George Zhang thought OpenClaw might make him wealthy, regardless that he didn’t actually perceive how the viral AI agent software program labored. But he noticed a video of a Chinese social media influencer demonstrating the way it could possibly be deployed to handle inventory portfolios and make funding selections autonomously. Zhang, who works in cross-border ecommerce within the Chinese metropolis of Xiamen, was intrigued sufficient that he determined to attempt putting in OpenClaw in late February.
Zhang is likely one of the many individuals in China who bought swept up within the craze over OpenClaw lately. Workshops educating individuals the way to use the AI agent have popped up in cities throughout the nation, drawing crowds of tons of. Tech firms are racing to combine OpenClaw into their platforms, whereas native governments have introduced subsidies for entrepreneurs constructing merchandise with it. Late final week, pictures of grandpas and grandmas lining as much as set up the software program went viral throughout the web.
After renting a cloud server from Tencent and shopping for a subscription to the Chinese massive language mannequin Kimi, Zhang might begin chatting together with his OpenClaw agent, or his “lobster,” as many Chinese individuals name theirs. At first, Zhang tells me, he was impressed by the AI agent as he watched it rapidly generate an extended market evaluation based mostly on the most recent breaking information. But a couple of days in, his lobster began slacking off, and it could generate solely a primary define of market developments as an alternative of an in depth report. He requested OpenClaw to generate one thing like what it had executed on the primary day, to which the agent perpetually responded that it was “working on it” earlier than by no means returning any outcomes.
Zhang’s conclusion was that OpenClaw just isn’t designed for individuals like him who have no coding expertise. “It would tell me I needed to configure the API port. But that’s a technical task, not something I can do unless I had a tutorial walking me through it step-by-step,“ he says. In the end, he gave up on letting his lobster trade stocks, settling instead on asking it to aggregate AI industry news, which he used to build a social media content farm on WeChat.
This week, I checked in with half a dozen users of OpenClaw in China about their experiences with the agent, and a clear picture of division emerged between the adopters who are technologically savvy and those who are not. People who are proficient in AI see OpenClaw as a game changer in productivity, but those with no technical background feel they were promised a miraculously powerful AI product that ultimately didn’t deliver. But by the time the bubble burst, they had already started paying for cloud servers and LLM tokens.
The real driver of the OpenClaw mania in China isn’t everyday users, but rather the Chinese companies that stand to benefit financially from its widespread adoption. Major tech firms like Tencent, Alibaba, ByteDance, Minimax, Moonshot, and Z.ai all saw the AI productivity FOMO as a rare chance to get normal people to start paying for AI services, and they are reaping the biggest rewards from it.
“A chatbot uses only a few hundred tokens per conversation; a single active OpenClaw instance can consume tens or even hundreds of times more tokens per day,” says Poe Zhao, a tech analyst and founding father of the publication Hello China Tech. Every new person of OpenClaw is somebody who’s paying 24/7 for LLM API calls. “That’s why Tencent engineers were setting up tables outside headquarters to help people install the software for free,” he says.
“I Couldn’t Understand Any of It”
Song Zhuoqun, a university scholar in China, says she began operating into issues with OpenClaw as quickly as she tried putting in it. Song is a social media intern at an AI startup however has no programming expertise, so determining the way to get OpenClaw operating turned out to be tough. She requested Doubao, ByteDance’s fashionable AI chatbot, to generate a step-by-step tutorial for her, nevertheless it wasn’t a lot assist.
“There were pages full of code, and I couldn’t understand any of it. I just kept asking the AI to generate a response for me, then I’d paste it over, run it, and it would run into an error, so I’d try a new response,” she says. The set up ended up being probably the most irritating a part of making an attempt out OpenClaw for Song, and he or she didn’t really feel like she realized something from it.
https://www.wired.com/story/china-is-going-all-in-on-openclaw/