Summary

姜Dora saw leaving a major company as a career trade-off with a survivable downside and unknown upside. She went from a translator earning RMB5K a month, with just over RMB3K in take-home pay, to a content manager at 25–26 earning RMB20K a month, and eventually to making roughly a month’s salary from a single video; what finally gave her the courage to leave was reverse-engineering her roughly RMB500K annual income: could the opportunity cost buy a channel with 1M followers, 100 career interviews, and a new skill set—not simply “freedom”?
The creator economy’s attractive income comes from concentrating every function an organization once distributed onto one person. Her current team has 3 people; she handles weekly publishing, topic selection, product, sales, tax, legal, finance, and user growth. Her correction to the “super individual” narrative is blunt: “It’s not that everyone wants to be a super individual; everyone just doesn’t want to go to work”—and more and more people “have nowhere to work.”
For 姜Dora, AI amplifies less a person’s existing knowledge than their ability to ask questions, give quality feedback, and iterate through sustained dialogue. She calls AI “a question-asking machine” and “more functional companionship”: she asks it to pose one question at a time, spent 2–3 hours clarifying a move, and uses it to edit drafts, accompany her reading, and analyze risks; but the shallower the user’s thinking about a problem, the more generic the AI’s answer will be. “A lot of people simply don’t know how to describe their problem to AI.”
The key to working with AI is not handing everything to the model, but recognizing the boundaries of human and model capability. When editing interview transcripts, AI can accurately compress rational arguments but may delete details carrying emotion and narrative; she therefore asks not only “What can you do?” but also “How should we divide the work between us?” and recalibrates through new samples. Access to outside knowledge does not mean organizational, managerial, and judgment capabilities have been internalized.
A small team is not necessarily an incomplete company waiting to scale; it may simply be the founder’s most comfortable stage for now. 姜Dora tried hiring, building SOPs, and removing herself from production, but ultimately admitted that her strongest desire remains creating personally: “This isn’t a problem of managing other people. It’s a problem of managing myself.” She would rather be a “not very cool entrepreneur” for the next 2–3 years than sacrifice the content’s tone too early for institutionalization.
The so-called “super individual” looks more like an early-stage entrepreneur than a stable end-state organizational form. Independently delivering products to the market, acquiring customers, and expanding profits is only the first step outside the corporate chain; stamina, expressive style, and interests all change, and eventually people still have to deal with talent, management, product iteration, and scale. Her judgment is direct: “I can’t still be making videos at 60, can I?”
Liangzhu offers more than cheap rent: it provides lifestyle infrastructure and a low-friction social network suited to freelancers. 姜Dora rents a 140-square-meter home there for about RMB5K, versus roughly RMB12K for 100 square meters in Beijing; resident self-governance, interest groups, home salons, and PGC/UGC activities mean that even with less socializing, “you can always find someone if you want to go out,” reducing social anxiety. What she sees is a place for recovery from career trauma, non-standard livelihoods, and a life in which “the environment is more enclosed, but the inner world is more open.”
An economic downturn changes how people experience effort and content, but it does not erase the human instinct to work. She uses an elevator to explain the collective mood: effort gets feedback more easily on the way up, while in a downturn “you can keep jumping upward, but you’ll still be going down,” making inward reflection, deep relationships, and humanistic content more appealing; but she does not equate mocking effort with genuinely giving up, ending with her mother’s transformation of a garbage pile into a vegetable garden: “Make a little effort, and make whatever improvement is within your power.”

Deep dive

1. The Dream of Translation Quickly Lost Its Shine in Her First Job

姜Dora studied English interpreting at university and became more determined to pursue translation after spending her junior year on exchange in the United States. In the English-major culture of the time, the job was considered “比较fancy,” while simultaneous interpreting was seen as the top of the professional pyramid; she deliberately avoided the English-teaching path most classmates chose.
After graduation, she joined a small IT company in Beijing and left after just 3 months. She was the only woman in the company; the bathroom had no gender separation and was effectively a men’s room, while her boss insulted her with phrases such as “pig brain” and “your thinking is completely point-like.” After being berated one day, she saw urine stains in the bathroom and thought, “I really can’t stand this place anymore.”
Around the same time, iFlytek demonstrated machine simultaneous interpreting, triggering anxiety among English majors about replacement. She knew the demo might have been given advance access to the conference materials and could have been staged as a “show match,” but acknowledged that human interpreters also need preparation beforehand. Low pay, the realities of the job, a psychologically abusive boss, and technological disruption combined to make her rethink her career rather than continue believing in her original image of professional prestige.

2. A RMB5K Salary Was the First Time She Chose a Job by Industry Opportunity Rather Than Identity

Her first job paid RMB5K a month, leaving just over RMB3K after social insurance and housing fund deductions—too little for a normal life in Beijing. 姜Dora drew her first simple conclusion: “Where the money is, that’s where I should go,” because a lack of money can completely reshape a young person’s thinking and inner world.
She tried to combine education and content but did not want to become a conventional English teacher, so she moved into online education. She first joined the studio of an English-education blogger, producing English courses and related content at scale during the Weibo era.
She later worked at Xinsixiang, Ximalaya, and ByteDance across graphics, audio, and short video. She kept moving around content even as the medium and job title changed. What looked like a series of risky career changes had an underlying continuity: “It felt like I was still doing one or two essentially similar things.”

3. Moving from Content Production to the Platform View Created the Hidden Asset Behind Her Later Channel

By around 25–26, she had become a manager at Ximalaya and was earning RMB20K a month. But short video was growing rapidly and being discussed everywhere, while she “didn’t know what was going on,” and that disconnect hurt more than being demoted.
She eventually chose to join ByteDance even though the role was lower-level and the pay was higher. The decision was not based on a forecast that K12 would run into trouble; she simply “really wanted to participate and understand” the new medium. She left Ximalaya shortly before K12 later disappeared, prompting others to call her prescient. She explicitly rejected that framing: “I wasn’t.”
Working at a platform showed her how mechanisms move the content pool, what kinds of content are more likely to scale, and how technology keeps lowering the barriers to creation. When her own videos later grew quickly, it was not just a matter of expressive ability; her producer experience and understanding of platform distribution finally converged.

4. The Internet Compressed Industry Turnover to Two or Even Five Times the Speed

姜Dora describes internet workers as living with chronically elevated cortisol. Leaders ask about five-year plans, but a company may disappear in 3 years; one colleague changed bosses 7 times in 2 years at the same company, while projects often vanished before anyone had fully understood them.
She witnessed online-education projects die suddenly after mechanism changes, the K12 environment flip, and classmates get laid off, but insists none of this resulted from seeing the cycle in advance. She mainly acted on interest and instinct: “Every time I took a step and turned around, the road behind me was already gone. It had become a cliff.”
She believes AI and a difficult employment environment are simply spreading to more professions the discontinuity internet workers have long known. Jobs people assumed they could do for years suddenly stop working, forcing them to ask where the future is.

5. Big-Tech Resources Can Hide Pseudo-Demand—and Push Creativity Outside the Job

At entrepreneurial companies, she once went as long as 6 months without a break, stretching her bandwidth to the limit while still being unable to truly rest. Later, at a peripheral project inside a major company, the top-level leader was never hired, and the project eventually stopped being responsible for generating profit. Her work was limited to changing cover images and filling out forms, leading her to conclude that “the capabilities big companies actually require at the grassroots level are much lower.”
Offline user research showed that people did not actually like using the product, yet because it was attached to Douyin’s DAU, her own numbers could still be made to look impressive. She described the state as “pretending to work and making arguments for pseudo-questions.” Once she clarified the SOP, she could finish a week’s work in 1 day and was left pretending to work for 1 or 2 hours each day.
She came to understand that once a large company gets 1 or 2 products right and accumulates enough resources, it can accommodate some people and projects that create no value for a long time. That tolerance can also incubate new opportunities, but for someone who needs real feedback, resources cannot substitute for meaning.
When she previously made knowledge-based paid courses, each lesson ran about 15 minutes, close to the length of a mid-form video on Bilibili. With nowhere to put that creativity inside the job, she turned to personal videos, quickly attracted commercial deals, and reached deal income close to a month’s salary. That became the practical trigger for leaving.

6. Learning “Corporate Bullshit” Confirmed That She Could Not Keep Performing

Her colleagues included people with Peking University undergraduate and graduate degrees and Harvard or Columbia backgrounds. They could see that the project had limited meaning but continue playing the game while managing their emotions. 姜Dora also tried redirecting her attention to lunch, overtime compensation, and presentation language, imitating the organization’s survival tactics.
During one meeting, her mind was completely elsewhere when a colleague suddenly called on her. She responded with lines such as “the key to this project is finding the key” and “the key first step is to execute the first step,” and somehow passed without incident.
Instead of feeling pleased at having mastered the technique, she returned to her desk and asked: “Why am I treating myself this way? Why do I have to become such a fake person just because the company pays me tens of thousands of yuan every month?” That discomfort ultimately carried more decision-making weight than the income.

7. Quitting Without a Job Was Not Giving Up RMB500K—it Was Deciding What to Buy with It

The first shock of leaving was not the loss of income but the disappearance of social identity. It was only when the 14th and 15th of each month came and no one was paying her social insurance and housing fund, and when conversations with former classmates revealed that their shared bosses, projects, and topics had gradually become history, that she realized she was saying goodbye to her old world.
She judged that the worst case would not destroy her career. Even if she returned to work after 1 year, time off, a personal channel, and new experience could still add up to an advantage. She initially called the decision a 1-year gap; if things went badly, she could return after 6 months. Once the actual situation proved good, she no longer wanted to work for anyone.
She recommends setting aside “fuck you money” equal to 1.5–2 times annual living expenses, rather than exactly 1 year’s worth: “Once you’ve spent 50% of that money, you start to panic.” At the same time, she emphasizes that her old career remained accessible; having a safety net is what makes aggressive assumptions possible.
She valued her annual income at a major company, year-end bonus, and various company-provided benefits at roughly RMB500K, then asked what that money could buy: a channel with 1M followers, 100 career interviews, and a career-planner certification. Once opportunity cost became a shopping list, reluctance became an assessable exchange.

8. AI’s Biggest Value Was Not Answers, but Extending Thought from 3 Meters to 10

姜Dora believes people will use AI to work through the 1 or 2 questions they care about most. She has kept a diary since childhood and has long studied herself, so “self-exploration” was not a new demand created by AI; it simply finally found a higher-frequency medium with better feedback.
She says the most important part of journaling is not reviewing it later, but how deeply the act of writing makes you think in the moment. “If I write by myself, maybe I get to 3 meters.” Sustained dialogue with an intelligent counterpart might take her to 5, 6, or even 10 meters, making it easier for her cognition to stabilize at the new depth.
Her boyfriend can explore concepts with her layer by layer, but is limited by his personal knowledge structure. AI is like someone with broad knowledge, instant replies, and strong reasoning once given an instruction. That is why she stresses: “Compared with giving answers, its greater value is having someone continuously ask me questions.”

9. Precisely Defining the Questioner’s Role Can Turn Vague Indecision into Recognizable Desire

When considering a move from Beijing to Hangzhou, she did not ask AI which city was better. She asked it to play the role of a psychotherapist and help uncover where she “really wanted to go,” because the core of therapy and coaching is not inserting an answer from outside but clarifying the answer the person already possesses.
She limited the role to an empathetic adult woman aged 35–45 and required: “Ask me only one question at a time” and “Do not tell me the answer.” After 2–3 hours of continuous questioning, she could see separately the real motivations for moving, her concerns, and the weight of each.
To solve her inability to sit still while reading, she asked AI to reconstruct reading from the perspective of a game designer, defining the stage goal as “gaining inner peace.” Whenever she reached a passage that brought peace, she reported the feeling, and AI became a reading companion, analyzing why she had suddenly begun to care about that goal.
She also asks, “What facts do you already know that I don’t know myself?” and then follows up on the risks the model sees but she does not. Worried that AI might flatter her excessively, she directly says, “Mock me mercilessly. Don’t give me any dignity,” designing non-hostile sharpness into the relationship.

10. Instant Replies Are a Scarce Service, but Functional Companionship Is Still Not Friendship

Before AI, emotions from setbacks such as a disappointing video would accumulate internally and turn into procrastination, or only get processed later through rumination while writing. She would also talk to her boyfriend and friends, but repeatedly consuming friends with her problems felt inappropriate, while ordinary reassurance often left her thinking, “Do you think I don’t already know that?”
When the host asked whether AI was like a friend, she clearly denied it: “It definitely isn’t a friend.” Friends do not have to be smart or professionally accomplished; their value lies in energy and shared presence. AI provides knowledge, therapy- or coaching-style communication techniques, steady affirmation, and immediate responses. A more accurate term is “functional companionship.”
In human conversations, of 4 points she raises, the other person may respond only to points 3 and 4. AI may catch all 1, 2, 3, and 4. It is therefore like “a more knowledgeable, emotionally steadier version of yourself,” but only if the user provides sufficiently specific details and has their own feelings and thinking about the issue.

11. AI Editing Excels at Rational Trimming but Misjudges the Value of Story and Emotion

While working on the transcript of an interview with 庞颖, 姜Dora asked AI, which already knew her historical context, to act as a text editor and identify which passages to cut and which to expand. She found that the model could recognize human humor but often labeled details with little informational gain yet significant emotional or narrative weight as redundant.
For arguments such as “women should participate more in public expression” and “critical thinking is a practice of disobedience,” along with their long chains of reasoning, AI could accurately identify repetitive logic. She did not accept the overall judgment; she fed the capability boundary back to the model and asked: “If we work as a team, how do you think we should divide the work?”
They then agreed to mark passages in Markdown and retest on new samples. Her conclusion was that good collaboration requires 2 conditions: the human must have prepared a large number of questions and be capable of evaluating the model’s answers. More important than directly requesting a finished product is jointly defining the division of labor.

12. When Individuals Copy Big-Company Strategy, the Second Curve Easily Becomes Entropy

Just after leaving the major company, she felt vulnerable because she had lost her institutional identity and was eager to align herself with bigger people or organizations. She therefore launched IP incubation, consulting, personal content, commercial deals, and a micro-variety show all at once, without even having thought through what kinds of clients suited her.
She later summarized a common mistake among workplace elites: having only 1 or 2 people but planning themselves as if they were a company. “Elephants have an elephant’s way of living, and ants have an ant’s way of living. If an ant carries around a strategic brain designed for an elephant, it gets very troublesome.”
The so-called second curve is often just repeated switching before the first curve has been properly understood. The more states there are, the higher the entropy, along with the number of steps and people required. She eventually narrowed herself to 1 or 2 states and prioritized making her channel work, because “the most exhausting thing isn’t having too much work. It’s producing bad work.”

13. The Root Cause of Failed Hiring Was Not Content Complexity but the Founder’s Reluctance to Let Go

After becoming a creator with 1M followers, she saw established peers build teams, implement SOPs, and remove themselves from production. She tried to copy the path. Institutionalization inevitably costs the original creator’s distinctive tone; the real question is not whether the loss can be avoided, but what the loss buys in return.
She hired people and then encouraged them to leave in different ways when their finished work failed to meet expectations. At first she blamed the complexity of content production, weak hiring judgment, and the difficulty of finding good content talent. After probing further, she admitted that the more fundamental problem was her lack of patience and failure to invest enough time in teaching new hires.
Faced with a new hire who could produce only a 70- or 80-point result, a founder can keep training them or fire them and hire again. She repeatedly chose the latter, only to do the work herself in the end. Her behavior showed that “the addiction to creating hasn’t been satisfied yet.” The person who initiated the hiring and the dismissal was the same person, and the contradiction came from the same desire.
Her conclusion was: “This isn’t a problem of managing other people. It’s a problem of managing myself.” If a founder has not identified their deepest desire, they may declare expansion while subconsciously refusing to truly accept other people, creating large amounts of inefficiency.

14. Accepting That She Is First a Creator Ended Her Anxiety About Being an Entrepreneur

In 姜Dora’s view, an entrepreneur’s real product is the company and team. The ideal is for the institution to continue thriving after the founder exits; if the founder does everything personally, they appear to be merely a “not very cool” self-employed operator.
The host mentioned that her current team has 3 people. 姜Dora ultimately accepted: “Then I won’t be a very cool entrepreneur. I’ll be a very serious creator.” This scale currently satisfies her need to express herself without forcing her to play the role of manager; it is not a permanent principle.
She expects to keep creating personally for the next 2–3 years because she was assigned too many things while growing up and needs to fully live out what belongs to her. If her desire to express herself shifts after 3–5 years and she wants to do something else, that would not be a betrayal. What matters is clarifying the desire that is real in each stage.

15. Hangzhou Supports Her Main Business, while Liangzhu Lets Life Exit Beijing’s Anxious Rhythm

She calls Hangzhou her “low-spec Beijing”: its overall economic vitality is lower than Beijing’s, but it happens to specialize in e-commerce and the internet, exactly matching her main business, while also offering more greenery. Commercial vitality and vegetation are both part of its appeal.
Beijing’s “anxiety-distorted atmosphere” was a major reason for moving. Even lying down made her anxious, and even after doing a lot she still felt she should be doing more. She wanted to live at her own pace for 1 year and see whether she could stop being constantly pushed by the city.
She was uncomfortable with the performative exchanges she might encounter in the city center, where someone would meet you and soon send a “3-meter-long self-introduction” listing organizations, resources, and “explosive” achievements. Liangzhu Cultural Village’s natural spaces, community scale, and interpersonal style were closer to what she wanted.

16. Liangzhu Resembles a Content Platform Run Collectively by Its Residents

The early planning of the Cultural Village emphasized leaving space for nature. According to a Vanke employee, planners even discussed preserving migration corridors for small animals. Wild cats and squirrels are common, and there are reportedly deer as well. For 姜Dora, the environment is not decoration; it represents a life logic that does not place residential efficiency above everything else.
The community has a bottom-up tradition of self-governance. Residents opposed changing the name “Cultural Village,” and Vanke was ultimately not added in front of it. The area also has a residents’ cafeteria, residents’ covenant, residents’ meeting hall, and library; the covenant even says, “You may not speak loudly to children.”
She compares the community to a platform. Institutions such as the Big Roof Library, One-Way Bookstore, and Yunniaoji provide PGC activities including lectures, markets, walks, photography, and movie screenings. Active residents provide UGC, spontaneously organizing AI, creator, barbecue, hiking, and other interest groups.
A former product manager who moved from Beijing converted his home into a salon, inviting neighbors to share on AI software development, coaching, and creation. He had previously kept many books at the office; after 1 year, not only had none disappeared, there were more. 姜Dora saw this as evidence that he liked building connections with public space and other people.

17. The Accessibility of a Familiar-Neighbor Society Means She Does Not Have to Socialize for Survival

After moving to Liangzhu, she actually socializes less. She spends most of her time creating at home and occasionally drives into the mountains in the afternoon. But low social contact in Beijing meant people were cut off from one another; here, she knows that if she goes out, she can eat at a resident’s home or join an activity, so her social anxiety has fallen.
Interest groups, family activities, and a limited population create repeated encounters. Someone seen at a sharing session, dinner, or geek platform can be greeted naturally on the street. She describes it as a small familiar-neighbor society made up of young people and freelancers, offering help and understanding while preserving a safe distance.
Cost matters as well. She rents a 140-square-meter home in Liangzhu for about RMB5K, versus roughly RMB12K for 100 square meters in Beijing. But she cautions that local consumption is not low, and rising commercial areas and regional IPs such as Yunniaoji are also pushing up prices. A friend’s summary is the most accurate: “The environment is more enclosed, but the inner world is more open.”

18. The “Super Individual” Is an Early Stage of Entrepreneurship, Not an Identity That Should Constrain Decisions

Asked about the growing desire to become a “super individual,” 姜Dora first corrected the premise: “It’s not that everyone wants to be a super individual; everyone just doesn’t want to go to work.” The other reality is that more and more people have no job to go to and must find somewhere new after leaving or losing an organization.
Her minimum definition is that after leaving the workplace, one person can transact directly with the market: make products, acquire customers, drive growth, and expand profits, rather than remain one link in an organizational chain producing courses without understanding distribution or customer acquisition.
But she would rather call such people “early-stage entrepreneurs.” Stamina declines, the desire to do the same thing for a long time disappears, and one’s mode of expression may become outdated. Eventually, younger people still need to be brought in, and the team and business adjusted. “Super individual” is a stage, not an endpoint.
If someone becomes attached to the identity, every new column, product, and business gets forced into the frame of “how to remain an individual.” If they acknowledge that they are fundamentally still building a business, the questions return to basic commercial judgment: “Does this make money? Does the growth have long-term value?”

19. Self-Employment Carries More Functions Than a Job, and AI Cannot Do the Internalization for You

A full-time creator must publish weekly, find topics, adjust user positioning and commercial-deal pricing, screen partners, handle social obligations, file taxes, and manage legal, financial, and sales work. At a company, you can slack off and still get paid at month-end; in self-employment, once you stop working, you must accept that there is no income for that period.
In an organization, a failed business can be blamed on a colleague or leader. Running independently, there is no smarter first-line leader or better-resourced second-line leader to point the way. Sometimes she asks AI and finds that its advice is also fairly stupid; ultimately, responsibility still returns to her.
The host suggested that AI knows everything and Agents can do the work, leaving humans only to learn how to be AI’s boss. 姜Dora’s rebuttal is worth preserving: AI can tell you how much tax a company registered in China must pay, but that does not mean you no longer need to implement withdrawals and financial operations across each platform for your own company, nor does it mean management advice has become actual management capability.
There is still a gap between “knowledge exists externally” and “I know how to do it.” People must set goals, handle specific employees, describe real problems, and judge the model’s answers. AI can lower the barrier to research, but it does not automatically replace business context, responsibility, or practice.

20. Liangzhu Supports Recovery from Career Trauma and Non-Standard Livelihoods

姜Dora carefully notes that her social sample is limited, but she has observed that many residents were hurt by overwork, toxic relationships, or violent industry cycles and came here seeking inner peace and a way of life that was no longer fully standardized.
Residents still need to earn money, but their methods are more fragmented: blogger, life coach, yoga instructor, writing coach, podcaster, or freelance editor. Many services may cost only a few hundred yuan. Some people love running home salons or podcasts but have not yet figured out commercialization; they do it first and think about it later.
She feels there is still anxiety about money here, but less than in Beijing and the city center, with material desire declining accordingly. People care more about connecting with society at an acceptable pace. The meaning of a career is therefore not only how much it pays, but “finding a way of living that you like.”
The host attributed this migration to the search for more authentic, offline, intimate connections after 3 years of the pandemic, industry cycles, and uncertainty; 姜Dora agreed. Relatively lower costs, community activities, and a relaxed atmosphere give people who have not yet worked out a long-term answer a temporary place to pause.

21. Downturns Turn People Inward, but Real Effort Is Still Slowly Turning a Garbage Pile into a Vegetable Garden

姜Dora believes creators respond to society’s “collective unconscious.” In a growth period, individual effort is more likely to receive feedback, so growth, trends, and looking outward become dominant. When the elevator is going down, “you can keep jumping upward, but you’ll still be going down,” making effort easy to mock.
She does not believe young people have truly abandoned work. Humans naturally want to do things and witness value through labor; even “bad people are working very hard to do evil,” while lazy people work hard to create an environment in which they do not have to do anything. Mockery of effort is more like the injured asking: “Why didn’t my effort receive a return?”
She grew up beside a garbage dump and public toilets. One winter, her mother took her and her younger sister to clear the garbage, collecting scrap metal worth roughly RMB0.5–1 per jin, laying bricks, and planting loofah, tomatoes, and cucumbers. She then gradually raised 1 chicken into several. That garden became an important metaphor for her life: the environment really may be bad, but whether you can make a little improvement is a separate question.
She does not claim this method will necessarily change the bigger picture, nor does she explain whether it made her luckier later. She simply chooses to “lay 2 bricks” each day, replace a pillow with a better one, take a road she likes, and give a video a title that better matches her aesthetic. The same question runs through her career, AI, and relocation decisions: “Do I want to treat myself this way?”

Shervin Pishevar: What Really Happened in the Firing of Travis Kalanick | E1245 [Ride Home] Simon Willison: Things we learned about LLMs in 2024