.png)
3 months ago
2
On a recent Monday evening, several Chinese participants joined a livestream on Zhihu, a Q&A platform similar to Quora, and were greeted by an instructor who said her name was Eva. She promised to teach them the skills they needed to work in the artificial intelligence industry.
Eva’s two-hour session covered topics such as “next token prediction” and “attention is all you need.” She also showed a chart detailing the salary range of AI engineers at different levels: The best ones are paid more than 2 million yuan ($279,000) a year, nearly 10 times the median income of software developers in China. In a related WeChat group, participants including software developers and a drone engineer said they were keen to find jobs working on large language models.
“AI has presented a great opportunity for ordinary people to change our lives,” Eva said in her lecture. “As long as you have good technical skills, you can get a job.” At the end of the trial session, she urged participants to pay 4,980 yuan ($695) for a 70-hour course.
As Chinese tech companies from Alibaba to ByteDance race to recruit people who can build cutting-edge AI models and develop novel applications, tech workers and computer science students in the country are turning to books, online classes, tutorials, and consultations to acquire skills they believe will help launch a career in AI.
China has a multibillion-dollar online learning market. Education platforms and influencers have previously touted courses on livestreaming, e-commerce, and the metaverse, capitalizing on an appetite for understanding new technologies and making money from them. Classes on AI have mushroomed since the launch of DeepSeek earlier this year.
Initially, these courses taught people how to use AI tools such as ChatGPT and Midjourney, which generate images based on text prompts. More recent offerings have been about building AI-powered products and securing well-paid jobs, with slogans like “Become an AI model engineer in 100 days,” and “AI models: a historic opportunity you must not miss.” They promise to teach everything from the history of LLMs to advanced techniques such as retrieval-augmented generation, which helps improve LLM outputs.
People may not be buying luxury goods, but they are willing to pay for these classes.
The AI boom has generated jobs not only for elite researchers but also for those working on small, experimental applications, and AI trainers, Lin Zhang, an associate professor at the University of New Hampshire who studies China’s digital economy, told Rest of World.
“No matter how useful or not these courses are, they address people’s anxiety over economic uncertainties,” she said. “People may not be buying luxury goods, but they are willing to pay for these classes.”
The rise of AI and the launch of DeepSeek have reenergized China’s trillion-dollar tech industry, which has grappled with widespread layoffs and greater government scrutiny in recent years amid slower economic growth. Search engine giant Baidu said earlier this year that 87% of its 3,000 summer internships would be AI-related. ByteDance, the parent company of TikTok, has a new “top seed” initiative, promising in a recruitment ad to provide AI researchers with sufficient data and computing power.
China’s need for workers skilled in AI will grow sixfold to reach 6 million by 2030, McKinsey estimated in 2023. A recent report by Chinese job platform Liepin showed that the number of AI-related job openings has increased by more than 36% in the first half of 2025.
The online courses are also filling a gap in formal education. Although universities in China have rushed to add AI majors, recent graduates from computer science programs told Rest of World they generally do not get hands-on training in cutting-edge LLM techniques. Even those enrolled in AI programs often need to supplement their education independently, they said. On social network Xiaohongshu, users frequently trade advice about breaking into the AI industry, sharing book lists and video tutorials, and swapping interview strategies.
Frankie Chen, an AI major at a university in Tianjin, said the courses in machine learning he took did little to prepare him to work with LLMs. To land an internship in AI, he spent two months reading books; practicing on LeetCode, a platform that helps users prepare for coding interviews; and purchasing material from influencers on Zhishi Xingqiu, a knowledge-sharing platform similar to Patreon. Chen didn’t sign up for any online courses, as a comprehensive curriculum cannot be offered online, he told Rest of World.
Chen recently found an internship working on an AI customer service tool. With the internship on his resume, he plans to apply for full-time jobs at China’s top internet companies.
“Traditional front-end and back-end engineers will be replaced by LLMs,” he said. “I want to plan for the future.”
On social network Xiaohongshu, users trade advice about breaking into the AI industry, sharing book lists and video tutorials, and interview strategies.
Chen is right to be cautious about online AI courses. There is little regulatory oversight, and there are several complaints about their quality and their claims. On consumer complaint platform Heimao Tousu, former students have demanded refunds, accusing some providers of poor instruction and exaggerated promises of securing jobs. Self-styled AI guru Li Yizhou’s tutorials on generative AI were removed from social media platforms WeChat and Douyin last year after customers complained they lacked substance and were filled with advertisements, according to media reports.
Other students think they have no choice but to sign up. One computer science graduate from Shandong province paid 4,000 yuan ($557) for a two-month AI engineering course online. During the live sessions, the instructor introduced concepts such as AI agents and retrieval-augmented generation, but the students didn’t gain any practical skills, he told Rest of World.
“The classes were too superficial,” he said, asking not to be named because he was worried that speaking to the media may hurt his job prospects. He turned to other free online resources, which he said were more useful, and helped him land a job building LLM applications at a small Beijing company.
Tech companies are also joining the AI education push, with courses that promote their own AI tools and cloud services. Tencent Cloud offers a course to Chinese universities on LLM deployment, using its own Hunyuan model. Nvidia, which sells chips used to power AI models in China, has both free and paid courses on deep learning and AI agents.
There are no estimates on how many people have signed up for AI courses online, but on video-sharing site Bilibili, some tutorials for LLM fine-tuning and deployment have more than 1 million views.
Despite the heightened interest, top-paying AI jobs only go to Ph.D. graduates from prestigious universities, or people with experience on well-known models, Berry Liu, founder of Shenzhen-based tech headhunting firm Zzcareer, told Rest of World. Self-taught developers can meet the demand for jobs in specialized AI applications, she said, as AI tools become deeply integrated across industries.
On recruitment sites in China, telecom, pharmaceutical, and machinery companies all have jobs related to AI models. The fear of falling behind has gripped veteran tech workers, too, as AI gets better at writing code.
One programmer at a leading tech company in Beijing, who requested anonymity because he wasn’t authorized to speak to the media, told Rest of World every department at his company was under pressure to incorporate AI.
In his downtime, the 30-year-old reads books on AI agents and hones his skills by building a chatbot to handle customer service inquiries for massage shops, he said. He plans to apply for LLM-focused positions — and, as a side project, create and sell AI tutorials to other developers.
“My own job is also at risk,” he said. “When the next round of layoffs happen, you will have more options if you have been learning [AI].”
