Aline Muylaert

The through-line

The recurring position is that democracy has been defined too narrowly and that technology's job is to widen it. "Democracy is not only about votes." [1] Voting every few years is a thin form of participation; the argument is for a continuous process, running between elections, held up by digital tools rather than by the occasional town hall [1]. The mission statement is deliberately plain: "Use technology to bridge the gap between local governments and citizens." [1] The gap is real on both sides. Citizens, younger ones especially, expect transparency from government as a default, while the channels available to them have not kept up [1].

On why the old participation channels fail

Physical meetings and paper surveys are treated as structurally inadequate rather than merely old-fashioned. They are non-interactive, and they reach only a small and unrepresentative slice of the population, which means the input a government collects through them tells it less than it appears to [1]. That failure matters more now because the expectation of transparency from government is rising among younger citizens at exactly the moment the traditional formats are least able to deliver it [1].

On making input legible: verification and feedback

Two mechanics are singled out as the difference between participation that works and participation that decorates. The first is citizen verification through digital signatures, described as a key success factor: when citizens identify themselves, the government knows who is actually participating, and input that carries an identity is more likely to be acted upon [1]. The second is transparent feedback, framed as a straightforward matter of incentives. If citizens do not know what happens with what they submitted, they have no reason to submit anything again, so governments should commit publicly to a timeline and a process before they ask [1].

On what AI can read out of citizen input

At volume, citizen input becomes unreadable by hand, and this is where AI and natural language processing earn their place: clustering large quantities of contributions and surfacing the structure inside them, including how different demographics think differently about the same policy question [1]. The Marche-en-Famenne square designs are the concrete case. Input on the plans exposed a clean generational split, with older residents wanting parking spots nearby and younger residents holding different priorities [1]. Cases from Vilvoorde and Vancouver sit alongside it as demonstrations of the same approach at local government level [1].

On digital exclusion

The enthusiasm comes with a limit attached. Digital participation carries a real risk of excluding minorities and the digitally excluded, so online tools should complement offline participation rather than replace it [1].

Takeaways

• Treat democracy as continuous rather than episodic: "Democracy is not only about votes," and technology exists to sustain participation between elections [1].
• Physical meetings and paper surveys are non-interactive and reach a small, unrepresentative group, which is why the input they produce misleads [1].
• Verify participants through digital signatures; when governments know who is contributing, the input is more likely to be acted on [1].
• Publish a timeline and process up front, because citizens who never learn what happened to their input have no reason to participate again [1].
• Use AI and natural language processing to cluster large volumes of contributions and reveal demographic differences in opinion, as in Marche-en-Famenne where older residents wanted nearby parking and younger residents wanted something else [1].
• Keep offline channels running alongside online ones so that minorities and the digitally excluded are not designed out [1].