Sam Altman Is Wrong: Parkinson's Law Will Protect Jobs from AI

Sam Altman loudly claims that AI will steal jobs from almost everyone. But this prediction overlooks one important law that people discovered long ago — Parkinson's Law. It states that work expands to fill the time allotted to it. Technology can make a process ten times more efficient, but the organization doesn't shrink — it simply invents new work for the same people.

How Parkinson's Law Works

British historian Cyril Parkinson published his observation in 1957, which seemed paradoxical. He studied the British Admiralty and discovered something strange: the number of officials was growing even as the number of warships was declining. During Her Majesty's reign, the fleet was shrinking while the administrative staff was expanding. The conclusion: the number of employees grows not depending on the volume of work, but according to completely different laws.

The mechanism of this phenomenon is simple and unpleasant. A person who fears losing their job begins to create work for themselves. They write reports about reports, hold meetings to plan meetings, create processes to manage processes. This is neither laziness nor cryptic employment — it's an instinct for survival in uncertainty. The busier I look, the harder it is to fire me.

When AI comes to a company and automates half of a department's tasks, nobody leaves. Instead, a new front of work emerges: AI implementation, AI monitoring, data preparation for AI, verification of AI results, integration of AI with legacy systems, documentation of AI usage. Work doesn't disappear — it transforms. Technology creates an illusion of its own necessity in management.

Organizations live like organisms

Companies have their own immunity. People inside the system know perfectly well how to survive and how to resist rationalization. There are many ways and all of them are used simultaneously:

• Bureaucratize the process — approval used to take three steps, now it will take nine with justifications for each
• Expand the sphere of influence — the analytics department becomes the "Strategy and Advanced Analytics" department with a new hierarchy
• Create coordinating positions — we need a manager of managers, a coordinator of coordinators, a head of the optimization department
• Multiply tools — if one tool increases productivity, then two tools will increase it even more; the main thing is to find a way to integrate them

In fact, companies often buy and implement technologies not for productivity, but to keep people occupied and justify their existence. This isn't visible in quarterly reports, but it works. Reports only grow.

The true purpose of any large organization is not profit, but to

create a safe place where its employees can comfortably wait until retirement, working as little as possible.

What this means

Altman isn't wrong that AI can do work better and faster than humans. His mistake is elsewhere — he overestimates the rationality of organizations and underestimates their resilience. Companies are not machines for optimization; they are living systems where people protect their positions using inertia, bureaucracy, and creative imagination. AI won't steal jobs as long as jobs remain a way to pay people salaries. Instead, AI will add a new layer of management, a new type of documentation, a new AI oversight committee. It will be sadder than outright layoffs, but more realistic.