When cloud service providers also start selling shovels, the AI bubble is not far from bursting.

Recently, OpenClaw has become a huge hit, with offline training classes popping up all over Beijing, Shanghai, Guangzhou, and Shenzhen. In this AI frenzy, it’s finally the ordinary people’s turn to jump in and take a shot.

In the past few years, the only players in the AI industry chain selling shovels were various hardware manufacturers—selling GPUs, optical modules, servers, and racks—all making huge profits. Cloud service providers faced rising hardware costs on one side and struggled with slow commercialization on the other, essentially burning money until the emergence of OpenClaw.

OpenClaw’s impressive capabilities finally convinced ordinary people to pay for tokens. Previously, AI was mostly about writing articles, creating beautiful PPTs, helping with coding, or improving search engines—tools to boost efficiency. But OpenClaw’s abilities as a personal assistant and virtual employee seem capable of fully replacing a real employee. This has caused real employees to panic, frantically learning about this AI that might replace them, even fantasizing about taming the AI to become their boss.

However, shifting perspective from workers to companies, if AI truly has the ability to fully replace employees, companies would simply lay off those employees and directly use AI, rather than relying on employees to use AI to do their original jobs. Companies don’t need middlemen. Additionally, companies naturally have advantages over individuals in AI adoption—they know which tasks AI can do better and which require human input, they have higher tolerance for errors, and they have more capital. AI cannot completely replace humans, and building a fully autonomous company solely with AI is a pipe dream. Ordinary people trying to start a business with AI assistance must learn AI and also learn how to manage people—more difficult than managing people alone.

In my view, if AI could fully replace my job in the future, the most important thing for me would not be to compete with my former employer using AI, but to switch to a job that AI cannot replace for now but is deeply integrated with AI—like an electrician. Because the AI wave is unstoppable, ordinary people need to learn to go with the flow rather than fight it. The panic to learn OpenClaw is actually an attempt by ordinary people to fight the speed of AI’s evolution, trying to tame AI, which is doomed to fail. Over the past few years, AI-based tools have evolved at an unprecedented pace, constantly pushing the frontiers of technology. Today’s cutting-edge knowledge might be replaced tomorrow by a new technology, rendering it worthless. Trying to keep up with these changes is a complete waste of time for ordinary people.

But this panic among ordinary people is a boon for cloud providers. They can finally sell tokens to individuals in large quantities. Cloud providers are no longer dreaming of endless spending to mine gold—they are now selling this dream through tokens, the shovels, to ordinary people.

When most ordinary people lose money in this game with nothing to show for it, the game will eventually come to an end. But the end of the gold rush isn’t the story’s conclusion. After the rush, roads, bridges, and towns are already built. Although some leave, population clusters have formed, and various businesses are sprouting on the once barren land. But that’s another story, and the new chapter has little to do with the gold prospectors anymore.