The recent explosion of interest in the AI agent software OpenClaw in China reveals a critical shift: ordinary users are now willing to pay for AI services, a behavior previously uncommon in a market accustomed to free, data-driven software. While many individuals hoped for effortless AI productivity, the reality has been far more complex, yet the underlying economic trend is clear.
The Rise of the “Lobster” Mania
OpenClaw, dubbed “lobster” by some Chinese users, went viral after social media influencers demonstrated its potential for automated stock trading and investment. The software prompted workshops across China, drawing hundreds eager to learn how to deploy it. Tech companies and local governments quickly responded, integrating OpenClaw into platforms and offering subsidies to entrepreneurs. The frenzy even led to viral images of elderly citizens lining up to install the software, highlighting the widespread appeal.
However, the experience has been uneven. Users without technical skills quickly found themselves struggling with API ports, configuration errors, and endless “working on it” loops from their agents. Despite the hype, many found OpenClaw unusable without coding expertise. One user, George Zhang, abandoned stock trading with his lobster and repurposed it for AI industry news aggregation instead.
The Business Behind the Buzz
The real winners aren’t individual users but the Chinese tech companies capitalizing on the surge in demand. Firms like Tencent, Alibaba, ByteDance, and Z.ai recognized the AI productivity FOMO (fear of missing out) as a rare opportunity to monetize AI services. The key lies in LLM API calls: a single OpenClaw instance consumes far more tokens than a typical chatbot, ensuring continuous revenue for providers.
Poe Zhao, a tech analyst, explains that Tencent engineers even set up tables outside headquarters to help users install the software for free – a clear incentive to drive API usage. The system works because even failed installations still generate token costs.
Technical Barriers and Hidden Costs
The installation process itself proved challenging for many. Song Zhuoqun, an AI startup intern, spent hours pasting code generated by ByteDance’s Doubao chatbot into OpenClaw, only to encounter errors repeatedly. Changpeng Zhao, founder of Binance, lamented that post-installation time is spent “tweaking that useless lobster that can’t do anything.”
Most nontechnical users rely on rented cloud servers and paid LLM models (like Kimi), adding significant costs. Running OpenClaw for a year can easily exceed $30, and complex tasks further drain token budgets. Some users joke that OpenClaw will soon be replaced by unpaid interns – a cheaper alternative to constant token consumption.
The Tech Giant Response: Proprietary Claws
Virtually every major Chinese tech company has rushed to create its own OpenClaw version: Tencent’s QClaw, ByteDance’s ArkClaw, Moonshot’s KimiClaw, and Z.ai’s AutoClaw. These proprietary clones promise easier installation and seamless integration with existing ecosystems, but primarily aim to lock users into their platforms.
The takeaway is that China’s OpenClaw craze demonstrates a willingness among ordinary people to pay for AI. This willingness will drive further monetization efforts by tech giants, even if the user experience remains flawed for many. The real gold rush isn’t about the software itself; it’s about the sustained revenue from paying customers.
