Blog

I posted yesterday about AI and taxation and I’ve been thinking further about AI’s societal impact, as well as reading the scenarios produced by the Windfall Trust.

That helped me realise that we are at an unusually powerful moment for well-designed citizen participation work to have impact on the future of AI and democracy.

Looking at Windfall’s scenarios, it is clear that no one knows quite where are going to end up. Unlike climate change – serious, slow, clear direction of travel – the AI transformation is moving quickly but its long-term results are still uncertain.

Speciflcally, we don’t know for sure who the ultimate winners will be.

Will OpenAI and Anthropic corner the market in advanced models, or will open weight and open source solutions destroy their business model with good-enough alternatives?

Will businesses profit massively from firing junior staff and automating coding and documentation processes, or will they need to hire more people to manage the hugely increased flow of AI-generated documentation?

Will workers find themselves commoditised, sidelined or unemployed, or will the new threat to jobs re-energise trade union and worker movements to ensure labour increases its share of GDP?

That uncertainty is what makes the moment powerful. The future coalitions have not yet formed, people and organisations have not entrenched their positions – we know that things will change, but we don’t know how they will change.

So we are in a Rawlsian moment. John Rawls proposed a philosophical approach to justice called “the original position” – what sort of society would you design if you were behind a “veil of ignorance” and didn’t know what position in society you would end up in?

That’s where we are, and it’s where Sam Altman is, along with every worker or community that might see disruption from AI. So now is the time to create the conversation about what that future might look like, and build the structures of resilience that make it more likely to come about.

As a democracy guy, I’m interested in how the economic and social changes AI will bring about will flow through into politics and democracy – whether through the loud voices of very rich people, or the collective voices of community.

Many of us working on democratic innovation are focused on using AI for participation events, or are concerned about bad information spread by AI. Others, including my colleagues at Democratic Society, are working on how to create good governance arrangements for AI use.

In this Rawlsian moment, though, we have an opportunity to think about what desirable post-AI futures might look like, and what investment we need in democratic and information infrastructures now – before the real social and economic impact hits, and when positions have not yet completely hardened.

There are lessons to be learned from another equally disruptive but slower-moving process – climate change. In my next post, I’ll think about how citizen participation in just transition thinking can and cannot be a model for shaping a democratic future after the AI transition.

A US-focused book but seeking to draw lessons from Europe and elsewhere on how to reverse what the authors see as the structural and constitutional bias towards minority government in the US.

The minoritarian institutions that they discuss are the Senate (equal representation overrepresents small states), the Supreme Court (appointed by often-minoritarian Presidents and confirmed by Senates that are usually not representing majorities), and the House of Representatives (seats being gerrymandered).

European constitutional flexibility is praised, with examples from Norway to Belgium, but the structures of the EU which risk repeating the mistakes of the US Constitution (particuarly in its inability to undertake constitutional reform) go unmentioned.

For policy makers:

• A 15-point plan for anti-minoritarian reform, useful in the US context but not so relevant for Europe.
• The need to maintain constitutional certainty and flexibility in balance.
• The need to protect the rule of law.

For citizens:

There is reference to movements for democracy alongside the existing civil liberties etc associations, but these are not as focused on constitutional change – the challenge as ever is how do you get the parties who are self-interested to change their prioritisation to handle constitutional reform.

Reading: Eric Corijn’s “De Verenigde Steden van Europa : ter verdediging van de democratie” (The United Cities of Europe to the rescue of democracy).

Overall, an argument that I like. Cities are spaces where national stories are quieter and local needs predominate. They are more multicultural than the average of the countries they are situated in. In an era where nationalism and division is rising again, cities can be the foundation of an effort to bring people and cultures together, and create a post-national democracy.

In Europe, specifically, there is an opportunity to create Europe from the cities, and do that founded on local democracy and democratic innovation, rather than getting stuck at the level of member states and their internal politics.

Brussels, in particular, is pointed out as the potential archetype, a city that is already hyperdiverse and needs to overcome the history of Belgian communitarianism (also written into its electoral system) to truly establish itself as the capital of Europe. This is something I’ve thought myself in the past, but the connection between Brussels, Belgium and Europe is more complex than that – something that needs more reflection.

The big missing point in Corijn’s argument is the rural areas. A large share of Europeans live in cities, but many live in rural or periurban areas which are not properly connected into the urban core. Take, for example, the statement of the mayor of Grimbergen, by any measure part of the Brussels economic area but just across the border in Vlaams-Brabant. When the plans for a tram from Grimbergen into central Brussels were cancelled, she celebrated, saying “it’s all very well to be able to get into Brussels easily but what will come out”. That “oil stain” effect of Brussels diversity into the surrounding suburbs and countryside is a significant point of resistance, and of nationalism. We need an approach to saving democracy that does not treat those areas as afterthoughts.

A brief braindump prompted by hearing talk of an AI minister. Better argued thoughts might follow later.

Thinking about “AI ministers” (as in, ministers who are AIs, not ministers for AIs) and citizen assemblies and increasingly convinced that a lot of democracy innovation at the moment is forgetting the perspective of any-given-citizen.

Not the theoretical citizen, averaged out across a thousand assemblies, but without real power in any of them, and with no choice about the issues he or she is asked to deliberate upon.

Not the citizen as policy consumer who can be as well served by a digital minister as by a human one.

But any-given-citizen, a real person who has a certain vision of the world, a certain set of things she or he is annoyed about, and wants to be able to do something – to have some agency.

They have a vote, which we can make more effective. They have a voice, which we can make more powerful. They are a customer of government, which we can make more responsive and more open.

But alongside those already huge challenges, we have to recreate the connection that they used to have through mass parties, unions or churches, from the life that they are living to the decision making level of politics.

I once thought that social media and digital platforms could be that route. I don’t think that any more – or at least, not ones that look like the places where people live online today.

But we need to empower any-given-citizen, not the theoretical citizen. We have to empower them across whatever issues they care about. And when they express themselves we have to have humans listening to humans taking decisions about humans.

FAZ is reporting on a campaign by the Left party for green energy that uses far-right-themed AI imagery and language to make its case.

Like the example above, slogans include “German energy from German sun”, and “German power from German wind” – but most strikingly, the art work uses a visual language strongly associated with fascism and the 1930s.

The argument that renewables are local and independent is a good one – though of course, “German” energy is as much European energy, given the interconnected nature of the grid.

However, making it in this way is a reminder that we are moving – as Alec Ryrie says – beyond the Hitler age. It’s not enough now to consider fascist language and imagery automatically disqualifying. Fascism is not yet used with pride, but adjacent concepts are, in some quarters.

As someone who grew up with older parents who had direct memory of the war, it is hard to think yourself into this position, but if you imagine something like the Boer War, perhaps you approach it.

Hitler has a resonance in history that Redvers Buller and Henning Pretorius will never have, but still, the issues at stake in the war are fuzzy for most, if they know about it at all. What was so high-stakes five generations ago is now a specialist history subject.

We aren’t there yet with WW2 and Hitler, and maybe it will be another hundred years before we are, but the Left’s campaign idea shows that even for those on the opposite end of the political spectrum, the imaginary language of the far right is a little marketing trick, not a taboo.

It challenges us on the democratic side of the spectrum – what are our images? What is the visual shorthand for democracy, inclusion and rights? We’ve relied for so long on just not being the other side, that we have nothing clear or identifiable to work with. It’s not that “the Left can’t meme”, as FAZ says, but that democracy as a process and set of institutional guarantees is hard to picture.

EU stars, the Portcullis of the UK Parliament, hands raised in voting, ballot boxes, post-it notes on a whiteboard – all of them are process or institutional images, and we need to find a way of showing the impact that they have, and the way they enable people to have a voice.