For Switzerland, the question of the future of democracy is existential. No other country in the world builds its identity so strongly on the direct co-determination of every citizen, on the control of power by the people. We vote quarterly. On laws, constitutional amendments, taxes, infrastructure. We decide for ourselves.
But what if an algorithm demonstrably makes better decisions than the electorate? Does one then vote on the algorithm? Or does one let the algorithm vote?
In this scenario, democracy does not die through a coup. It dies through irrelevance. [1]
Elections take place, but the real decisions are made by algorithms:
Elected politicians pass laws -- but the laws have no effect because technology is faster than regulation and corporations more powerful than states.
Iain Banks' Culture series describes a civilisation governed by superintelligent AI beings, the "Minds". The people in the Culture live in abundance and freedom. They can do whatever they want. They do not have to work. They have little to decide. [2]
Banks himself regarded the Culture as a utopia. But the disturbing question remains: is a society in which artificial intelligences make all the relevant decisions still a democracy? Even if the decisions are objectively better?
Switzerland's direct democracy is based on the premise that the people are capable of judging complex matters. The EEA vote of 1992 was an act of sovereignty. A vote on Google's algorithm is unthinkable -- not because it would be forbidden, but because nobody understands the algorithm.
In the worst-case scenario, three to five technology corporations control the AI infrastructure: the models, the data centres, the data. They offer their services for free -- just as Google offers search for free, just as Meta offers social networks for free. [3]
Free means: you are not the customer. You are the product.
Switzerland has built its identity over centuries on four pillars:
All four pillars are threatened by an AI-dominated world:
Your smartphone knows where you are. Your bank knows what you buy. Your health insurer knows how healthily you live. This data, combined with AI, enables a level of control that Stalin could only dream of. And the transition from "AI helps you" to "AI decides for you" is seamless. [4]
Digital Stalinism needs no gulags. It only needs convenience and ignorance. It needs people who voluntarily surrender their freedom -- in exchange for the comfort of no longer having to think for themselves.
Two of the nine political demands directly concern democracy:
Algorithmic transparency: All AI systems that prioritise, rank or filter content in the public information space must disclose their criteria. An algorithm that decides which news 300 million people see makes more political decisions every day than any parliamentarian. [5]
Ban on AI mass surveillance: AI-powered biometric real-time surveillance in public spaces must be prohibited in the Federal Constitution -- without exceptions.
The EEA vote of 1992 was an act of sovereignty. The question for the next generation will be whether such a vote is still possible -- or whether the algorithms have long since decided.
[1] Harari, Yuval Noah: Homo Deus: A Brief History of Tomorrow. Harvill Secker, 2016.
[2] Banks, Iain M.: The Player of Games. Macmillan, 1988. On the role of the "Minds" in Culture society.
[3] Zuboff, Shoshana: The Age of Surveillance Capitalism. New York: PublicAffairs, 2019. On data extraction by technology corporations.
[4] EU AI Act (Regulation EU 2024/1689), entered into force 1 August 2024. Provisions on biometric surveillance.
[5] AlgorithmWatch: Automated Decision-Making Systems and Discrimination. Berlin, 2023. On the transparency of algorithmic decision-making.