I taught small and medium-sized enterprises to use ChatGPT and I made $500,000 in 3 months

Under the influence of ChatGPT, the entire workplace environment is being reshaped. The author of this article has communicated with many PMs of large companies, self-employed entrepreneurs, programmers, etc., and compiled this "ChatGPT workplace survival guide", let's take a look.

"At the end of last year, Hippocampus formulated a strategy for implementing ID photos in 2023, and soon ChatGPT came. Midjourney does a much better job than human cutouts. It feels like an urgent strategic adjustment is needed."

"As a product manager who doesn't know much about technology, with the help of free ChatGPT, I was able to read and optimize a piece of original code in just a few hours. The content was reduced from more than 1,800 lines to more than 300 lines, and the execution time was reduced from more than 1 hour to 8 seconds, which increased the efficiency by 520 times."

"We provide combined training on AI series tools for small and medium-sized enterprises. One session costs about 50,000 yuan. Two people spend three months on it, and the estimated revenue is 500,000 yuan."

The excitement of ChatGPT is not only on the ice where big guys bring in funds and big companies take turns releasing big models to start a big fight; off the ice, the changes in people in the workplace are like threading a needle, stitching the AI ​​era and countless individuals more closely.

According to data from Boss Zhipin's "CSL Career Science Laboratory", 269 occupations have explicitly stated that they hope job seekers will use ChatGPT, mainly in computer technology, operations, product, design, and sales-related positions. Media and marketing-related positions have the highest requirements for proficiency in ChatGPT, including advertising copywriters, event planners, and marketing consultants.

At the same time, ChatGPT has also created new positions, such as prompt word engineers, language model trainers, ChatGPT optimizers, etc. The lab predicts that there will be 424 new occupations enabled by ChatGPT in the future, of which 7% will require applicants to be flexible or even proficient in the use of ChatGPT.

Under the sweep of ChatGPT, the entire workplace environment is being reshaped. Looking back at history, in the process of new things spreading downward, there are always two completely different voices, colliding and merging with each other. To this end, after communicating with many PMs of large companies, self-employed entrepreneurs, reporters, researchers, and programmers, we tried to restore a true "ChatGPT workplace survival guide".

01 As an AI service provider, I made 500,000 in 3 months

If you often use AI tools, hang out on Xiaohongshu and search for cutting-edge tutorials on the go, you must have seen "Xiaozheng".

On Xiaozheng's homepage, the use of different AI tools in specific scenarios is explained in great detail. Different from some training courses on the market that simply "sell tool tutorials", Xiaozheng's posts are mostly focused on scenarios and solving specific problems. Sometimes, several different tools may be used at the same time. This simple and direct way allowed Xiaozheng to quickly break through the circle, and within a week, his followers increased by nearly 2,000.

▲ Screenshots of homepage and post details

In fact, behind "Xiao Zheng" are two young people working in Internet giants: Acheng and Xiaozhi. They have the same experience in VC/PE and consulting industries, which makes them look at problems from a macro and framework perspective from top to bottom. Their unique judgment of the future and trained high execution ability have enabled them to keenly perceive the opportunities of generative AI since the second half of last year, and constantly think about what they can do.

"It's been a long time since anything new has made me so excited." Ah Cheng sighed.

This is not the first time that Cheng and Xiaozhi have started a business. When they were students, Xiaozhi started cross-border e-commerce with a few friends. When Shein and Shopify were not yet popular, Xiaozhi's team achieved a GMV of 2 million within a year through non-standard products.

After four years of online education going from its peak to a rapid decline, in 2022, Acheng began to look for a new direction and transformed into the Web3 industry, where he was responsible for the NFT and Gamefi projects. However, Web3 is still in its early stages, and there are less than 300,000 active Chinese-speaking NFT users in the market. Family and friends around him do not understand what he is doing, and he will inevitably feel a great sense of loneliness in the process. "But ChatGPT is different. This is a huge opportunity on the scale of mobile Internet."

During the conversation, we could always feel that these two people are "trendsetters" who have always been at the forefront of the times.

And what they are doing is not as simple as "posting tutorials and attracting fans".

"The C-end user tutorials and fans are only part of our display. In fact, we provide collective training for some companies in our own social circles to help small and medium-sized enterprises reduce costs and increase efficiency through AI tools. This is one of the core business models in the future." Xiaozhi said.

According to "Zi Quadrant", at present, "Xiaozheng" provides training to small and medium-sized enterprises for 1 to 2 days and charges about 50,000 yuan. It also includes subsequent long-term services and after-sales work.

This is not a result of a "snap guess", but an "entrepreneurial opportunity for ordinary people" discovered by Xiaozhi and Acheng after carefully analyzing the current entrepreneurial opportunities in generative AI.

Xiaozhi believes that there are four major entrepreneurial opportunities at present. The first is that large companies are competing for large models and AI Infra, and they are scrambling to become China's first ChatGPT. This is an opportunity for BATZ and Wang Huiwen. The second is the industry and vertical fine-tuning of general large models, and the creation of professional models for specific scenarios. All walks of life have also begun to move. The third is the application ecosystem. It is unknown which sub-segment will produce the next unicorn.

Ok, so far, it is basically consistent with the judgment of "Self-quadrant" in "ChatGPT Entrepreneurship Revelation" . However, these three types of entrepreneurial opportunities all have a problem: the threshold for entrepreneurship is high, the general large model > industry-specific model > application ecology, which basically isolates ordinary entrepreneurs thousands of miles away.

Acheng and Xiaozhi thought of the fourth option, which not only broke the high threshold of AI entrepreneurship, but also found irreplaceable value.

That is to be an AI service provider. Simply put, it is to become a builder of an inclusive ecosystem, break the high threshold of large models and the barriers between small and medium-sized enterprises, and enable more companies to reduce costs and increase efficiency faster through AI capabilities.

In the past, especially in the Chinese market, humans play a role in certain links that cannot be replaced by machines. For example, no matter how smart the voice robot is, we will still habitually press "0" to switch to manual service. This is especially true in the digitalization of small and medium-sized enterprises. Small and medium-sized enterprises have always been in the deep waters of digitalization. The core reason is that the digitalization levels of a large number of small and medium-sized enterprises are uneven, which requires more time, patience and methodology to irrigate.

Just like many service providers in the digital age, service providers are also needed in the AI ​​era.

"We believe that if large companies provide pure end-to-end services, they will not be able to form a large-scale business ecosystem. A large number of service providers will be needed to undertake the services. Microsoft's office suite, PPT and Excel product forms have been solidified for a long time, and still require professional training to get started. This is the opportunity we see." Xiaozhi said.

A Cheng introduced: "At present, in the small and medium-sized e-commerce industry, there are immediate effects after training, including in e-commerce marketing, store decoration, picture material production, live streaming and other links, AI tools can be used to quickly improve efficiency. The e-commerce industry is the first stop for our application."

At present, "Xiao Zheng" is "walking on two legs". On the one hand, it is refining the granularity of C-end tutorials and spreading them to more users through courses. On the other hand, it is accumulating know-how and solutions from all walks of life, and opening a new door for enterprise intelligence through services.

▲Image source: Xiaozheng Knowledge Planet homepage

“In the next five to ten years, if we compare the business world to mountains and rivers visible to the naked eye, I hope to be able to stand on a relatively high mountain, see the trends of the mountains and rivers below, and choose to invest in one of them.”

The wave of generative AI may be The Choosen One.

02 I use ChatGPT to break the anxiety of working at the age of 35

"Our company has established an AI learning group, and all employees hold regular discussions and learning, especially the development team. Model-oriented development is a definite trend in the future," Winnie told "Self-Quadrant".

I think Winnie’s company might be one of the seed clients of Xiaozheng training.

As a business analyst, Winnie's daily work is macro market analysis, including competitive product analysis, user usage, and data modeling in her company's track. In this process, she often needs the help of ChatGPT.

"When doing market analysis, we often need to aggregate and summarize a large amount of public information, which involves crawlers. For example, you need to tell ChatGPT the source code, data structure, whether it is dynamic, whether it is encrypted, and other information of the website, and it will automatically help you grab all the information you need."

"In the past, this work was usually done by the company's developers, but the developers may not have the right time, so cooperation is needed. Now with ChatGPT, I can do it myself. If the content of a small website is not large, it only takes a few minutes to crawl it. On average, crawling 100,000 pages per hour is possible. In general, the work that used to take half a day can now be done in 10 minutes."

▲Picture source: Winnie photographed the crawling process

So, if the basic work of programmers is done by ChatGPT, will programmers panic?

For this purpose, "Self-quadrant" specially asked several senior programmers, and unexpectedly the results were surprisingly consistent: not only are they not panicking, they are even very much looking forward to it.

In fact, almost 70% of programmers' work is repetitive code structure, 20% is secondary innovation, and 10% is absolute innovation. If 70% of repetitive work can be replaced by ChatGPT, they will have more time to invest in the remaining 30%.

"In the future, there will be many applications developed based on AI natively, and programmers should also carry out self-revolution, first use the tools to improve efficiency, and then innovate and create." A 35-year-old programmer told "Self Quadrant".

"However, even if the efficiency is improved, your work will not be reduced from 8 hours to 6 hours. It will only increase from completing 8 tasks in 8 hours to completing 16 tasks in 8 hours. Work efficiency will be improved, but working hours will not be reduced." Winnie said.

When asked, "Do you think you will be eliminated if you don't know how to use ChatGPT?" Winnie said with certainty: Yes. Even though the company does not have a clear "blood transfusion order" now, the judgment of GPT generation results and the issuance of instructions are more based on people's understanding of the market and years of experience. This is another form of test.

In this big test, it is difficult to fish in troubled waters.

In addition, Winnie believes that ChatGPT can break the workplace anxiety and layoff curse of 35-year-olds.

"The basic work that GPT replaces happens to be the value played by some newbies and interns in the workplace. People aged 35 can be replaced because young people are more studious, more competitive, and more willing to work overtime. For programmers, 70% of the code is written in the same way, and the gap between new and old employees is not big, so companies tend to use new employees. But now GPT is more cost-effective. What is being competed is experience and decision-making ability. The point is that how much role GPT can play depends entirely on the ceiling of people, which is an irreplaceable value."

▲Image source: Das Kapital

In this social atmosphere where "35-year-old workplace anxiety" is being discussed everywhere recently, Winnie's views are refreshing. But if you can rely on experience to get a job at the age of 35, then the pressure will naturally be on the young people. In addition to the lower cost-effectiveness, what other value do the newbies have? This is something that needs to be considered.

03 Engage in AI innovation to enable every product manager to become a "Zhang Xiaolong"

Akai, born in 1992, is already at the end of the group of 35-year-old working professionals.

But he showed the exploration and curiosity of a young man.

Since November 2022, Akai has been almost immersed in AI tools and has integrated them into every process of his work as a product manager.

So far, AI tools have largely replaced the repetitive aspects of his job.

For example, when writing standardized requirements, there is often only one sentence of requirements, but it needs to be expanded in multiple dimensions to align with development. Akai uses Notion AI to convert a sentence into a product requirements document, and then simply modify the requirements document, which can reduce the original work of half an hour to 40 minutes to 5-10 minutes.

For example, when a product manager proposes a requirement, he needs to conduct research and evaluation. Why is this requirement proposed? How is the user market research? These can all be done through the large model integration platform POE. Akai only needs to propose a problem and turn it into a solution, which can basically cover 99% of the problems.

▲Image source: POE official website

In addition, there are some more tedious tasks. Akai's company is mainly engaged in overseas banking and financial services, which often involve finance and numbers, and mistakes will occur. For example, extracting bank account numbers from eBay's database requires a certain amount of experience and 100% accuracy. After packaging the data and "throwing" it to ChatGPT, it can automatically summarize and capture the bank account numbers, quickly and accurately.

"Under the current trend, almost all jobs can be replaced by GPT, but what it cannot replace is the subjective emotions in the human brain. Product manager is actually a very emotional role, and there are many classic problems and very interesting work, but because we are trapped in a programmatic model, we rarely have the energy to take care of both."

In fact, every product manager aspires to be "Zhang Xiaolong" and hopes to be able to think down-to-earth about the "Ten Methodologies of Product Managers", but there is only one Zhang Xiaolong, and there are not many people who can slow down in this fast-paced era.

Akai also admitted that the company's use of AI tools is very subtle. The boss encourages us to make some different attempts and explorations, but some older employees are quite resistant to this.

“They cannot resist the arrival of the times, but they can try their best to slow down the speed at which the times change individuals.”

Under the whole wave of AI2.0, some people took the opportunity to find the right niche to make money, some transformed themselves into excellent employees, and some became the gems of the era, just like "Nokia", hard and stubbornly remaining in that era forever. When technology becomes universal, we find that the Matthew effect in the industry is also intensifying.

How to respond proactively is the true survival guide for people in the workplace under ChatGPT.

*The pictures in this article are from the Internet

Author: Cheng Xin, Editor: Yun Tianming

Source public account: Zixiangxian (ID: zixiangxian), between the squares, there is a quadrant. Care about science and technology, economy, humanities, and life.