Last night, watching the Spring Festival Gala, I saw a dense row of robots standing on stage—dancing, turning, syncing to the beat, their movements more precise than many human performers.

In the past, when we saw robots in the news, they were mostly on factory assembly lines—welding,搬运, assembling. Now, they stand center stage, facing a national audience, becoming part of the show.

For a moment, I felt a bit dazed.

Not because the robots danced so well, but because I suddenly realized something: years ago, when we talked about “automation,” it was to make people’s lives easier. But today, when we talk about automation again, its true endpoint is actually—unmanned operation.

This isn’t a sensational claim, but a path we’ve been walking for over two hundred years.

During the First Industrial Revolution, the steam engine replaced human muscle. A machine that once required a dozen people to pull could now be handled by a single steam engine. Back then, the anxiety was, “Will manual labor disappear?” Then came electricity, then the assembly line—Ford’s standardized production turned people into part of the process. Later, in the information age, ERP, CRM, and automation systems emerged, and managers began pursuing “process standardization and data visualization.”

Every technological advance, on the surface, is about efficiency. But at its core, it’s always doing one thing: reducing dependence on people.

You can understand this as a simple logic in management—people are the most unstable variable. They get tired, get sick, get emotional, quit, and ask for raises. Machines don’t.

So from a boss’s perspective, the core of automation isn’t “cool” or “flashy”—it’s two words: certainty.

What does certainty mean? It means predictable costs, predictable output, and controllable risk. You don’t have to worry every day about a key person suddenly leaving, or about a mistake in some process due to human error. Once the system is built, it runs steadily.

In the past, we said, “People are a company’s most important asset.” That’s emotionally true, but from a business logic standpoint, it’s not entirely accurate. More precisely, people are the most important creative asset. But at the execution level, companies are always looking for ways to minimize the parts that “must rely on people.”

Why do I feel so strongly about this issue right now?

Because this wave of AI—especially large language models and AI-assisted programming—is no longer just optimizing processes. It’s beginning to touch the boundaries of “knowledge work.” In the past, automation replaced physical labor. Now, it’s starting to replace some cognitive labor.

You’ll notice a phenomenon: many companies are no longer rushing to hire. Instead, they first ask, “Can this be solved with AI? Can we restructure the process so the system runs itself? Can we even design it from the start to be ‘unattended’?”

This is different from the “internet thinking” of ten years ago. Back then, we talked about connections, traffic, and platforms. Now, we talk about models, data, computing power, and agents. The core change is that machines are no longer just executing clear instructions—they’ve begun to possess a certain degree of “judgment ability.”

This makes “unmanned operation” logically feasible for the first time.

Of course, there’s an easily misunderstood point here. Unmanned operation doesn’t mean “no people at all.” It means “the system’s operation no longer depends on specific individuals.” People are more like designers, rule-makers, and boundary managers, rather than daily executors.

If a system hasn’t been changed in three months, it’s not because “no one is working,” but because it was designed to require minimal human intervention. However, when the environment or requirements change, people still need to step in. The difference is that people have shifted from being “cogs in the machine” to being “architects.”

Behind this is actually a logic similar to what Drucker described as “making knowledge workers productive”—the truly valuable thing is the ability to define problems and restructure processes, not to execute repeatedly.

So when I say, “The ultimate destination of automation is unmanned operation,” I’m not advocating for layoffs or stoking anxiety. I’m acknowledging a trend: the ultimate form of an enterprise will inevitably be a highly systematized, low-labor-dependent organization.

From a boss’s perspective, this is an almost irresistible direction.

If your competitor can produce more stable products with fewer people and offer lower-cost services, it’s hard to insist, “We just need more people.” The market will make that choice for you.

So what about people?

My own thinking is that what will truly be scarce in the future is not “people who can work,” but “people who can design unmanned systems.” Not “people who can write code,” but “people who can define problems, break down tasks, and orchestrate collaboration between AI and humans.”

The endpoint of automation is unmanned operation, but the starting point of unmanned operation is a higher-order human.

This is also the direction I’ve been thinking about and adjusting toward recently. Inside our company, we ask ourselves: Which things must be done by people? Which are historical baggage? Which are done out of habit, not necessity? Can we design our business from the start to be “unmanned by default”?

This isn’t cold-heartedness. It’s a form of long-termism.

Because only when the system is stable enough and self-driving enough can people be freed from tedious tasks to do truly value-creating work. Otherwise, we’re forever trapped in daily operations, running ourselves ragged.

The robots at last night’s Spring Festival Gala were just a symbol.

What’s truly worth paying attention to isn’t how well they danced, but the very fact that they appeared on that stage. It means machines are no longer just backstage tools—they’re stepping into the spotlight, becoming part of how society operates.

This path of automation won’t stop. And it won’t reverse.

The question has never been, “Should we go unmanned?” It’s, “Will you be eliminated by unmanned operation, or will you participate in building the unmanned system?”

For me, the answer is already clear.