Layoffs Will Continue, But They Solve Nothing

Article

Layoffs will continue, but they solve nothing.

Recently, news about layoffs has been increasing both domestically and internationally.

Overseas layoffs can be described as brazen — after laying off employees, CEOs even come forward to write articles explaining and boasting about their achievements.

Domestic layoffs, by contrast, are primarily addressed through rumor debunking. Information spreads informally among employees, followed by official denials.

I won’t name specific companies, but some reportedly cut 50%, others 30%, and some 20%.

However, most of these percentage figures have already been officially denied — the denials are true, because this is a process; it’s impossible for so many cuts to have been finalized.

The last major wave of internet layoffs occurred in 2022–2023.

At that time, opinions were relatively rational and polite:

Layoffs are the consequence of corporate overexpansion and require mean reversion.

During the internet era, management became accustomed to solving problems by hiring more people: if a project fails, hire more. If two teams compete, hire more. If launching a new business, hire more.

The myth of “man-months” was long ago debunked, yet no one listened.

Eventually, growth plateaued, and personnel became unsustainable — so they were laid off.

That round of layoffs stemmed from management hiring too many people — a decision error.

Though calling it an “error” might slightly insult mid- to senior-level managers at big tech firms. In large companies, headcount often serves merely to pad numbers — perhaps “strategy” is a better term.

I personally experienced that round. The situation was absurd: our entire department was shut down overnight, justified by “research showing users don’t need this service.”

It was a lightning operation — within a single day, the entire department left, and the service was taken offline.

Then something interesting happened: users began complaining. Complaints piled up until the company had no choice but to urgently restore the service.

To this day, that feature remains online.

But everyone who built it has long since left.

This is just one of countless human follies.

How laughable humans are — we love making terrible decisions, then wrapping them in “rational” explanations.

I’ve discussed this insight before:

Humans are creatures driven by hormones but who like to disguise their behavior with rational thought.

Anthropological research shows that human decision-making is primarily driven by hormones; knowledge, experience, and reason play minimal roles in this process. We typically make decisions first, then use intellect to find evidence supporting those decisions. If decision-makers don’t bear the risk or loss of poor decisions, they cannot generate the necessary hormonal responses under pressure — and thus cannot make correct decisions. — *Skin in the Game*

As noted in *The Righteous Mind*, our so-called reasoning isn’t aimed at discovering truth — it’s meant to justify our intuitive, emotional reactions.

CEOs and executives are also human. Their anxiety, pressure from capital markets, and herd mentality (“everyone else is cutting”) reached a tipping point one day. Their bodies made the decision: lay off staff.

Judgment came first — then scholars emerged to justify it.

Looking back today, I believe the real reason behind that wave of layoffs was:

The industry had hit its ceiling — everything was a ripple effect of “the internet is dead.”

The 2026 round of layoffs is different. This time, the justification is: “Agents are immortal?”

One overseas company announced record profits and simultaneous layoffs of 40% of its workforce on the same day, citing AI-driven operational changes. Its stock surged 24% that day.

The irony? The profit surge occurred *before* the layoffs...

I recall my own experience — it also happened after the company’s revenue and profits hit all-time highs.

Blame AI — what a convenient, weight-off-your-shoulders explanation.

Then the machinery kicks in: gather data, calculate per-employee efficiency, identify departments below thresholds. Package it as an “AI efficiency” narrative — for the board, for media, for laid-off employees.

We don’t know the final outcomes, but some absurd things are indeed happening.

One overseas company replaced hundreds of customer service agents with AI — satisfaction plummeted, so they quietly rehired humans.

Two major foreign firms used AI to boost employee productivity — only to discover labor costs spiked, realizing that in many scenarios, AI usage was actually more expensive than human labor.

What a perfect scapegoat — even the laid-off employees believed it.

The narrative of “agents replacing humans” sounds flawless — yet why are teams truly excelling with agents still growing in size?

Has OpenAI laid off staff? Has Anthropic?

If we stop at such superficial explanations, we’ll never uncover the real root causes.

So what are the real underlying reasons?

Let me start with the term “AI native,” which everyone loves discussing these past two years.

I break “AI native” into three layers:

1. AI-native tasks yield 10x revenue.
2. AI-native organizations achieve 10x capability.
3. AI-native talent delivers 10x efficiency.

What defines an “AI-native task”? It’s something that could only exist — and generate 10x returns — in a world where AI exists.

With such a task, a company can design a new organization optimized for 10x efficiency.

With such an organization, talent within it gains space to unleash AI’s potential 10x.

Without layer one, there’s no layer two — and without layer two, no layer three.

Compared to layers one and two, layer three is easiest — and least important. Yet it’s the layer people talk about most: using Claude, Codex, letting AI write code, distilling colleagues’ work, automating operations — achieving 10x efficiency.

Most companies do exactly this: skipping layers one and two, they push middle management to implement layer three — using reduced headcount to “prove” successful transformation.

Let’s set aside the question of how valuable easily automatable tasks really are. Think carefully: if we skip layers one and two and only pursue layer three, then in a 10-person team, if one person achieves 10x efficiency, what happens to the other nine? They’re likely just... laid off?

Even achieving 10x organizational efficiency (layer two) becomes impossible this way.

After all, middle managers can never answer one fundamental question:

Does an AI-native organization even need middle management?

Layer one, however, is the responsibility of CEOs and senior executives.

Finding a 10x-scale AI-native task requires creativity, risk-taking, and redefining the company’s very purpose.

This is a process of reinvention.

Layoffs may continue — but we must first confront the real issues.

During China’s Warring States period, King Wuling of Zhao sought national strength and military improvement by mandating universal adoption of nomadic-style clothing and cavalry tactics.

When King Wuling introduced “Hu clothing and horseback archery,” the entire court of nobles fiercely opposed him. Their argument: “Changing ancient ways violates human nature — ancestral laws must not be altered.”

Behind their opposition lay a refusal to acknowledge a simple truth: Northern nomadic cavalry moved like birds in flight and vanished like snapped strings — chariots couldn’t match them, old methods were obsolete.

The wars he needed to fight had changed — so his tactics must change — so his clothes must change.

Today’s big tech firms, using “AI efficiency” as justification for layoffs, sound like they’re embracing change — but in reality, they’re doing the opposite.

They’re making the smallest possible gesture to pretend change has occurred — avoiding the truly painful question:

The old war is over. What’s the next war?

If you can’t answer that, laying off another 10,000 people won’t help.