DeepSeek TUI Founder "Whale Brothers" Gets a Harsh Lesson in China

This article is republished from the WeChat public account: AI Humanist by Shansen Nan , author: Shansen Nan. Original title: "DeepSeek TUI Founder 'Whale Brothers' Gets a Harsh Lesson in China"

A few days ago, the AI community was buzzing with gossip about the founder of DeepSeek TUI — it was everywhere on social media groups.

Everyone was just fixated on a single screenshot of a WeChat group chat, with almost no one sorting out the full story from beginning to end.

I traced through the entire timeline and found that the complete picture was far more complicated than that screenshot suggested.

An American named Hunter Bown created an open-source project called DeepSeek TUI. Shortly after releasing it, he came to China — and got a harsh "lesson." Before leaving for the US, he said in a WeChat group that he felt he had been "exploited." The screenshot went viral, and the internet exploded.

DeepSeek TUI is a command-line Agent "specially adapted" for DeepSeek — essentially a DeepSeek version of Claude Code.

But whether "specially adapted" is genuine optimization or just marketing buzz has never been settled.

When the project was first released, the response was lukewarm.

The turning point came when Hunter posted a Chinese tweet and created a fan organization called "Whale Brothers." The entire tweet was clearly targeting Chinese developers, and he earnestly added —

that he still didn't know how to use WeChat yet.

Then something magical happened.

Within half a month, the project surpassed 10,000 stars, and now sits at 34.5K stars. The chart below shows its star growth curve — rising from flat ground with absolutely no warm-up —

Looking at just this curve might not mean much. So I picked two comparison groups to put it in perspective.

The first is Karpathy's star project, currently at 148K stars. Everyone knows who Karpathy is — former OpenAI founding team member, AI top influencer. His project's growth curve looks like this —

Maybe that's still not intuitive enough — after all, Karpathy isn't "that that that famous" in the Chinese internet.

So let me pick one everyone knows — Anthropic's Claude Code, currently at 126K stars. No introduction needed for this one. Its growth curve is "relatively gentle" —

Looking back at DeepSeek TUI's curve, it's truly explosive.

Within just a few days, the project received massive help and donations, with community contributors flooding in wave after wave. This coincided with DeepSeek's recent V4 model release, so the DeepSeek buzz was already at a peak, and this wave pushed the project through the roof.

Even Fu Sheng joined the "action," retweeting Hunter's post with the comment — DeepSeek TUI is really great.

I actually deployed this project myself at the time, and my experience was that it's a fairly ordinary project.

It didn't have any special adaptation for Chinese users — the project already supported many languages, with Chinese being just one of them.

But there was one thing different from other open-source projects — it had a dedicated Chinese website.

Within days, DeepSeek TUI topped the GitHub repositories chart, gaining thousands of stars each day.

Hunter posted another tweet, mixing Chinese and English, thanking all the "Whale Brothers" for their contributions to the project.

Later, Hunter renamed the project.

Perhaps because "DeepSeek TUI" clashed too much with the official DeepSeek branding, the project was renamed to Codewhale.

After the rename, Hunter began giving back to his fans.

He tweeted, welcoming everyone to join him and "the Whale Brothers organization spread all across China" —

And then his China tour began. As for how this trip was initiated — it could have been that people in China invited him, or Hunter wanted to come himself, and both sides hit it off.

That part is impossible for outsiders to say for sure.

But soon, an event page appeared on Luma —

Dates: May 15 to June 5, 2026

Just looking at the event page, combined with the events that actually took place afterward, the scale was absurdly large.

The organizer was the DeepSeek TUI community, listed below were co-organizers, strategic partners, exclusive AI ecosystem partners, supporting media, and supporting enterprises — so many entities you couldn't count them all.

Plus a large number of partner organizations and communities.

Some prestigious universities even appeared — Peking University, Tsinghua University, Wuhan University — all on the list.

I later discovered another detail — the DeepSeek TUI project actually came with a 210,000-word open-source guide.

It explained the project's code logic, NPM installation methods, and all underlying principles in clear detail — at a very professional level.

A guide this comprehensive most likely wasn't the work of Hunter alone; it seems more like it was put together with help from supporters in the Chinese community. Just looking at this guide, you can see a tremendous amount of developer effort went into it.

And then, Hunter actually came to China.

He gave a self-introduction in Hangzhou, with his slides almost entirely in Chinese, sharing his background and introducing himself as the creator of the open-source tool DeepSeek TUI.

What many didn't expect was that he wasn't originally a programmer.

(Image from @tualatrix)

And then, that opening group chat screenshot was leaked and spread online.

According to information circulating on the internet —

This was Hunter's first time in China, and there was a language and cultural barrier. All his itinerary and coordination were handled by people on the ground in China. But later, he gradually realized that he was being used as a bargaining chip for business and traffic, with the DeepSeek name being casually exploited, and the events deviating from the original intent of pure community exchange.

He had no choice but to terminate his entire trip early and leave China urgently.

The incident continued to ferment in open-source community discussions, with many developers calling for the relevant parties to respond publicly.

At this point, the whole thing turned into a Rashomon 🌀

The real concern isn't about whether Hunter himself was exploited by people in China — that part is a mix of truth and fiction, and everyone has their own judgment, so I won't provide additional commentary here.

What truly sparked the controversy are the following points.

First, the project's bizarre viral success.

That star growth curve has exceeded the natural rhythm of any normal project. In hindsight, people attribute it to two main reasons —

One, DeepSeek had just released a new model, so the buzz around DeepSeek was already high. Two, Hunter's tweet was written in Chinese, specifically targeting the Chinese community.

So a significant part of why the project blew up was that it captured this wave of enthusiasm from Chinese developers.

Second, right after the project launched, he immediately came to China for exchanges.

And it was precisely this China trip that ultimately produced the group chat screenshot saying "I may have been commercially exploited."

Third, there are technical controversies.

The project claims to have been specially adapted for DeepSeek. But an Agent framework similar to Claude Code, claiming to be adapted for a specific model — that itself is a black box.

Writing a few hundred lines of code to adapt to a model's specific features can be called "adaptation"; doing nothing and just tweaking a few prompts can also be called "adaptation." Outsiders simply cannot verify it.

On the other hand, people also noticed Hunter's background —

High school music teacher → MBA → musical instrument salesman → learned vibe coding → built this project.

For instance, during his speech in Hangzhou, as a former music teacher, he drew an analogy between Agent and tuning, discussing how to optimize the entire Agent system.

Whether the technical analogy is correct aside, the fact that someone with a music teacher background was talking about AI Agent already struck many netizens as off 🤔 (Image from @tualatrix)

Someone also dug up a detail —

Before DeepSeek TUI, all of Hunter's previous projects had almost no stars.

Yet this one project suddenly took off out of nowhere.

When you put all of this together, in the specific context of the internet, the whole thing gets interpreted as —

Hunter executed a precision marketing campaign targeting the Chinese community for DeepSeek TUI.

The Chinese community gave him plenty of face, with various corporate entities and media outlets aggressively promoting the project, driving the hype through the roof. Seeing this response, Hunter came directly to China and held a special exchange.

This kind of scenario was unprecedented — an overseas project suddenly going viral, claiming to be adapted for the Chinese community, with the founder personally flying over for exchanges.

And then he came to China, "got a harsh lesson from the domestic AI environment, and left urgently."

The whole thing has reached an absurd new level.

Some well-known developers on X are also criticizing this situation, and the public discourse is quite heated.

Their core argument goes something like this —

Whether an open-source project gets stars on GitHub doesn't have that much to do with code quality, release timing, or even functionality itself.

Take Hunter himself — someone specifically counted his GitHub project history —

In his first 90 days, he released over 60 projects, and the total stars from all of them combined was less than 100.

During those three months, he barely wrote any code. Instead, he spent every day on X, frequently showing up in the comment sections of AI influencers to build familiarity.

Through this sustained, high-density exposure, he built an identity label for himself — an American guy building Chinese AI tools.

This cultural contrast created a naturally viral topic, making it easy for the Chinese community and media to keep the hype going.

The comments under his tweet were quite interesting —

Broadly speaking, people used the analogy of "a foreign netizen commenting that one Chinese dish is delicious, and it instantly goes viral in China" to describe this event.

The logic is the same — we have a natural tendency to get excited when "foreigners actively acknowledge us."

As for why this incident generated such intense and absurd public discourse, in my personal opinion —

The first point — "ranking manipulation" has appeared on every platform.

The most obvious example early on was Product Hunt.

It was originally a platform for tracking product popularity, but over time, because it's inherently a ranking system — higher rankings bring more traffic — people started aggressively gaming the system.

In the past year, almost every AI product promotion includes a line like "Product Hunt #X of the day" or "top 3 of the week," which has become such a cliché that people now mock it 😆

Eventually, this playbook spread to other platforms — like GitHub.

GitHub is a very old platform.

It originally represented the spirit of open-source internet — a group of people collaboratively building the underlying infrastructure of code, driven by passion and enthusiasm, a sense of building a better world.

But now you'll find that many projects' READMEs are extremely abstract and exaggerated, full of buzzwords, while the projects themselves have little actual value.

Someone recently pointed something out —

A huge number of so-called Skill projects are just shells wrapping yt-dlp.

Many of these shell projects don't even work, but as long as the shell is there and the README is flashy enough, they can get 300+ stars.

I'm not trying to criticize this.

In this era, this kind of thing is perfectly normal.

As long as a platform has quantifiable ranking numbers — whether it's Product Hunt rankings or GitHub stars — ranking manipulation will inevitably emerge.

Ranking manipulation itself isn't the problem. What's really controversial is "what kind of manipulation counts as normal, and what kind counts as abnormal."

A while back, the "Resident Evil" actress posted a project targeting Claude Code's Memory mechanism, claiming she built it entirely by hand.

Later, people discovered that there was another crypto founder behind her who was the one with actual coding experience. The "pure handmade" story had a fair amount of marketing spin.

But that didn't stop the project from being wildly shared and driven to viral heights.

Once the hype died down, no one paid attention to the project's actual value.

The second point — "adapting for the Chinese community" struck a chord with many Chinese developers.

Everyone's initial intentions were good.

With AI exploding over the past two years, Chinese developers rarely see this kind of story —

An overseas developer specifically posts a Chinese tweet for the Chinese community and creates an adaptation for Chinese users.

After the project went viral, he even flew to China in person for an exchange.

There had been almost no precedent for this in the past two years. So enthusiasm was instantly ignited, and Hunter was treated like a VIP.

But once enthusiasm takes hold, economic factors automatically creep in.

The nature of the whole situation became incredibly complicated, to the point where it's impossible to fully explain.

You could see it from Hunter's Luma event page.

The entire event was packaged in a very "domestic Chinese" style — organizer, co-organizer, strategic partner, supporting media, supporting universities, supporting enterprises — a long, long list filled with many well-known entities.

For an overseas independent developer, and for anyone who genuinely loved and supported this project, this was something they had never seen before.

Hunter's later comment "I seem to have been exploited" ultimately boils down to this — "I've never seen this kind of setup before."

On one side, you have GitHub where stars can be quietly bought and sold; on the other, a Chinese community that easily gets hooked on foreign recognition.

Given all this, it's not surprising that things turned out this way.

But to be fair, this can't be entirely blamed on the Whale Brothers. Put AI traffic on anyone, and they wouldn't know how to handle it either.

In the past six months, the pace of domestic model development has been incredibly fast — DeepSeek, Qwen, Kimi, Zhipu, one after another releasing new versions, with Chinese names appearing almost daily on open-source leaderboards.

The enthusiasm for participation in the entire AI open-source community has followed suit.

As enthusiasm rises, money, resources, and opportunities follow.

Money comes in, and chaos comes with it.

This is an inescapable cycle for any field simultaneously targeted by capital and attention.

Once this wave of chaos passes, what remains will outweigh what is forgotten.

Behind that 210,000-word open-source guide, there are a large number of passionate Chinese AI builders sincerely writing code; among the developers who joined the "Whale Brothers," some initially came out of curiosity.

These things won't be scared off by a single group chat screenshot.

Finally, forgive me, I must add a disclaimer.

All the above information comes from public sources. For key information requiring source verification, you can DM the backend.

I have no personal connection to this incident. I don't guarantee the accuracy of this gossip, nor do I intend to criticize any party —

I'm simply organizing the entire timeline of events 🐳

Thank you for reading and for being "present"!