2016, even if Lu Qi announced at the Microsoft Build 2016 developer conference that "today, we finally have the conditions for dialogue as a platform" - the Microsoft Xiaoice born for two years has tens of millions of users, accumulated more than 20 billion conversations, and accumulated a huge amount of data, but back to reality, Microsoft actually did not get any tickets for the mobile Internet era, it tried to cooperate with Motorola, used their hardware + Microsoft software to do, did not do it, and finally bought Nokia, Nor did I find a more suitable product form.

Voice interaction is nothing new in Silicon Valley. In 2009, Microsoft integrated voice functions in the Win 7 operating system, and then Google and Apple launched their own voice search products. However, these products did not really open the market, until Amazon found an interactive form of smart audio, it set off a "hundred box war" of Internet manufacturers, and these lively things have nothing to do with Lu Qi and Bing.

Lu Qi really entered the Chinese media scene on January 28, 2017, which was his first appearance after joining Baidu. In Baidu's "Qingyu Case" conference room, Li Yanhong, wearing a blue and white T-shirt, and Lu Qi, dressed in Polo shirt and jeans, sat side by side, responding to more than a dozen reporters' curiosity about Lu Qi's joining Baidu.

It was a very early year for China to embrace the concept of artificial intelligence. Alpha Dog defeated world Go champion Lee Sedol in 2016, setting off a wave of artificial intelligence around the world. According to CB Insights, Chinese AI startups surpassed the United States for the first time in 2017 and ranked first in the world. The concept of artificial intelligence was first written into the government work report, and representatives of the two sessions, such as Ma Huateng, Li Yanhong and Lei Jun, all mentioned artificial intelligence in their proposals.

Lu Qi returned to China when China's technology circle is very hungry for artificial intelligence. Bill. Gates repeatedly retained him, saying that the conditions given by Baidu can be given by Microsoft. Lu Qi's answer was: You can't give me China.

Lu Qi wants to contribute to China's AI development. He said that he chose Baidu because of technology, but also because of Baidu's accumulation in commercialization and changing scenes. He also brings new stories to Baidu. After Lu Qi joined, he reorganized Baidu's business lines and led the integration and establishment of three AI-related business units: Baidu Intelligent Driving Business Group (IDG), Baidu AI Technology Platform System (AIG), and Intelligent Life Business Group (SLG). In addition, he also led the launch of the conversational AI operating system DuerOS and autonomous driving Apollo, making Baidu's AI products initially clear.

As a professional manager, Lu Qi has accumulated more good reputations in Chinese Internet companies that may be the most suitable for him.

However, the development of artificial intelligence has always been extremely cyclical, it is often aroused by an exciting technological progress, and it will always enter the winter slowly because of the slow commercialization process. Baidu, which is obsessed with AI, continues to show its technical genes and determination to the outside world, but it is still unable to change the laws of the entire industry. What's more, at that time, it had not yet gained a foothold on the mobile Internet, and the competition for information flow with ByteDance also involved great energy.

Lu Qi's time in Houchang Village did not last long. In July 2018, Lu Qi stepped down as president and chief operating officer of Baidu Group, Baidu's stock price fell in response, and the capital market expressed direct concern about the prospects of Baidu's AI business.

03

After 7 years, technology and capital are once again rushing to the AI track.

Wang Huiwen, who has nothing to do with AI, ended her retirement life and hurriedly joined in, bringing her own funds and flesh into the shares, and forming a game. Kai-Fu Lee, who has dealt with AI for nearly 40 years, also announced that he personally organized Project Al 2.0. Wang Xiaochuan, who was voted by the media to be the Chinese version of ChatGPT, has established an AI company. Go out to ask CEO Li Zhifei and former Amazon chief scientist Li Mu, both of whom have successively voiced their desire to create a Chinese version of OpenAI or ChatGPT.

The new outbreak of the AI wave has allowed the big manufacturers and bigwigs who have swayed and hesitated in the technology trend in the past few years to find a new channel. But among them, Lu Qi's voice did not come out.

He is very close to Open AI, very far away.

In addition to his years of accumulation in artificial intelligence, among the contacts he has established in Silicon Valley and even the global technology circle, there are many influential figures who have long paid attention to artificial intelligence, such as Bill Gates, Musk and so on. He has also known Sam Altman, the former CEO of Open AI, for many years and has a close personal relationship.

But comparing his personal achievements with technological trends, it is not difficult to find that in the face of those key nodes that change the direction of the trend, he is either a little early, a little late, or misses the golden period of a company that fits the times. He went to Microsoft at the time of the rise of mobile search, and Microsoft and Google were basically not on the same order of magnitude. His choice to use speakers to present the concept of "dialogue as a platform" fell behind Amazon.

From any context, Lu Qi stood very close to the creation of "OpenAI", but he was always a breath away. It is obviously very close, but it has never really been reached, and this regret may also be the farthest distance.

But 62-year-old Lu Qi is still on the road. He has publicly stated more than once, "We are optimistic about the innovation ecology and investment environment of China's technology, and there will be more and more opportunities in the future." Past experience, including those missed regrets, can become the blessing of today's entrepreneurial mentorship.

The abundant period of AI development will appear every few years, which means that the opportunity to enter the game will always exist.

China's first wave of AI began in 2012, after deep learning made a breakthrough in using AlexNet and Dropout to process ImageNet, AI entrepreneurs appeared in the market, and companies such as SenseTime and Megvii that started with face recognition were popular in the capital market. Tsinghua's "Yao class" students are popular candidates in this field, and Megvii's co-founders Yin Qi, Tang Wenbin and Yang Mu are all from Yao class.

Star entrepreneurs have emerged one after another, bringing a steady stream of new topics, but during this period, AI has not made much splash in the capital market. Take a closer look at the application direction of AI technology: face recognition, video surveillance, AR interaction, etc., can basically be classified as to B business, not the C-end mass products that capital likes.

The change happened in 2016. Google's AlphaGo and South Korea's Lee Sedol's game of the century has pushed AI to the forefront. Tang Xiaoou, founder of SenseTime, once described it as follows, "After this chess game is played, artificial intelligence does not need us to explain, and everyone suddenly understands." Happiness comes too quickly, Zhu Long, founder of YITU Technology, one of the four AI tigers, once described the scene of being sought after by capital: "After AlphaGo, we started financing in March, and we can add $100 million in April (valuation), another $100 million in May, and another $100 million in June." I didn't do anything, that's it. ”

Capital threw money and heated up the field. The ideas output by some industry leaders have brought the tech community into the dream of "the eve of the outbreak of artificial intelligence". For example, Lee's views at that time can to some extent represent the scientific and technological community's view of AI: artificial intelligence will be the best opportunity for China to lead the world, because China not only has a relatively open policy, but also has the key to the development of artificial intelligence - massive data.

One of the important signs of the industry's rise is that valuations have skyrocketed. Kai-Fu Lee described that era, "three AI experts can estimate 700 million." There are doubts in the market, such as is the AI bubble too big? How do these companies make money? But capital can't take care of it, and the fear of missing out on the wind overcomes the rational thinking of business.

When the capital accumulated more than 220 billion yuan in China's AI field, the result was that the four AI tigers entered the "entity list" of the US Department of Commerce one after another. The dream of an AI business empire is still far away, and the shackles have shown a cold metallic glow.

Since then, the keywords surrounding AI companies have become "financing difficulties", "layoffs", and "unclear business models", and the AI industry has become a bubble that has been quickly blown up and quickly burst. Many investment institutions said that they no longer look at the AI track, like the scene when the couple in love broke up dismally.

04

When China's AI industry entered the dark night, the dawn shone into Silicon Valley.

The deep learning model of Transformer proposed by Google scientists is greatly more efficient than the previous recurrent neural network (RNN) in the field of natural language processing (NLP). Based on the Transformer model, OpenAI, the top stream of Silicon Valley, launched the GPT model, and iterated all the way to GPT-4. The launch of the popular product ChatGPT upgraded the advantages of artificial intelligence in processing data and information to provide user knowledge and insights, and the C-end was ignited.

Different from any previous wave of AI, this time, ordinary users may feel the charm and power of artificial intelligence. It can talk, write scripts, draw, design, and many people engaged in professional work are beginning to worry that their jobs will be replaced by AI - like social animals waiting for layoffs in the wind for the past two years.

In the era when artificial intelligence was still an intellectual disability, human beings have long established a vigilance against the excessive evolution of artificial intelligence, and film and television works with similar themes have become their own school, from the movie "HER" to the TV series "Westworld", which essentially explores the reaction of artificial intelligence to human society. Radical as Musk also warned in December that OpenAI-developed ChatGPT was "too good to be scary, and we are not far from an AI that is dangerously powerful." ”

Silicon Valley continues to play a driving role. A new round of technology competition is raging, Google, Meta, NVIDIA, Adobe and other companies, rolling towards AI large models, competing to hold a press conference, new products intensively unveiled. China's long-dormant venture capital circle has finally waited for the wind of consensus, and large-scale integrated AI has become the hope of the entire Internet village. A group of middle-aged entrepreneurs are struggling to find their last stop before retirement.

For some, this may also be the last chance to spend a fortune on the ideal.

Kai-Fu Lee has gone from coach to entrepreneur. He recently officially announced in a Moments post that he is personally organizing Project Al 2.0, a global company dedicated to building a new platform for Al 2.0 and Al-first productivity applications. Project Al 2.0 funds and computing power are in place, and the new company options are absolutely dominated by the new team.

Wang Huiwen, who retired from Meituan, also invested in the field of AI after lingering in the blockchain and currency circles. The new company invested $50 million and was valued at $200 million. The real names of the two co-creations are not convenient to reveal. He has already roared and roared to publish a hero post immediately, inviting product managers to create the AGI era. Wang Huiwen, who started the business again, did not become Wang Xing's deputy, and on the contrary, Wang Xing became the honorary director of Wang Huiwen. He participated in Wang Huiwen's AI startup company in his personal capacity and served as a director. The WeChat picture is a "3425839" that makes people guess what it means.

Many of the entrepreneurs in this round have become wealth-free, deeply hidden in their achievements and fame. But most of their past achievements are due to the selection of good companies and good partners. At this moment, what they are longing for is a vigor that truly belongs to them - just like Lei Jun, who founded Xiaomi.

After Lei Jun led Kingsoft to go public at the age of 37, he experienced another side of financial freedom: boring. He worked as an angel investor for a few years, and many of the companies he selected that year went public. But these numbers could not make Lei Jun, who was addicted to "Silicon Valley Fire" in college, really satisfied, so he seized the window of the mobile Internet and drank the bowl of millet porridge with the dream of influencing the world.

After the explosion of ChatGPT, Wang Xiaochuan said on Weibo that the success of OpenAI is first and foremost a victory for technical idealism. China needs its own OpenAI, and it needs technological idealism.

Lei Jun back then, Wang Huiwen, Wang Xiaochuan, Li Kaifu and others who are now starting a business, this group of entrepreneurs have all gone through the stage of working hard for money. They worry that they will miss out on the opportunity to participate in the next industry megatrend, and want to complete "The next big thing!" before they retire! ”。 This group of new entrepreneurs entering the AI track, the most abundant motivation is ideal, and the resources that can serve it.

Lu Qi once said: "If you want to really change the world on a large scale, then you must be the founder of this company, otherwise you will always be limited by your employer."

As an employee, Lu Qi is excellent to near perfect. His entry and departure have brought billions of dollars to Baidu's market value changes. When he left Baidu to join YC, YC gave him an affirmation in the announcement that "Qi represents our values." Earlier, then-Microsoft CEO Ballmer described him as "a very rare wizard in the industry." He is an important enough figure in the technology company, and after becoming an entrepreneurial mentor, he also has enough calling.

Now, it depends on how Lu Qi chooses. When he returned to China, the market's expectations for artificial intelligence products were slightly suspended, and this wave of AI is more like an idealistic moment rarely seen in China's venture capital market. Lu Qi, one of the few people in the Chinese technology circle who is suitable for taking the lead in large-scale integrated AI products, do you still have the urge to personally "change the world on a large scale"? Will he "take a step forward" again?