Yesterday, while introducing our product to a friend, I suddenly realized something.
Foundational technology is evolving fast—how fast? By the day. But the speed at which the general public—and even many white-collar professionals—adopt these things is far slower than imagined.
Take a recent example. OpenClaw went viral, and everyone was drawn to this technical breakthrough. Someone asked me: "You use Claude Code? That's for writing code, right?"
I said yes.
Their first reaction was almost identical: "That sounds complicated. I probably won't need it in my lifetime."
Not What You Think
I demonstrated in the most intuitive way: installed in 5 minutes, ready to go with just an account, as long as you can talk.
They stared at my terminal for a moment and fired off a series of questions. Is the installation tedious? Is it difficult? What problems can it solve?
I said, installation is just one line of code, 5 minutes. An account costs just a few dozen yuan per month to get started.
They were stunned. "Is that really it? I assumed something this powerful must be complicated to install."
Not at all.
What's even more surreal is OpenClaw itself. Online, there are 15-step, 20-step installation guides that look headache-inducing. Many people complain: the "crayfish" (OpenClaw's nickname) works great, but installation is too difficult. On Xianyu (a Chinese secondhand marketplace), services charging 500 yuan for on-site installation even appeared.
I demonstrated right then and there. When you have Claude Code installed on your computer, you only need to say one sentence—"Help me install OpenClaw on this computer."
It will find out what OpenClaw is, where to download it, and how to deploy it all by itself. You tell it what the Feishu (Lark) App ID is, and it helps you configure and test everything until it's running. You go grab a coffee, and by the time you're back, it's deployed.
One line of code, 5 minutes. That simple.
Why Is AI Coding a Meta-Ability?
Someone asked me: "Does your software have any out-of-the-box applications? Like text-to-image?"
I didn't answer directly. I was thinking about something else.
If we let everyone master an efficient AI coding tool—a coding agent—you want something, you have it install it for you.
Text-to-image? You say: "I want a text-to-image application, download the model and deploy it locally so I can generate images with a single sentence." It might take an hour or two to help you download the model, deploy it, and get it running.
This ability has already transcended the concept of an "application."
It is helping every ordinary person—at an extremely low barrier to entry—truly take control of their computer. By "control," I mean at a level surpassing that of 90% of engineers and senior engineers. And your computer is connected to the internet.
So the old idea of "I need to install an out-of-the-box application" is no longer necessary. You need something, you tell it. It does it for you.
What Can It Do?
Honestly, at this point, if you asked me to use a computer without software like this, I wouldn't know how.
It can unlock the 90% of capabilities on your computer that normally go unused. This is also why edge AI devices should run Linux—Linux is inherently designed for programs to control other programs, and agents experience zero friction on it.
You have 20 Excel files to process? Put them in a folder, toss it to the agent, and tell it how you need them analyzed. It will invoke Python, download plugins if they're missing, and solve problems on its own.
Want to implement some interesting networking features? It can help with that too.
As long as the model is strong enough and fast enough, things that previously required professional skills become within reach.
What Does the Future Look Like?
In the future, people will become increasingly lazy about using computers, unwilling to manually operate them. I rarely actually operate my computer anymore—at most, I open a few Claude Code instances and switch between them to assign different tasks.
The way we work will change.
Before: Human → Tool → Computer.
In the future, it will likely be: Human → Agent → Computer.
What's the point of building a complex interface? It's a burden for AI, and humans don't look at it either. Just give instructions directly to the agent and let it execute. Done.
Software infrastructure may change. I'm not sure exactly what it will look like, but the general direction is probably this.
The Difference Between Claude Code and OpenClaw
Many people ask me this.
Strictly speaking, OpenClaw puts a shell around Claude Code's capabilities. Some call this shell the "crayfish shell"—I think that makes sense.
What Claude Code does: you give it a vague instruction, and it executes end-to-end, largely meeting expectations along the way without needing you to correct its course. It can run for tens of minutes or even hours, and finally gives you a result.
What OpenClaw does: it adds another layer on top—distributing instructions, managing results, forming long-term memory, acting like a project manager or supervisor. It can interact with you at a higher level, becoming your personal assistant.
But the prerequisite for all of this is that the underlying coding agent executes each task beautifully, without you needing to watch over it.
This is also why OpenClaw connected to different models is an entirely different species.
Currently, there aren't many models that can truly "execute well from a vague instruction." In my view, only Claude's Opus and possibly Codex's latest model can do it. Versions connected to other models feel like toys—each execution is unreliable, making the coordination layer above even more difficult. However, model progress is faster than expected, and the million-level gap in this area is rapidly narrowing.
The reason OpenClaw wowed everyone is because it was connected to Opus. But Opus's coding plan is no longer available—too expensive.
This is the current pain point. I don't know when it will be resolved.