When Hu Yanbin developed an app in just one month, the AI era had already arrived.
```
If you've been paying attention to trending topics recently, you might feel as if “am I still dreaming?”
The literary master Yu Hua, author of “To Live”, proudly announced on camera that he has learned “local deployment”.
Meanwhile, singer Hu Yanbin, who rose to fame with “Take Everything You Want”, posted a photo of his computer screen on social media, captioned:
Anyone who does Vibe Coding knows this pose! Fixing bugs on the road…
Even netizens couldn’t help but leave witty comments:
The tokens you want, take them all, turn the memory into empty…
This sounds like a magical crossover between entertainment and literature, but if we look past the entertainment surface and dig deeper, we’ll find a business signal powerful enough to shake the entire industry: The era where AI makes programming accessible to everyone has truly arrived.
A Singer’s First Programming Experience and the Vibe Coding Wave
The app Hu Yanbin created himself is called “Yanhuo”, a community platform for fans.
The astonishing thing is Hu Yanbin didn’t hire an outsourcing team or find a tech partner.
He simply sat in front of his computer, described his requirements to AI in natural language, and let AI write the code.
When encountering bugs, he pasted the error messages to AI for fixing; when wanting new features, he kept “outputting” via his mouth.
He compared this process to beating levels in a game, also like music production he’s familiar with—polishing over and over, scrutinizing every detail.
What Hu Yanbin used is exactly the new programming paradigm that OpenAI co-founder Andrej Karpathy ignited in Silicon Valley—Vibe Coding (coding by intuition).
Its key features thoroughly overturn traditional software engineering:
Natural language is code: You don’t need to understand syntax or APIs, just say “I want a login page with phone number verification”, and the code is automatically generated.Conversational debugging: Just throw errors to the AI, goodbye to wild searches.Progressive iteration: Start with the smallest workable version and add on, perfectly in line with lean startup.
Because of a leap in large language model coding abilities, the threshold for app development is, for the first time in history, being lowered from “needing an expensive tech team” to “one person + one computer”.
Even Hollywood Hardcore Female Stars Are Getting 50,000 GitHub Stars
This bold cross-industry wave has even swept Hollywood.
(Right: Milla Jovovich)
Milla Jovovich, star of the “Resident Evil” series, troubled by AI assistants’ inability to "remember" previous conversations, teamed up with an engineer friend to create an open-source AI long-term memory system called MemPalace using Claude Code.
MemPalace is a locally run AI long-term memory system. It saves your conversations and project files word for word, and retrieves them using semantic search. Note: word-for-word storage, no summarization, no rewriting.
The advantage is no loss of information; the disadvantage is high storage cost.
The project was inspired by the ancient Greek “method of loci”, using four-layer progressive loading memory stacks to achieve an astonishing 96.6% pure local retrieval recall rate.
Although the project has technical flaws—the team found that the touted palace architecture actually decreased retrieval performance by about 12%, and early marketing was exaggerated.
The team subsequently spent time apologizing and correcting, but it still garnered over 50,000 stars on GitHub.
Setting aside technical imperfections, the most shocking thing is: An action movie star, starting from a real personal pain point, used AI to lead a cutting-edge technology experiment.
Why “Domain Expertise” Is More Important Than “Coding Expertise”
Whether it’s musician Hu Yanbin or Hollywood actress Milla Jovovich, their cross-industry app creation actually reveals the core significance of the AI era.
As Claude Code founder Boris Cherny said:
The best accounting software may not be written by engineers, but by outstanding accountants—because coding is the easy part, domain knowledge is key.
We are witnessing an unprecedented “mass application creation” movement. Just look at who’s joining this wave:
70-year-old grandpa using AI to make a bookkeeping app for managing retirement funds;Yiwu small-goods merchants made multi-platform order-sync inventory tools;Middle school teachers made an assistant that automatically identifies handwritten answers and scores them;Stay-at-home moms made baby food apps that recommend combos based on month-age.
In the past, these specific, scattered, small “pain points” that can’t support millions of users had no commercial value to justify hiring a tech team. But now, the power of technology is decentralized.
It’s like Gutenberg’s printing press in 15th century Europe. Before its invention, only 10% knew how to read and write—literacy was an exclusive privilege.
Now, “writing software” is becoming like “sending texts” or “using Word”, a truly democratized basic skill.
When development is no longer an engineer’s privilege, those who truly know the business and pain points will be the best software creators.
Will Programmers Panic? From "Code Worker" to "Code Director"
Now that celebrities and stay-at-home moms can make apps, will programmers really be replaced?
The answer: No, but their role will fundamentally change.
Vibe Coding only eliminates the “repetitive coding work” that occupies 80% of junior programmers’ time (like CRUD, page layout).
In the future, programmers’ core value will shift toward system architecture design, complex logical reasoning, and security offense/defense.
Programmers will evolve from hands-on “code workers” to “code directors”, overseeing the big picture and auditing quality.
Big tech companies have already sensed this trend.
Tencent and Ant Group have launched low-code & AI coding assistants, Bytedance’s Duobao programming assistant already has over a million monthly active users.
The industry consensus is forming: The future of app development is a right for anyone with ideas.
Conclusion: An Unlimited Future
Hu Yanbin “talked out” an app in a month, showing us the ultimate form of democratized creative tools.
Just like music production once required a million-dollar recording studio, now a single computer with Logic Pro can handle the whole process.
Vibe Coding is the “Logic Pro” of app development.
Both Hu Yanbin’s “Yanhuo” and the “mass app creation” movement are also the start of a recursive model of “using AI to make AI apps”.
What’s most moving about this is not just that Vibe Coding’s real meaning is not just that everyone can make an app, but that everyone clearly knows: in this era where AI is rushing in, you can make an app.
As long as you have pain points, ideas, and an open mind facing new things.
This “can” itself is the greatest force to reshape future business and creativity.
Risk warning and disclaimerThe market involves risk, investment requires caution. This article does not constitute personal investment advice nor consider the special investment goals, financial conditions, or requirements of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article fit their specific situation. If invested accordingly, the responsibility is your own. ```