I am forty-two years old. I live in a small municipality in the easternmost corner of Belgium, where Dutch, French, and German-speaking cultures meet. Today I run an AI automation consultancy. My job, in the simplest terms, is to make small businesses more efficient using artificial intelligence and workflow automation. Forged as a digital jack-of-all-trades and completely flourished in AI workflows.

I am also, according to the Belgian pension system, supposed to retire at sixty-seven. That is twenty-five years from now. Twenty-five years during which, if current trends continue, artificial intelligence will have fundamentally reshaped what work means, what employment looks like, and what taxation can fund.

And yet, in Brussels, Berlin or Paris, serious people in serious suits are having serious debates about adjusting the statutory retirement age by one year. They are arguing about whether sixty-seven is too high or too low, about early retirement provisions, about career length calculations — as if the fundamental architecture of the system will still be standing in 2052.

It will not.

The arithmetic no one wants to do

Belgium spends approximately 150 billion euros annually on social protection. The overwhelming majority of this is funded through social contributions — taxes on labour. Every worker who collects a salary feeds the machine that pays for pensions, healthcare, unemployment benefits, and disability insurance.

This system works — as long as there are enough workers paying in. But here is what I see every day from my desk: the number of tasks that require a human worker is shrinking. Not gradually. Not someday. Now.

Every workflow I build for a client eliminates hours of human labour. Every automation I deploy makes a certain category of employee less necessary. I am, in the most literal sense, part of the force that is eroding the tax base on which my own pension depends.

We are arguing about the bedside manner while the patient is bleeding out.

Last month, I built a workflow for a small accounting firm that replaced hours of data entry with a one-minute automated process. The firm is delighted. The three employees who used to do that work are now doing higher-value tasks — for now. But the trajectory is clear. Next year, the workflow will handle those higher-value tasks too. And the year after that, the firm will maybe grow without hiring further.

Multiply this by every small business in Belgium. Then multiply it by the large ones, which are automating even faster. Then add the wave of humanoid robotics that is about to hit manufacturing, logistics, and every kind of manual work. The pension system is funded by a tax on something that is becoming scarce: human labour.

The debate that is not happening

What frustrates me is not that politicians are wrong about pensions. It is that they are having the wrong conversation entirely. The question is not whether the retirement age should be sixty-five or sixty-seven or sixty-eight. The question is: what happens to a social security system built on taxing labour when there is dramatically less labour to tax?

This is not a hypothetical. This is arithmetic. And the arithmetic does not care about coalition agreements.

Belgium's social protection spending of 150 billion is more than a universal basic income of one thousand euros per month for every adult citizen would cost. The money is already being spent. It is just being spent through a labyrinthine bureaucracy that employs thousands of people to means-test, categorise, and administer what could, in principle, be a single monthly transfer.

I am not arguing for UBI. I am arguing that the fact this comparison is arithmetically valid — and yet politically unspeakable — tells you everything about the gap between what the system can discuss and what it needs to discuss.

Why Belgium, why now, why me

Belgium is not a random place to write about democratic dysfunction. Six governments. Three language communities. A world record for the longest period without a federal government — 541 days, in case you were wondering. If democracy can survive here, it can survive anywhere. But the question this series asks is: should it have to?

I write this not as a political scientist, but as a citizen who has watched politics closely for decades — and never found a home in it.

Not as an expert with a framework to defend, but as someone with no affiliation to protect and no boundary to observe — only a determination, long overdue, to be completely honest towards the topic.

And I have come to believe that the biggest broken workflow of all is the one that governs how eleven million Belgians — and by extension, billions of humans — make collective decisions about their future.

This is not a blog about pensions

This is a blog about a democracy that was built for a world that no longer exists. A system of governance designed for an era of scarcity, slow communication, and manageable complexity — now operating in an age of abundance, instant information, and problems so interconnected that no single ministry, no single election cycle, no single human mind can fully grasp them.

Over the coming months, I will write about my thoughts on the structural failures of democratic governance, about the voters who enable them, about the parasites who profit from them, and about the technologies — particularly artificial intelligence — that could help fix them. I will write about Belgium, because it is what I know, but the arguments apply everywhere.

The pension debate was my wake-up call. I hope this blog is yours.

Welcome to The Unfinished Democracy.

‍