Source: in-cyprus.philenews.com
More and deeper democracy means a greater capacity to manage our disagreements. Technology should help us live with them — not make them disappear.
In his final years, the philosopher and theorist of the public sphere Jürgen Habermas — who died last March at a great age — must have been deeply uneasy watching Google DeepMind introduce what it called the “Habermas Machine” in 2024. It is an experimental AI system designed to help groups of citizens arrive at shared formulations and points of consensus after deliberating on political or social issues.
A digital deliberation engine, in other words, that sands down the sharp edges of political disagreement — a practical application of large language models deployed among people with conflicting political views.
The misreading of Habermas is this: he championed an open communicative rationality aimed at achieving mutual understanding, while thinkers such as Jacques Rancière argued the opposite — that democratic practice is defined by public disagreement, which does not paper over inequalities but exposes them.
That is one thing. A not-exactly-insignificant private technology giant getting to design the criteria by which “good” public debate is defined is quite another. And the very spirit of this machine smacks of the technocratic logic Habermas himself frequently criticised. It is anything but a faithful realisation of his theory.
The most striking finding of the experiment was that participants overwhelmingly preferred the consensus texts produced by the AI over those written by human facilitators, rating them as fairer, clearer, and more balanced. If that is not a textbook example of critical thinking in retreat, what is?
Where are we heading, you might ask. Probably exactly where we all fear. The idea that AI could act as referee — tirelessly processing thousands of views to identify points of convergence — might, at a very superficial level, sound almost appealing. But the assumption that consensus is the ideal outcome for a democratic society is far less innocent than it sounds. For a start, deep conflicts of interest and values are hardly going away, while minority voices get forcibly rounded off in the name of unanimity. The more fundamental problem, though, is that democracy requires not just synthesis but confrontation.
Imagine throwing every opinion into a blender and pressing the button. The result would not be a democratic mix. It would be mush. The prospect of algorithmically resolving pluralist political tensions is not democratic in the least. It is rather the opposite. Achieving homogenised consensus through algorithms is a political act of the highest temperature — and the highest danger.
Friction and conflict are in democracy’s nature, given that human values are sometimes genuinely irreconcilable. The prospect of an artificially engineered harmony should frighten us.
The use of technology — and AI in particular — in modern democratic processes is not inherently threatening. Far from it. As long as it does not undermine the nature of political contestation, and does not restrict or dictate the political agenda. If it can shed light on the structure and character of disagreement without replacing human decision-making or discouraging critical participation, then it is worth talking about.
In the Land of the Rising Sun, a new star has recently risen — Takahiro Ano, born in 1990. Perhaps the new political demons now haunting Cyprus have led some of you to notice the case of Team Mirai — Team Future — which used a chatbot to process thousands of voter questions and proposals, helping surface issues that conventional politics had been ignoring.
This seemingly out-of-nowhere party put in a remarkable performance at the last two elections — the 2024 lower house vote and the 2025 Senate race — proposing a new model of “digital democracy,” with a campaign built around an AI avatar of Ano himself, available to answer citizens’ questions twenty-four hours a day.
To head off any lazy comparisons: Ano is not a populist YouTuber. He is an AI engineer, a graduate of the University of Tokyo’s prestigious School of Engineering, and an established science fiction author. His case could, with a little imaginative licence, provide a theoretical framework for examining the Cypriot debate around technologically assisted “direct democracy.”
In Cyprus, the basic — and clearly simplistic — assumption is that if we listen directly to the people, solutions will come more easily. The premise is that citizens know what they want, that parties and institutions merely get in the way, and that the popular will would express itself more clearly and effectively if the middlemen just got out of the room.
Beyond that, citizens with genuinely conflicting interests, values, and visions — on the Cyprus problem, migration, the environment, development, taxation, human rights, Church-state relations, and so on — will somehow find the magic formula through a Cypriot version of the Habermas Machine that solves everything at once. Which means, inevitably, bypassing the very mechanisms that protect dissent and give organised voice to different social groups.
Take the Cyprus problem as an example. If a referendum were held tomorrow, the wisest possible verdict would not necessarily emerge. Because there is no such thing. A painful and extraordinarily complex historical and political conflict gets reduced to a binary answer — and as the traumatic and divisive experience of 2004 showed, regardless of where anyone stands, there was no right answer and no wrong answer. Every vote was cast for the right reasons and the wrong reasons simultaneously.
What I want to conclude is this: the novel debate about direct democracy in Cyprus may present itself as anti-establishment, but it contains a contradiction. Beyond the complexities of society’s many layers and divergent views, direct democracy promises to bypass “middlemen” — but in practice it requires new ones: platforms, algorithms, influencers, social media. Good luck with that.
The will of the people is not singular. It is not even binary. If it were, we would not need democracy at all.