Skip to content
Articles
ArticlesAI translation· 21 July 2026

For a New Social Digital Economy

We need public-service data centres

by 5 min read

We are facing a paradoxical problem that concerns anyone who thinks about the future of our digital society. On the one hand, artificial intelligence and algorithms are automating more and more jobs, from data processing in public administration to creative and analytical professions. On the other hand, the physical infrastructure of this AI is growing at breakneck speed: gigantic data centres are springing up like mushrooms.

The problem? These two developments are currently on a collision course with local prosperity.

Let's take a real example, modelled on the dynamics in cities like Abilene, Texas: over the course of a decade, a mid-sized city loses hundreds of routine office jobs to clever AI software. At the same time, a tech giant builds a multi-billion-dollar AI data centre in that very city. City hall celebrates the size of the investment. But after a short construction boom, what remains is a facility that consumes vast amounts of electricity, land and water, yet permanently employs only 50 to 100 people.

The value created by automation flows out globally, while the social and infrastructural costs (rising rents, strained power grids, lost jobs) stay stuck locally.

The End of Classic Redistribution

To understand why we won't solve this problem with old methods, we need to take a step back. As I recently described in 2026 – the Tipping Point of the Social Innovator's Dilemma, we are currently witnessing the final decoupling of value creation and distribution.

2026: The Tipping Point of the Social Innovator's DilemmaDocIsInDaHouse

Until now, the dogma was: prosperity is redistributed through human labour and the taxation of it. That was the foundation of the classic social market economy. But this year we have reached the Tipping Point. Capital has become footloose, and AI is pushing labour costs in the white-collar sector down to the level of pure energy costs. Kilowatts instead of Christmas bonuses.

When human labour is no longer the primary means of creating value, the historical social market economy loses its basis for redistribution. We have to update it. What we urgently need now is the foundation for a New Social Digital Economy.

The Future Scenario: A Dual System for AI

As Ludwig Erhard knew, a functioning market needs both the free scaling of the economy and strong social balance. For the social digital economy, this means:

We have to stop treating server farms as purely private real estate projects.

As a future scenario, we need a “public-service” foundation in addition to the private-sector infrastructure. A dual system that reconciles the incredible scaling engine of the private AI giants with strong, local public provision.

In the 20th century, our society recognised that information is too important to be left solely to private profit interests. So it created the dual system: alongside private broadcasters, public-service broadcasting was established to guarantee a democratic basic provision. Today, in 2026, the critical core resource is compute: computing power and machine intelligence. If this infrastructure remains exclusively in the hands of a few global hyperscalers, we give up our ability to shape things.

This Datacenter as a Public Service concept for the social digital economy could rest on three pillars:

1. “Basic Provision of Intelligence” (The Public Compute Space)
Computing power today is like water or electricity: whoever lacks access gets left behind by society. In a dual system, computing power becomes a digital commons. As part of building permits, private data centres agree to hand over a fixed percentage of their capacity. In parallel, independent public-service data centres are built. Together they form a “public compute space”: their mandate is not profit but providing highly secure AI capacity to citizens, schools, municipalities and start-ups. Anyone who loses their job to AI finds a non-commercial infrastructure here to learn new skills and build their own ideas, without being defeated by inflated cloud fees.

2. The Local “Digital Dividend” (A Solidarity Pact for Resilience)
Private data centres remain, but a public-service pact holds them to their social obligations. Instead of classic tax breaks for infrastructure that creates hardly any jobs, part of the revenue (or a usage fee for the enormous consumption of resources) feeds a municipal fund. When case workers are rationalised away by the AI running in that very data centre, this fund pays for retraining in future-proof, empathetic or skilled-trade professions that make up genuine public provision.

3. Thermodynamic Symbiosis: Common Good Instead of Wasted Heat
When algorithms work, the hardware gets hot. In 2026, this enormous energy can no longer simply dissipate into thin air. Private and public facilities could be legally coupled to local district heating networks. If AI is already upending jobs, then in return the data centre fulfils a public-service supply mandate and subsidises the heating costs of the affected population. Private tech infrastructure thus gives rise to public provision that is actually lived on the ground.

Infrastructure for Participation

We must not fear radical change in our society; we have to actively shape it. Stopping technological change is neither possible nor desirable. Freeing people from repetitive tasks holds enormous potential.

But humanity and foresight demand that, while we're at it, we repair the mechanics of how we distribute prosperity. The machines will take over tasks. But in a genuine Social Digital Economy, the physical spaces where these machines stand become our allies.

It is time for us as a society to stop acting like blind consumers or frightened victims of optimisation when it comes to AI. We need a public-service pact for the age of AI: infrastructure for participation.

If we don't develop new concepts such as this New Social Digital Economy, we could, to borrow the words of my Tipping Point piece, become the Kodak of the history of civilisation.

Live long and prosper 😉✋

▶Listen