Openclaw, from a sweaty Manhattan ballroom to a Shenzhen line: what ‘raising a lobster’ looks like on the ground
Openclaw was the reason a thousand-plus people squeezed into a Manhattan apartment-turned-ballroom, sweating through a night of presentations, a rap performance, and a seafood spread of free lobster tails—while, across the Pacific, nearly 1, 000 people lined up outside Tencent Holdings’ Shenzhen headquarters to get the same software installed on their computers.
What is Openclaw, and why are people building “agents” around it?
At its simplest, Openclaw is a free software package that lets people create “agents”—AI systems that can perform autonomous tasks with limited human oversight. In the Manhattan crowd, the appeal was practical and immediate: users described sending text or WhatsApp messages to their agents, then stepping back while the software handled tasks within the reach of today’s AI systems.
ClawCon NYC, held Wednesday night, drew an eclectic mix that ranged from college students and working moms to hedge fund technology teams. The scene was part costume party, part recruitment fair: a man in a neon-blue jellyfish hat scrolled for messages from his personal AI assistant while people wearing Pegasus wings tried to recruit users for their latest AI solution.
“It’s getting hot, and the lobster is getting warm, ” said Michael Galpert, one of the event’s hosts, as he urged the crowd to settle before the evening’s presentations began. “Welcome to ClawCon. ”
How did a lobster-themed project become a global magnet—from New York energy to Shenzhen queues?
The software’s origin story is a reminder that the AI ecosystem is also a culture—one built on naming, inside jokes, and volunteer labor. The project was launched in November by software engineer Peter Steinberger and was first called “Clawd, ” a nod to Anthropic’s Claude AI system. After Anthropic strongly suggested Steinberger change the name to avoid legal issues, the project kept its lobster-themed heritage and settled on the OpenClaw moniker.
From there, momentum became a social force. Several ClawCon attendees who started using the tool in January described themselves as “veterans, ” a small word that hinted at how quickly identities form around new technology. Tomas Taylor, a programmer and ClawCon organizer, said the energy was palpable: “There’s a kind of electricity and energy you can just feel in the room, ” he said. Taylor used his own system to help plan the event and interact with vendors, presenting Openclaw not as a distant promise but as a working assistant in the background of real logistics.
In southern China, the surge looked different but carried the same undertone: ordinary people trying to bring a new capability into their daily routines. On Friday, nearly 1, 000 people lined up outside Tencent’s Shenzhen headquarters to install the open-source AI agent software after an invitation from Tencent’s cloud-computing unit. Company engineers installed it for free.
Who is using it, what are they asking it to do, and what worries follow the excitement?
In New York, users described a range of hands-off uses: having agents listen to podcast episodes and email summaries; negotiating with car dealers over the price of a vehicle; and ordering and paying for grocery deliveries—without direct human input. Openclaw positions itself as a bridge between powerful AI systems and the real-world tasks people want done, working with paid AI systems from OpenAI and Anthropic or freely downloadable AI models, including models from Chinese companies like DeepSeek or Alibaba.
In China, users have leaned into productivity and personal leverage. Chinese consumers are using the software for stock picking, report writing, slide decks, emails, and coding. The line outside Tencent pulled together a mix of amateur developers, retired space engineers, housewives, students, and AI enthusiasts—an intergenerational, cross-profession snapshot of who feels the pull of “personal AI. ”
Mark Yang, a Shanghai-based designer and early adopter, described the experience in workplace language that many people understand immediately: with the AI assistant, he said, it felt like having “virtual staff” that handled assignments and reduced workload.
Yet the excitement carries friction. In southern China, the rush to embrace AI agents is unfolding as privacy concerns intensify, even as global concern rises over how AI agents could disrupt lives and industries. In the same moment that a user marvels at an agent remembering preferences and keeping detailed notes to better serve them over time, the shadow question follows: where does that information live, and who controls it?
What responses are emerging—from volunteer code to corporate installation drives?
The phenomenon is being shaped by two kinds of infrastructure: informal communities and formal institutions.
On the community side, ClawCon functioned like a high-energy meet-and-greet—main stage presentations downstairs and a less-crowded VIP area upstairs with a livestream of the event unfolding below. The software itself relies on volunteers to maintain its code and respond to users, a reminder that “open-source” is not a slogan so much as ongoing, often invisible work.
On the institutional side, Tencent’s cloud-computing unit turned adoption into a physical event—an invitation, a line, engineers installing the software for free. At the same time, social media posts in China offered similar installation services for fees ranging from tens to hundreds of yuan, creating a parallel market that follows wherever demand outpaces confidence or know-how.
Seen together, the New York ballroom and the Shenzhen queue suggest that Openclaw is not spreading only through downloads and developer forums, but through gatherings where people watch each other use it, talk about it, and make it feel safe—or at least possible—to try.
Back in Manhattan, the room that felt like a party also felt like a rehearsal for something more serious: people learning how to hand off pieces of daily life to autonomous systems, one message at a time. In Shenzhen, the line outside Tencent’s headquarters carried a similar impatience—less spectacle, more urgency. Between the warm lobster trays and the careful laptop installations, Openclaw is becoming a shared experiment: not just what AI agents can do, but how quickly ordinary people will reorganize their routines around them.
