A reorganisation of productivity, happening quietly
Originally written in Chinese. Translated with AI, reviewed by Gary. Read the Chinese original →
I just saw the news: Block has cut 40% of its staff in one go, from 10,000 people to 6,000. The computers in the office are still on, the coffee is still warm, but the people are gone.
If the story stopped there, it would just be another ordinary tech layoff headline. What's really thought-provoking is something else: Block isn't a loss-making company. On the contrary, it makes money, and its profits look decent.
The market's reaction was just as direct: the share price went up.
This scene is quietly announcing something: the business instincts many of us carry around are starting to date.
From farmers to tractors
For a long time, the business world had an almost default path. When a company does well, it hires more people. As the team grows, scale naturally follows. More people meant more capacity, and made the company look more "like a big company". That logic held in the industrial age and in the early internet era.
But AI is bending that path, bit by bit.
A more intuitive metaphor: you used to need a hundred farmers for a hundred acres. One hoe each; people were the capacity, people were the scale. Then the tractor arrived. The same hundred acres, three people, a few machines, done in a day. The field is the same field, but what "people" mean has changed.
And AI is harsher than the tractor.
A tractor only amplifies physical strength. AI amplifies cognitive work that used to require people: decisions, content, customer service, analysis. It doesn't just save a company a bit of effort. It rewrites the function of output per unit.
So when a company is already profitable and can use AI to push productivity per person up another notch, the capital markets don't see risk. They see profit elasticity.
Ripples below the surface
Many people are still staring at the word "layoffs", but the bigger shock is slowly spreading below the surface.
Imagine a company that had 10,000 employees and 10,000 parking spaces. Now there are 4,000 fewer people. The first to be hurt may well not be the company itself, but the office building's parking revenue, the coffee shop downstairs with its steady daily cups, the economy-rice stalls that are busiest at lunch, the insurance agents who sell corporate group cover, even the valuation model of an entire row of commercial property.
AI never affects just one industry. It's more like a stone dropped into water, rippling outward ring by ring. Many businesses that look like "steady rent collectors" today are fundamentally built on the density of people. Once productivity per head keeps rising and headcount keeps falling, the whole surrounding economy gets repriced.
Over the next year or two, we'll most likely see several completely different kinds of companies existing side by side. One kind has truly embedded AI into its workflows: its cost structure has been rewritten and output per person has made a visible leap. Another kind looks very active and uses plenty of tools, but has essentially just swapped Excel for an AI chatbot; the production function hasn't changed. And a good number have simply added an "AI strategy" page to their PowerPoint while organisation, process and KPIs stay exactly the same.
When the tide rises, everyone floats together. When it goes out, you soon find out who's been swimming naked.
The hard part is organisation, not technology
Look one layer deeper and you'll find that the real difficulty has never been a technical problem. It's an organisational one.
Many managers say they embrace AI, but the moment it means touching headcount, changing processes or rewriting KPIs, resistance appears immediately. Because what it touches isn't tools. It's the power structure, vested interests, the sense of position many people have built over a decade or more.
Another change is already happening that many haven't fully noticed: the squeeze on the middle layer. What AI replaces first often isn't the physical labour at the bottom, but the large number of knowledge-intermediary roles that can be turned into processes and standardised. These people used to be the vital lubricant that kept companies running. Under the new production function, they may well become the most anxious group.
Crueller still, AI won't change all companies evenly. It will only make the small group that was already fast run faster, while the slow ones look more and more cumbersome on the same timeline.
But if you look at it from a founder's position, all of this is clearer than ever.
As AI takes over more and more standardised content production, customer service responses and product selection judgements, the barrier for small businesses really is falling. What truly opens the gap is system capability. In the future, not everyone will be replaced by AI. Rather, the systems that genuinely embed AI into their workflows, their data loops and their decision rhythm will systematically eat the players still stuck in a tools mindset.
This change won't happen overnight. The inertia of most organisations is far greater than we imagine.
But if you've felt a faint unease lately, that's normal.
Because what we're going through is probably not an ordinary technology upgrade, but a reorganisation of productivity that is happening quietly.
And history keeps proving it: the real turning points are usually completed in the quiet.