Teaching others how to use AI has earned me a massive fortune.
AI is making more and more skills cheaper, but "teaching people how to use AI" has instead become an increasingly specific business.
Its price range is extremely wide. On Xianyu, AI tutorials priced at 9.9 yuan can sell over 10,000 copies in half a year; some people gather thousands of students through a paid community model costing hundreds of yuan; others sell one-on-one consultations at 2,000 yuan per hour, charge 20,000 yuan per month for corporate consulting, and even charge 300,000 yuan for a single in-house corporate training session.
On the surface, this seems to be the kind of business most easily replaced by AI itself. Tools are becoming more and more user-friendly, free tutorials are increasingly available, and people can even ask AI directly if they do not know how to operate. But what people who actually pay for these services buy is often not "how to register an account" or "how to write prompts".
Some people spend 3,000 yuan on courses because the teacher has clients and commercial orders, and they can take on projects right after finishing the study; some are willing to pay nearly 20,000 yuan to truly integrate AI into their work; lawyers learn AI not just for fear of being replaced, but also because issues such as client privacy, local deployment, and work processes are hard to find direct answers in general tutorials.
These trainers are not all "teachers" in the traditional sense. Some used to be screenwriters, lawyers, or psychology lecturers, and some just started using AI half a year earlier than their peers. Their common advantage is often not that they understand models better, but that they figured out earlier what problems AI can solve in various industries, and turned these experiences into sellable products earlier.
As a result, a somewhat contradictory market has emerged: while AI eliminates information gaps, it also creates new ones; while it lowers the learning threshold, it makes more people worry that they cannot keep up.
Is this business essentially selling knowledge, experience, resources, or anxiety?
We talked to several people who are making money by "teaching others to use AI". Some regard training as a long-term education business, while others are well aware that their profitable window only lasts for the next one or two years.
01.
A course costs 3,000 yuan, but I do not plan to make AI training a long-term business
Wang Tian | Founder of Mengjing Jingming, Beijing
I used to work at an internet company, and my total annual income including base salary, performance bonus, year-end bonus and stock options was quite substantial. Later, I left my job to start my own business for various reasons.
I have a background in screenwriting, I have always loved writing, and I have won several awards in writing competitions. Back when I was still employed, I always wondered if I could create my own content one day.
After starting my business, I first invested in a short drama, but ran into a lot of pitfalls. After that experience, I became wary of shooting live-action dramas, because there were too many links to coordinate and the work was extremely tiring.
It was just at that time that AI comic dramas became popular. They do not require organizing live shoots, and one person can complete most of the work. With my foundation in screenwriting and writing, if I learned AI production, the ideas in my head could be directly turned into videos.
At the beginning, I explored everything on my own, and spent hundreds of thousands of yuan in total. Later, more and more people came to me to cooperate on projects. Now our studio mainly serves enterprise clients, producing brand stories and creative advertisements, and also takes on scriptwriting and planning projects. We basically cover the whole process from early planning, script writing, AI video production, to traffic delivery and agency operation. The core team only has 3 to 5 people, and we have more than a dozen long-term part-time collaborators.
For a period of time, we had more orders than our team could handle, so I thought, instead of constantly recruiting new employees, it is better to teach some people the skills. After they master the skills, they can not only take their own orders, but also help us share the workload of our projects. That is why our original intention of launching training was not simply selling courses, but hoping that students can actually take on commercial orders after finishing learning.
Now we have three courses: AI Video Production, Master Screenwriter, and Master Operator, each priced at 3,000 yuan. The best-selling one is AI Video Production.
I did not spend a lot of time specially designing the courses. Most of the content is the problems we repeatedly encountered in previous projects, which are already stored in my mind. We stream the courses live via Tencent Meeting, with about 10 students per class. We do not do paid traffic delivery, and student recruitment mainly relies on referrals from friends.
At present, we have less than 20 students, and their backgrounds vary greatly. Some are looking for jobs, some are stay-at-home moms who want to do side hustles, and some are recent graduates. One of the new graduates even borrowed money from his family to pay the tuition. Some others are already working in branding and content creation, and they just want to figure out how to integrate AI into their business.
They are willing to learn not just because of the courses themselves. Many of them value the fact that we have clients and orders in hand, and there is the possibility of continued cooperation after they finish the courses.
Now whenever there are suitable commercial orders, we will distribute them to our students. Almost all of them have earned some money, and some have even nearly earned back their tuition fees.
But I do not plan to make training a very large-scale business.
I may run a few more sessions first, and continue if there is demand. Whether I will continue to teach in the future also depends on what new things I can learn. If I find a new direction that is really useful, I may come back and share it with students again.
AI is evolving faster and faster, but I am not that anxious now.
Because we entered the track early, we already have clients, resources, and businesses that we can actually deliver. For me, what is more important than "whether the next model will make the current technology obsolete" is: what new capabilities it has, whether this capability can be turned into a project, and whether it can help clients create new value.
02.
Use AI to make courses, then use courses to teach people to use AI
Zhang Yuming | AI Application Consultant, Corporate Trainer, Beijing
I do not specialize in AI training. My main business is AI application, including medical AI, corporate consulting, and short video creation. Giving lectures to teach others to use AI only takes up about one third of my time.
This matter was not even related to "selling courses" at the very beginning.
I used to teach psychology and philosophy. After my classes, people did not ask me about psychology first, but asked me: why your PPT is so beautiful? Why can you make videos so fast? I told them I used AI. Later, more and more people came to consult me, and some even directly asked me to teach them artificial intelligence. That was how I started charging for the courses.
Now I have 33 major courses in total, which can be split into more than 400 individual lessons. In the past, this was almost impossible to complete with manpower, but now all the courses can be launched within about a month. I no longer record each lesson in person. My own image appears in the opening and closing clips, and the voice, text and pictures in the rest of the content are basically generated by AI. The teaching plans are made in the same way: I set the topic and outline, AI generates the first draft, I am responsible for reviewing it, and modify the parts that are not correct.
Our team uses Claude, ChatGPT and Gemini, and the API cost alone is 3,000 to 5,000 US dollars per month.
"Teaching AI" currently has three sources of income. The most stable one is online courses, which are mainly offered on WeChat Channels. The personal course has 4 sessions per month, priced at 19,800 yuan; we also have crash courses for people with solid foundations priced at several thousand yuan. Offline personal training can generate tens of thousands of yuan per month, but the income is not stable. Corporate training has the largest single transaction value, and a weekend course can be priced at 300,000 yuan, but we need to participate in bidding, so the income is also not very stable.
I found that people's willingness to pay for AI is not necessarily related to the size of the company or their capital strength. I have seen companies that are reluctant to pay tens of thousands of yuan for training, and people from well-off families who keep asking "what is the use of learning this" and finally request a refund. The people who are truly willing to pay and learn fast often have one thing in common: they are curious about new things, and willing to try them out first.
Doctors and lawyers make up the largest proportion of my students.
Lawyers used to ask their assistants to write legal documents, and they had to revise the drafts after that. The revision process was more troublesome than writing, and they also had to pay the assistants. Now they use AI to generate the drafts directly and then optimize them themselves. Doctors use AI to organize a large amount of patient data first, so that the information is sorted out clearly. There is also a student in Hangzhou who used to develop mini-programs. A single project would take 3 months, and clients would keep making revisions, which often dragged the project to half a year. After using AI, the number of projects he can take on at the same time has become 5 to 6 times of the previous level.
Of course, some people lost contact with me after finishing the courses, and I do not know whether they have achieved results or made money.
Many people ask me, there are so many free tutorials online, how long can paid courses last?
I am not worried. The internet has been around for decades, and to this day, there are still people who pay to learn about Douyin and self-media. Free content will always exist, and paid education will never disappear.
03.
Teaching lawyers to use AI earns me more than working as a lawyer myself
Lin Zhou | Legal Consultant, Lecturer, Nanjing
I used to be a lawyer. Later, introduced by a friend, I joined a legal information software company, where I was responsible for providing regular training and digital services to lawyers associations across the country. In those years, I dealt with lawyers every day, gave lectures, made plans, and visited various associations, so I got very familiar with this industry.
Around August and September 2024, I clearly felt that lawyers were getting more and more anxious about AI.
Among all the people I have come into contact with, lawyers are the group that worry the most about being replaced by AI. Everyone wants to know what exactly AI can do, and wants to learn to use it, but at that time, there were very few trainings specially designed for lawyers. I realized there was market demand here, so I left the company and started my own business. In the early stage, I still got clients through associations, and later I ran the business completely on my own.
Now I mainly offer two types of courses. Offline small classes have less than 20 students per session, and the revenue for each class is usually 30,000 to 50,000 yuan, but the income fluctuates greatly; online courses are held three times a month on a fixed schedule, with an average monthly income of 20,000 to 30,000 yuan, which is relatively stable. Occasionally I also give large lectures for lawyers associations, but those sessions basically do not make money.
Lawyers themselves have strong payment capacity. I think my pricing is not high. If I really want to expand this business, I should be able to earn more. But now the total income from all types of courses is much higher than when I worked as a lawyer, and the work is also less tiring.
I do not have a technical background. If we really compare how much AI knowledge we master, I may not be better than many of my peers. But after running this business, I am more and more certain of one thing: You do not have to be the person who knows AI best. What is really valuable is that you know exactly what people in a certain industry need.
Many people will ask, why do you still need to pay to learn to use AI? Is not it enough to just ask AI directly?
If you just want to chat or look up information in daily life, you can indeed explore it by yourself. But when you really integrate AI into lawyers' work, the problems will become much more complicated.
The most typical one is privacy. The materials lawyers handle involve a large amount of client information, and not all content can be directly imported into public models. Some people choose local deployment for data security, but the capability of local models may be relatively weaker. What materials can be uploaded, what tools can be used, and how to balance privacy and efficiency, these problems cannot be solved by just learning a few prompts.
So for me, what I sell is not just AI knowledge, but a combination of information gap and industry experience. My peers need someone to translate the tools into specific working methods, and they are willing to pay for this kind of experience.
Of course, I also know that not all the money people pay is for knowledge. Part of it is for experience, and part of it is for relieving anxiety. Practical problems in using AI do exist, and the fear of being replaced by AI is also real.
I am not sure how long this kind of information gap can last.
The skills that need to be specially taught today may be mastered by everyone in a year or two. When AI truly becomes a daily tool, many services will naturally no longer need people to pay for learning.
So I do not plan to take AI training as a long-term business. As long as there are people who need it and are willing to pay for it now, I will keep running it.
When this wave of popularity passes, there should be other opportunities, and I will decide what to do then.
04.
After turning down a 480,000-yuan "guaranteed employment" order, I decided to run my own training business
Yang Bo | AIGC Director, Xuzhou
I have a background in news video production, and I have also been involved in short drama projects before. Around September 2025, I started to use AI in real work. At that time, the models were still not very mature, but I already vaguely felt that AI would change the way of video production.
I have a habit of learning: I cannot just watch tutorials, I have to use what I learn to do real work. So while I was testing various new tools, I took on small orders of copywriting and video production on Taobao. My clients included teachers, civil servants, and professional film and television practitioners. A director I knew from Hengdian once asked me to make a 10-minute video with interactive scenes between elves and real people. Since traditional shooting was very difficult to achieve that effect, we tried to complete it with AI.
At the beginning, AI was just my side hustle, and I earned 3,000 to 4,000 yuan per month. Later, there were more and more orders, and I could not handle all of them alone, so I built a group to teach others for free. There were freelancers and stay-at-home moms in the group, some of whom could not