985 Graduates Returning to Vocational Schools? In the AI Era, Companies Don't Want Diplomas

During the May Day holiday, my cousin called me.

He graduated from a vocational technical school and built his e-commerce business in Shenzhen from scratch—a true doer.

Due to his own educational background, he used to be anxious about his child’s education, wanting them to attend a top university.

But this time, he seemed at peace.

He said: Even 985 university graduates are going back to vocational schools—what am I worried about?

I was stunned.

This isn’t a joke. It’s reality.

985 Graduates Returning to Vocational Schools: Not Downgrading, but Surviving

985 university graduates returning to vocational schools sounds like a punchline.

But this is the market voting with its feet.

Since 2024, many private universities can’t fill their quotas even with lowered admission scores.

It’s not that students don’t want to go to college—it’s that college can’t give them what they want.

Four years of university: learning theories, writing papers, getting a diploma.

After graduation, companies don’t want them.

Why?

Because companies want people who can do the work, not people who can pass exams.

What about vocational schools?

Though quality varies, at least they teach skills.

Welding, CNC machining, e-commerce operations, short video editing.

After learning, you can start working. Companies are willing to hire.

It’s not that vocational schools are better than universities.

It’s that universities have become too weak.

Universities Sell Credentials, Vocational Schools Sell Skills—Both Got It Wrong

The problems with university education didn’t start today.

Hollowed-out teaching, formalistic employment preparation—essentially selling credentials rather than cultivating people.

Professors read from textbooks, curricula unchanged for ten years.

Students scroll their phones in class, cram for exams at the last minute.

Internships are perfunctory, graduation theses copied.

After four years, they have a diploma but can’t do anything.

This isn’t an isolated phenomenon—it’s systemic.

The entire operating mechanism is designed this way.

Universities aren’t cultivating talent—they’re mass-producing diplomas.

Among 23 private undergraduate universities in Guangdong, 14 failed to fill their quotas, with a total shortage exceeding 25,000 students.

This isn’t an exception—it’s nationwide.

Private universities are hit first, unable to fill quotas even with lowered standards.

Without reform, they’re waiting to die.

What about vocational schools?

Domestic vocational schools are also weak.

But the market is forcing them to change.

Companies need skilled workers, so vocational schools teach skills.

Those that teach well have students competing to enroll and companies competing to hire.

Those that teach poorly can’t recruit students and naturally get eliminated.

In the future, a batch of premium vocational education institutions will emerge.

Not because of policy support, but because of market demand.

AI Has Arrived—What Kind of People Do Companies Want?

Now the issue is that AI has arrived.

Companies don’t just want skilled workers—they want AI-literate workers.

What does that mean?

Before, companies hired an operations person who could write copy, run ads, and analyze data.

Now, companies hire an operations person who can use AI to write copy, use AI to optimize campaigns, and use AI to analyze data.

Before, companies hired a programmer who could write code, debug, and optimize performance.

Now, companies hire a programmer who can use AI to generate code, use AI for code review, and use AI to refactor systems.

AI is replacing our work.

Not completely, but replacing routine, repetitive, low-judgment work.

Entrepreneurs are all thinking about digital transformation.

Using AI to reduce costs and increase efficiency, using AI to enhance competitiveness.

In the future, most companies will be “AI+” companies.

Not just “using AI,” but “AI-native.”

The entire workflow assumes AI is present from the start.

What kind of people do these companies need?

What Companies Urgently Need: Not Executors, but Decision-Makers

In the AI era, companies don’t lack executors.

They lack decision-makers who can use AI.

What is a decision-maker?

• Initiative: Not waiting to be assigned tasks, but proactively discovering problems, defining problems, solving problems
• Insight: Not copying past experience, but rapidly iterating, rapidly testing, rapidly adjusting
• Leadership: Not managing people, but creating an open-minded atmosphere where teams are willing to try and dare to challenge

These three capabilities cannot be replaced by AI.

But these three capabilities aren’t taught by universities or vocational schools.

Because these capabilities can’t be learned from textbooks—they’re developed through real practice.

Product development requires building actual products, not course assignments.

Marketing requires running real ad campaigns, not writing marketing plans.

Operations requires actually managing projects, not making PowerPoint presentations.

Project practice and case analysis—this is the only path to cultivating judgment.

Our Company’s Experiment: Reset to Zero, Build an AI Self-Learning System

Our company is conducting this experiment.

When hiring, we don’t look at degrees or certificates.

What do we look at?

Do you know how to use AI? Do you have judgment?

How do we cultivate this?

Three steps.

Step one: Reset to zero.

Forget what you learned in school.

Don’t think knowing Python makes you a programmer, or knowing how to write copy makes you an operations specialist.

In the AI era, these are basic capabilities.

Real capability is whether you can use AI to reconstruct your workflow.

Step two: Build an AI self-learning system.

Not teaching you to use ChatGPT, but teaching you to build your own AI toolbox.

Every role has its own AI workflow.

Operations has operations agents, development has development agents, design has design agents.

When you learn to create various AI agents to assist yourself, you’re halfway there.

Step three: Project practice and case analysis.

No toy projects—real projects.

After completion, review.

What went well, what went poorly, how to improve next time.

Break down every win and loss into reviewable structure.

Don’t rely on inspiration—rely on muscle memory.

Trainable, iterable, reviewable.

Rethinking Education: Teachers Aren’t Couriers, Students Aren’t Test-Takers

Our education model must be reconstructed.

Teachers should be spiritual guides, not knowledge couriers.

Knowledge? AI explains it better than you.

A teacher’s value is guiding students to think, sparking student curiosity, helping students build judgment.

Not reading from textbooks, but igniting flames.

Students should be direction explorers, not rote learners.

Problem-solving? AI does it faster than you.

A student’s value is exploring directions, defining problems, creating value.

Not memorizing answers, but finding questions.

If we continue with force-feeding discipline, it’s equivalent to “harming lives for profit.”

Lu Xun said, save the children.

Today, I also want to say, save the children.

Not save them from being replaced by AI.

Save them from being destroyed by the wrong education model.

In the AI Era, Diplomas Are Worthless—Judgment Is Valuable

985 graduates returning to vocational schools isn’t a joke—it’s a warning bell.

Who is it ringing for?

For universities: If you don’t reform, you’ll truly only be selling credentials—and that won’t last long.

For vocational schools: You have an opportunity, but don’t repeat universities’ mistakes.

For companies: Redefine your hiring standards—don’t just look at diplomas.

For students: Learn to use AI, build your self-learning system, don’t wait to be spoon-fed.

In the AI era, diplomas are worthless.

Judgment is valuable.

Professional depth + AI leverage + judgment = moat.

This is the reality we all must face.

Not the future—now.

Follow Now

More AI insights and deep thinking