Last updated: April 2026

China’s OpenClaw Mania — What the Red Claw Hats Mean for the Rest of Us

At AI events in Shenzhen right now, they’re serving piles of lobster. There are lobster balloons. Lobster headbands. Lobster plushies in claw machines. A kiddie pool with live lobsters.

The crowd isn’t there for the crustaceans. They’re there for OpenClaw.

In the past few weeks, OpenClaw — the open-source AI agent framework built by Austrian developer Peter Steinberger — has gone completely viral in China. Not viral among tech workers. Viral among everyone. Reuters reported on a 60-year-old retired electronics worker named Fan Xinquan attending an OpenClaw training event hosted by an AI startup. A Baidu executive said primary school parent group chats have been “overwhelmed” by OpenClaw discussions. His daughter came to him and asked: “Dad, I see you raising a lobster every day. Can I have one too?”

This is wild. And it matters more than it might seem.

Why “Raising a Lobster”?

The nickname comes from OpenClaw’s lobster branding — and it stuck in China because it captures something true about the experience. You don’t just use an agent. You raise one. You teach it your preferences, your workflows, your way of doing things. Over time it gets better at being yours.

That’s not how people think about software. It’s how they think about a relationship. And the Chinese framing — “my lobster,” “raising a lobster,” “feeding it good information” — might actually be the most honest description of what working with an AI agent feels like once it’s set up.

Chinese tech companies are fully leaning in. Baidu, Tencent, Zhipu — they’re all hosting events, running workshops, building products around OpenClaw. The country’s tech shares jumped as much as 22% in recent weeks as companies scrambled to get into the agent space.

Nvidia CEO Jensen Huang called OpenClaw “the next ChatGPT.” OpenClaw has become one of the fastest-growing projects in the history of GitHub.

Who’s Actually Showing Up to These Events

Here’s the part that should reframe how you’re thinking about this.

It’s not developers. It’s not AI researchers. It’s retirees hoping to organize their industry knowledge. It’s people looking for a side hustle during retirement. It’s parents whose kids are asking about it at the dinner table.

Bai Yiyun, a retiree at the Zhipu event, told Reuters she hopes to use her agent to start a side hustle. “Some people use it to buy lottery tickets or for stock picking, others use it to create money-making content,” she said.

These are not people who grew up with computers as a creative tool. These are not people who think in terms of APIs and workflows and automation. They showed up at an event because they heard something that could help them — and they wanted to learn.

And they’re learning. That’s the story.

What This Signals (It’s Bigger Than China)

A technology that starts with developers and ends up with retirees at lobster-themed meetups has crossed a line. That line is called mainstream adoption, and once a technology crosses it, the dynamics change fundamentally.

AI analyst Wei Sun at Counterpoint Research put it clearly: “If DeepSeek marked a milestone for open-source large language models, then OpenClaw represents a similar turning point for open-source agents.”

DeepSeek was a moment when the AI world realized that frontier-level models didn’t have to cost frontier-level money. OpenClaw appears to be a similar moment — when AI agents stopped being something you needed a developer to set up and started being something a retired electrician in Beijing could figure out at a Saturday morning event.

That’s the signal. The barrier to entry just dropped off a cliff.

The Window of “Early” Is Closing

There’s a useful window in any technology’s lifecycle — the period after it works reliably but before everyone knows about it. In that window, the people who show up early get a real advantage: they figure things out, they build the habits, they develop the intuition that later adopters have to learn from scratch.

That window is closing for AI agents. Fast.

A few months ago, OpenClaw was a GitHub project that developers knew about. Today, it’s the subject of a Bloomberg opinion piece, a CNN business feature, and Reuters reporting from Beijing — all in the same week. Chinese tech companies are mobilizing entire product lines around it. Jensen Huang is calling it the next ChatGPT.

When Bloomberg is writing about the cultural phenomenon, you’re no longer in the early-adopter phase. You’re in the mainstream adoption phase.

What This Means for You

If you’ve been curious about setting up an AI agent but told yourself it’s too technical, the Chinese retiree story is your answer. It’s not too technical. Fan Xinquan, 60 years old, retired from electronics work, figured it out at a training event on a Saturday.

The people lined up at these lobster-themed meetups in Shenzhen aren’t asking “what is an AI agent?” They’re asking “what should I have my agent do first?” That’s the right question. And the answer is simpler than most guides make it sound.

Tell your agent what you need. Let it figure out the how. Correct it when it’s wrong. Repeat.

That’s raising a lobster. That’s what’s happening in Beijing and Shenzhen right now. And if that’s what Chinese retirees are doing on their weekends, there’s no version of this that’s too complicated for you.

The One Thing Worth Taking From This

The red claw hat crowd isn’t technically sophisticated. They’re not optimization-obsessed. They’re just people who heard that this thing could help them, showed up to learn, and started.

That’s actually the best possible approach. Not “wait until I understand it fully.” Not “wait until the perfect setup.” Just — start, and let the agent teach you what it can do.

If you haven’t set up an agent yet, this week is a good week. The technology is mature enough to work. The resources to learn it are good. And if you wait until AI agents are as ubiquitous as smartphones — which this moment suggests is closer than anyone thought — you’ll be starting from zero while everyone else has a year of experience.

Your lobster is waiting.

Sources: Reuters — OpenClaw enthusiasm grips China, schoolkids and retirees alike raise ‘lobsters’, CNN Business — Behind the lobster merch, China’s latest tech obsession could be a game changer, Bloomberg — China’s OpenClaw Madness Has a Method, CNBC — How China is getting everyone on OpenClaw