Young People Diving into AI: Huawei Genius Gives Up Million-Dollar Salary, Gen Z Drops Out to Start a Business, Fearing Failure Less Than Not Making Money
(Reprinted from Sohu Technology, author: Liang Changjun)
Editor’s Note:
Life reignites, like spring willows sprouting, after enduring the harsh winter, finally bursting with vitality.
Everyone is a navigator, in the journey of life, we inevitably encounter difficulties, setbacks, and failures. Facing the baptism of storms, we constantly adjust our course, move forward firmly, and search for our own shore.
Life reignites, is also a re-recognition of self-worth. We need to learn to appreciate our strengths, like the harmony of a qin and se, and accept our shortcomings, like raw jade that needs to be polished to shine.
This road is not easy, but like a spring in the stone, accumulating day by day, eventually converging into the sea.
On this New Year’s Eve, Sohu Finance and Sohu Technology jointly launch a planned report, focusing on the journey of life reignition of individual small characters, bravely facing the challenges of life.
Produced by | Sohu Technology
Author | Liang Changjun
Operations Editor | Zhang Lezhou
After proposing resignation four times, Li Bojie finally left Huawei, where he had worked for four years.
Before resigning, as one of Huawei’s first “genius youths,” he was already the deputy chief expert of a laboratory under the Central Software Institute of Huawei’s 2012 Lab, having received more than 20 company awards, at grade 20 (Technical Expert A level).
Just over thirty, he had a stable upward trajectory at Huawei, “able to continuously level up and defeat monsters on the technical expert path”.
But Li Bojie wanted more experiences and freedom. In July 2023, he and a friend from the University of California, Berkeley, co-founded Logenic AI, adding another AI company near Tsinghua Science Park, the holy land of entrepreneurship in Beijing.
In 2023, the explosive ChatGPT catalyzed a new wave of AI, the calm waters for many years, dropped a big model bomb, Beijing, Shanghai, Silicon Valley, and other places around the world stirred up long-lost waves, countless big names entered the field, hoping to use this opportunity to reignite their lives.
The crossroads of fate also awakened the restless hearts of many restless young people.
If Gen Z Chi Guangyao didn’t take a break from school, he might be preparing for the final exams of the first semester of his junior year. But four months ago, he also faced a choice: continue studying or take a break to start a business?
In April last year, while still a sophomore at Shanghai International Studies University, he and his partner established Yuhex Technology in Shanghai. The two agreed that if the product reached 100,000 users or received an investment before the new semester started, he would take a break to start a business.
But by the last week before school started in September, the product users had just reached thirty thousand. However, fate timely snapped its fingers. Chi Guangyao received a call from Miracle Creation Camp, after a brief interview with him and his team, the founder Lu Qi finally decided to invest.
“Rather than saying it’s about the money, it’s more like the entry ticket to the life I hoped for, maximizing the value of life.” Immediately, Chi Guangyao decided to take a break from school, and by early September, he had completed the procedures to leave school, turning his life completely at the age of 20.
The Biggest Pressure is “How to Make Money”
After starting full-time entrepreneurship, Chi Guangyao only leaves the company to go home at eight or nine o’clock every night, and when he gets home, he reads papers, writes code, and learns about various “king bomb” new technologies.
“Entrepreneurship is more fun than studying, the process of making products is very happy, like a child building with Lego.” Although the pace is faster and more tiring, Chi Guangyao enjoys it.
Currently, Yuhex Technology has launched several products for both B and C ends, among which the fastest-growing is CopyAsk, a software that can translate, summarize, explain, or search with one click using AI, with over 80,000 downloads and more than 60,000 users.
Compared to the once-popular generative AI products like Miaoya Camera, CopyAsk is not a hit. C-end products often only have one chance to make an impression, and the first impression is very important, Chi Guangyao understands this principle.
But he said, he did not make products towards the direction of hits, but more about making products he wanted to make, whatever can bring value, enjoy the process, he would do. “Whether it can become popular is not my focus.”
These C-end products were initially promoted by Chi Guangyao himself, attracting a lot of overly harsh insults and personal attacks. At first, he felt aggrieved, but over time, he took it in stride.
Li Bojie is also adapting to the new pace. Half a year into entrepreneurship, his biggest feeling is freedom, no longer having to clock in every day, work overtime until nine or ten o’clock at night, he can arrange his time more flexibly, and more importantly, he can explore new ideas at any time.
Sohu Technology first met Li Bojie at his company, which is an office that can only accommodate six or seven people, basically his core team: co-founders and early AI, product were mostly his classmates, front-end and back-end development were experts recruited from the society. Due to heavy snow that day, Li Bojie arranged for them to work from home.
Over the past half year, Li Bojie has been leading the team to polish the first virtual digital image product, expected to be released in the first quarter of this year to the United States and other overseas markets.
His goal is to make Logenic AI into a high-quality Agent creation platform, and develop AI Agents with good-looking skins and interesting souls, based on multimodality with personality and emotions, etc.
Although the company is just starting, Li Bojie has gradually figured out the business model—brand and platform revenue sharing, hoping to cooperate with Hollywood, Disney, and other star animations and copyright holders, to validate the market with value, “will accumulate from seed users slowly”.
“How to make money” is also the biggest pressure Chi Guangyao is currently facing.
Chi Guangyao can’t bear to set too high a fee for the product, the current price is still discounted by more than four times, only 14.9/month, but because of the high free quota, the product’s payment rate is less than 1%, monthly payments are not enough to cover employee salaries, he also once compressed his own expenses to within 1500 yuan per month.
Now, Chi Guangyao’s company has more than ten people, product development and iteration are still his responsibility, and he has planned 10 C-end and B-end products for the company, including AI companion project Yuhex Life, and intelligent customer service products developed for catering companies, etc.
Entrepreneurial days are even busier. When Sohu Technology contacted him for the second time, it was already after nine o’clock at night, he had just finished fixing a product bug and was preparing to leave work, rushing to a small shop about to close for a cheap dinner. “I especially hope to be able to clone myself, but unfortunately I can’t, so I’ll do as much as I can.”
“When There’s a Choice, Run Fast”
Before choosing to start a business, both the post-90s Li Bojie and Gen Z Chi Guangyao had stable lives that many people envied. But they leaped forward, breaking free, and stepping out of their original life trajectories.
The first time Li Bojie thought about resigning was in September 2022, when he became interested in AI applications and had the idea of starting a business, “wanting to do something more high-level”.
Five years ago, after graduating with a Ph.D. from the University of Science and Technology of China and joining Huawei, Li Bojie had been responsible for high-performance data center networks and systems. Later, he came into contact with Huawei’s deep learning framework MindSpore, and after OpenAI released GPT-3 in 2020, he gradually realized the potential of large models and began to participate in large-scale high-performance network interconnection projects.
With the emergence of ChatGPT, the seed of entrepreneurship grew even more wildly. At that time, he would spend two to three hours a day chatting with it, testing it with various questions, to the extent that it caused dissatisfaction among his family members.
Li Bojie quickly realized that this was a disruptive technology, but also found that ChatGPT’s interaction mode had not yet reached a state that could be widely used by most people, and there were not many good applications on the market.
“There’s definitely a lot of entrepreneurial opportunities between powerful base models and products mismatch,” not wanting to always “climb the ladder” in a big company Li Bojie, in June last year, hit it off with a friend and decided to start a business.
Thus, he submitted his fourth resignation, giving up a million-dollar annual salary, and resolutely chose to start a business. Since then, Li Bojie’s destination has changed from Huawei’s North Research Institute to Zhongguancun Zhizao Street in Beijing, where he began his entrepreneurial days.
Compared to Li Bojie’s dissatisfaction with the status quo and pursuit of new technologies, the younger Chi Guangyao’s choice to take a break from school to start a business was full of coincidences.
Chi Guangyao majored in Internet and New Media in college, and his GPA was the first in his major for the first four semesters, which could guarantee him a recommendation for postgraduate studies at Tsinghua or Peking University. Although he did not love the courses, he did not skip classes, always sitting in the first row, but in fact, he was tinkering with code and products.
Before college, Chi Guangyao had come into contact with novel continuation models, image generation DALLE, and other AI products, later also trying to use ChatGPT to write code, etc., and began to create various fun and interesting AI products.
During the winter break of his sophomore year, he developed a product, Time Trace, which could record everything on the computer and allow users to search, and later launched BillyGPT, which could help novice users use AI more conveniently. It was these two products that unexpectedly changed Chi Guangyao’s life.
The open-source BillyGPT fell into the hands of developer Zhai Xingji. One afternoon in early April 2023, Chi Guangyao met with Zhai Xingji, who had come from out of town, on the second floor of a bookstore on Wenhui Road near the school, and the two talked for most of the day. Zhai’s idea of GPT autonomous task planning later became a hot topic in the industry.
This made Chi Guangyao realize that Zhai Xingji was a person with many new ideas and strong execution. So when Zhai suggested starting a company, he readily agreed. They registered the company in Shanghai that month, and agreed on two conditions for taking a break from school to start a business.
Time Trace then connected Chi Guangyao with Miracle Creation Camp. At that time, he asked a friend from Tsinghua to help find a UI designer for this product, and the latter posted a hero post on the Tsinghua forum. Although he did not find a front-end developer, he attracted the attention of people from Miracle Creation Camp: “Will you come to the entrepreneurship camp?”
Miracle Creation Camp is an entrepreneurship incubator founded by Lu Qi, a former executive of Microsoft and Baidu, which holds entrepreneurship camps in spring and autumn every year to recruit potential seed projects and provide $300,000 in funding. Under the big model craze, they began to widely search for AI projects.
Chi Guangyao filled out the application form with a try-and-see attitude, applied for the entrepreneurship camp with several products, and interviewed five or six times, but there was no definite news. Later, busy with product iteration and operation promotion, Chi Guangyao almost forgot about it.
On August 25, 2023, Chi Guangyao received a call, and Lu Qi’s voice came through, asking him about his entrepreneurial intentions, project progress, etc. Soon, he received an invitation email from Miracle Creation Camp to join the camp, confirming the investment intention.
“At that time, I didn’t even know this was an important moment in my life.” Mentioning this 8-minute call, Chi Guangyao couldn’t hide his excitement. In fact, he was selected for Miracle Creation Camp’s early admission list a long time ago, this belated surprise allowed him to fulfill the agreement just in time.
Chi Guangyao didn’t know much about why Lu Qi invested in him, but he felt that, compared to the project, Miracle Creation Camp valued people more, especially the fit between people and the mission. He believes that his learning ability, coding ability, and alignment with the entrepreneurial vision are his entrepreneurial advantages.
“Young people still have opportunities, without the pressure of children, car loans, and mortgages, stepping out, you can still survive. Many people step out and ‘die’, anyway, when there’s a choice, run fast.” Chi Guangyao said.
Life Principles VS Business Principles
Chi Guangyao seems to have known what he wanted from a young age, interested in programming and game development in elementary school, he began to self-learn. Later, he also became interested in video editing, embedded development, as well as 3D modeling, rendering sculpting, animation composition, artificial intelligence, etc., he self-identifies as an instinctively “desire-driven beast”.
“Life is a game, how much content you can experience depends on how many skills you are willing to master, how many new fields you are willing to explore.” This understanding of life gives Chi Guangyao a maturity beyond his peers.
Chi Guangyao had been pondering the value of life before the age of 20. “The moment you die, everything resets to zero. If there’s a way to calculate the score of life, I think it should be happiness.”
To gain more happiness points, he wishes to engage in activities that have positive feedback during the process, positive outcomes for himself, and positive outcomes for others. For Chi Guangyao, programming and entrepreneurship are activities that satisfy all three positive feedbacks.
However, the principle of pursuing happiness above all else, when confronted with the harsh realities of business, has also left Chi Guangyao feeling lost. “I thought I could focus on coding after starting a business, but in reality, I had to bear the pressure of paying employee salaries, endure rumors, and make money to keep the company alive.”
Chi Guangyao knows he lacks business management skills, but he says he must deliver on the free quota promised to users. “If I take charge of commercialization, the cost is the company going bankrupt.” As a novice in the workplace, he also struggles with the intricate maneuvers of business negotiations.
From a student to an entrepreneur, and then to a businessman who needs to consider how to make money, Chi Guangyao is clearly not yet accustomed to the transformation into new roles in life. These tasks he is not good at, he has handed over to his partners.
Chi Guangyao also always has a sense of crisis. At the end of November last year, after completing the entrepreneurship camp training at Miracle Makers, he spent half a month in Beijing looking for investment, and the seven or eight investors he met mostly declined, the financing market is like the recent winter.
Not long ago, he moved out of his family home, rented a place in the suburbs of Shanghai to save money, and commuted for more than an hour each day. The 5000 yuan salary he draws from the company each month barely covers living expenses, and now he tries to keep each food delivery order under 20 yuan, which is an increase of 5 yuan from the early tough times.
“If it weren’t for Dr. Qi (Lu Qi), we would have gone bankrupt by now.” Chi Guangyao said, now he no longer worries about not receiving money for products or not being able to pay salaries, and he still has enough money on hand, financing can be postponed.
Li Bojie was also troubled by financing in the early stages of entrepreneurship, but fortunately, a classmate introduced resources in capital, customers, and the market, and he and his partners focused on technology and products. However, the technological shift after starting a business also put a lot of pressure on Li Bojie.
Previously, he had little exposure to AI models and algorithms, so he spent the last six months catching up, reading two to three hundred related papers, and systematically understanding the basic principles, but there is still a distance from understanding papers to designing algorithms and writing papers.
Li Bojie also encountered many unexpected difficulties in the development process. He originally thought training speech would be simple, but the first demo produced was unclear in articulation. This made him realize that it is a comprehensive system engineering project that needs to consider the quality of the corpus, data cleaning, fine-tuning parameters, and other aspects, for which he and his team spent several months making dozens of optimizations.
In terms of products and business, Li Bojie said he also needs to catch up. However, Huawei taught him a lot of experience—how to understand customer needs, how to build technical competitiveness, how to demonstrate business value, which is very similar to entrepreneurship.
“My experience at Huawei taught me how to be more certain of success in entrepreneurship.” Currently, the AI Agent track is fiercely competitive, with severe homogenization. Li Bojie hopes to build a company with a technological moat.
Failure is nothing but starting over
Li Bojie chose to start a business, essentially because of his faith in AI.
He once said, if he continued to engage in network research at Huawei, he might end up climbing the Taihang Mountains; but if he pursued new technology, the goal might be Mount Everest. For Li Bojie, AGI (General Artificial Intelligence) is the Everest he wants to climb.
Li Bojie’s personal domain name includes 0 and 1, the only two numbers in binary. He said choosing this domain name was to devote himself to the AGI cause, to make a small contribution to silicon-based life based on 0 and 1.
“AGI will definitely come, and this is also a question I must ask when interviewing people.” Li Bojie said, entrepreneurship requires a lot of self-motivation, believing that AGI can be achieved, and it will definitely be better than humans, you need to have this belief to believe it can be done.
Li Bojie even looks forward to living with AI, believing it is already feasible. AI Agent will also become an extension of the real world, and may even transition from practical value to emotional value, AI lovers will no longer just be a story in science fiction movies.
As a firm technology neutralist, Li Bojie admires effective accelerationists like OpenAI CEO Sam Altman. “Technology used by kind people will be used for good, but we shouldn’t stop developing technology just because some people use it for evil.”
Chi Guangyao, who has thrown himself into the AI wave, is also a believer in AGI. However, he believes his stance is not important. “No matter which side I am on, it will not change this terrifying AI war.”
Chi Guangyao is not very interested in these abstract issues, but he has clear judgment. “Whether the capability becomes extremely strong, or the computational power consumption becomes extremely low, any of these situations will lead to killer applications for large models.”
In his view, large models are currently like the 3G era of the mobile internet, which may flourish after 4G comes out. For young entrepreneurs like him and Li Bojie, life’s new opportunities are just beginning.
Of course, Li Bojie has also anticipated that it is very likely that the business will fail due to the product or commercialization not being good enough.
“As long as there is technology, failure means nothing but starting over.” Li Bojie said, if this attempt fails, he will most likely continue to start a business.
Chi Guangyao feels that it will take another 2-4 years to know the outcome of his project. If in the future he is forced to do things he doesn’t want to do, leading to a significant decrease in the happiness weight of programming, then he might do something else.
“For example, returning to the path of arranging music, to become an independent musician.” Chi Guangyao said, no matter what form it takes, he will create the greatest value he can.