I'd be happy to chat about any cryptographic primitives (signatures, encryption, multi-party computation, voting, etc.)
You’re viewing this forum as a guest. Sign in to ❤️ and comment.
The ultimate collective challenge: organizing human societies within planetary boundaries while leveraging technological potential. This topic examines automation as political technology and ecological thinking as collective intelligence.
-
Bastani, Aaron. Fully Automated Luxury Communism. Chapter 3: "What Is FALC?"
-
Pepper, David. "Utopianism and environmentalism"
-
Ernest Callenbach – Ecotopia
-
"What Is Solarpunk?" solarpunkstories.com/what-is-solarpunk
Even when you never shared any data about you, AI systems trained on the data of your peers can be used to predict your future behavior and interests with ever greater accuracy.
Given this reality, we cannot only consider privacy as an individual concern, without also looking at the collective consequences of each of us making individual decisions about our data.
So it turns out selling to government is hard. Procurement frameworks are a labyrinth, incentives are completely warped, and the sales cycles can outlast your runway. If you're trying to figure out how to navigate the bureaucracy without losing your mind, I've got some scars to share.
In 2016 the European Union adopted the General Data Protection Regulation (GDPR) that seeks to protect the privacy rights of Europeans. However, since the GDPR came into force in 2018, regulators have been slow to enforce it. To ensure that the privacy rights of Europeans are sufficiently protected, a growing number of private actors have taken it upon themselves to bring cases against companies that are in violation of the GDPR.
How does private enforcement work? What mechanisms are available? And how do these differ between different European member states?
How do we calculate the value of privacy violation and specifically the material and non material damages?
We are seeing a growth in transparency and data access laws, to support scrutiny of tech companies. This is particularly the case in the EU, where the Digital Services Act supports researchers (including those outside the EU) to request data from very large platforms and search engines (Facebook, TikTok, Google Search etc.) for researching "systemic risks to the EU". This can, under some conditions, include access to internal company data.
Other rules and other countries, including the AI Act in the EU and implementation of the Data Access and Use Bill in the UK, may also open opportunities for access to platform data. All this builds on existing rules such as Freedom of Information and Subject Access Requests under the GDPR - though these existing rules are also somewhat at threat from political trends.
I can help Fellows who have projects involving access to data from large social media, search, and AI companies consider if and how they could use data access laws; or explain the rationale, debates, and politics underlying these rules and their ongoing development.
Some material for starters:
https://algorithmwatch.org/en/wp-content/uploads/2025/08/DSA-40-12-explainer-2.pdf
I have been developing a project to build a more connected and impactful UK democracy sector for over five years. The UK Democracy Network is a civil society infrastructure organisation, comprising an established and diverse network spanning hundreds of organisations across the country, working on a range of issues from civil liberties to civic education. Our work is based on and works to improve upon existing network development, network weaving and network infrastructure practice.
We started the work based on an acknowledgement that those working to improve democracy in the UK were fragmented, not talking to each other enough and lacking access - and the trust required to develop - shared resources. Based on this, we founded our predecessor organisation, the Centre for Democracy at Newspeak House in 2020 and undertook network mapping and seeding activities. Since then, we’ve built a directory of UK organisations working on democracy, regularly extracted their organisational outputs: newsletters, events, grants and socials - and fed this into our comprehensive sector bulletin, sector resources and the connection and capacity building activities we offer to improve the impact of UK democracy orgs and the vital work they do.
I can share my learnings, practical examples and experiences through undertaking this work. I can help fellows with questions like: what do British democracy organisations care about and work on? How do we build trust in large communities? How do we prevent information and resource bottlenecks in networks? How do you evaluate the impact of network activities? How can I apply network development practices to build trust and collaboration in my sector?
I’ve been involved in building and launching a few interesting pieces of govtech, like GOV.UK Chat, Caddy. If you want to dive into the messy reality of shipping code inside the state apparatus, and what it actually takes to get modern tools into the hands of citizens, let's talk.
One of the big challenges of group decision-making is that group members typically have different needs, interests, preferences, and priorities. Even when groups have perfectly accurate information, and a full set of possible ideas, they may struggle to reach agreement because everybody in the group wants something different.
One simple way to make a decision amidst disagreement is to take a vote. The analysis of the advantage and disadvantages of different voting methods is known as social choice theory, and may be covered in other modules.
However, voting has limitations. For example, it presupposes a relatively limited set of discrete options.
Moreover, it is sometimes desirable for groups to reach consensus. Consensus can be important when it is operationally necessary to have all group members involved (i.e. one dissenter can wield an effective veto) and can also be desirable simply for generating "buy in."
The practice of making decisions amidst agreement is known as negotiation.
It's important to realize that negotiation is not always about wheeling and dealing. Imagine a family trying to agree on vacation plans, a group of freelancers trying to agree on a shared office space, or a team of political technologists trying to design a project together.
Notably, reaching agreement is not only a social problem but a computational problem. It's a matter of searching the space of all possible solutions until you find one that gets everybody to "yes."
Negotiation is a huge topic, and I have a variety of syllabuses ranging from a 2 hour workshop to a full 30 hour university course. An important part of negotiation education is hands on skills practice, i.e. completing simulated negotiation exercises to build your experience and try out different techniques.
A big part of negotiation education is theoretical learning, and you can see excerpts from my MBA lecture videos in my free online course.
Some of that theory translates into direct techniques you can apply, like a negotiator's toolbox. For example, a common technique is "multiple equivalent simultaneous offers" aka MESOs. In this technique, someone will literally make multiple offers at the same time that are of equivalent value ("I can offer you either A, B, or C — which do you prefer?") which offers a variety of tactical and strategic benefits.
A model might be an equation, a map, a simulation, a story or a physical arrangement of stones. A Zen garden is not a landscape, yet it preserves or evokes something important about one. Metaphors work similarly: describing a career as a ladder or life as a journey creates an implicit model with particular structures, directions and possibilities. What happens when we make those assumptions explicit?
This topic weaves around the two core topics, Systems Thinking: Understanding Political Dynamics and Building Evidence: Data Science in the Age of AI. Those topics introduce ways to model systems and build evidence from data; this one turns those methods back on themselves. What assumptions are built into a model? What does it make visible, and what becomes harder to see? The aim is not to retreat from systems thinking or empirical evidence, but to use them more imaginatively and self-consciously.
Play gives us a way to explore and test models without treating them as settled truths. Roger Caillois distinguished competition, chance, simulation and vertigo, alongside a spectrum from spontaneous play to formal games. We can compete models against one another, introduce chance, inhabit them through role-play or deliberately destabilise their categories. Play both creates temporary systems of rules and allows us to bend, replace or step outside them.
I have spent much of my career constructing mathematical and computational models, but models need not begin with mathematics. We would build models of ourselves using metaphors, diagrams, objects, stories or games, then allow other people to interpret and disrupt them. The aim is to become more imaginative and self-conscious model-makers, better able to see what any representation reveals, assumes and excludes.
Core reading
- Roger Caillois, Man, Play and Games (University of Illinois Press, 2001).
- George Lakoff and Mark Johnson, Metaphors We Live By (University of Chicago Press, 1980).
- Ezequiel Di Paolo et al. "Simulation Models as Opaque Thought Experiments" (2000).
There is now a metalayer (metaverse?) of information. Content that feeds on other content, putting new distance between what we deal with online and lived reality. The same infrastructure that lets legitimate distant voices reach us also lets sources completely divorced from reality permeate and subvert the trust networks we once reserved for people and institutions close to us. Our trust circles evolved for a world where proximity was a filter. That filter is gone, and hostile or hollow sources can now occupy the same space as a trusted neighbour.
Key Question/s: When trust networks can be invaded by actors detached from any grounded reality, how can civic tech help people re-establish meaningful signals of proximity, accountability, and authenticity, without simply appointing (or becoming) a new gatekeeper?
Alternative title: “Network Subversion and Me”. If you’re familiar with “Cognitive Security”, then you have a basic handle on this. We can look at approaches for strengthening the integrity of tech-moderated trust networks against infiltration while preserving the openness that makes them valuable. Finding a balance between broad connectivity and local or authenticated networks. We can also look at the kind of social geometry our tech fosters. Both the stuff we use. And the stuff we’re building.
For those looking to spice up their political toolbox. I’ve been part of several direct action groups of varying sizes and approaches, from cycle safety campaigns, anti-fascism and student street movements, though my primary experience is based on my experience as a National Coordinator of Extinction Rebellion in 2019 & 2020.
I'm connected to people organising in that space who have thought about this a lot longer than I have - I can pull in various people currently organising and thinking about this to join us. Together we can discuss and evaluate the effectiveness of various approaches to direct action strategy and tactics, what has worked, what hasn’t - and what I, and others, have learned along the way. Whether that’s the use of mass arrest tactics, disruptive action, small scale targeted action or the challenges of growth, organising and mobilisation in mass movements. We can also discuss the current strategic predicament the current climate movement and the wider extra-parliamentary progressive movement faces. Beyond this, I can also offer specific training in the key skills of direct action: action design, affinity groups, spokescouncils and decision making, action roles, police liaison, legal observing, arrestee support and de-escalation.
Additionally, Newspeak House itself has hosted many such groups and acted as a valuable organising space. Welcoming and inviting the current generation of organisers to the space, to explore, experiment and collaborate would be a worthwhile and fascinating opportunity - if Fellows were interested in this I would be happy to assist and facilitate.
Universities around the Western world are becoming less relevant. Anybody can learn anything online on their own. Yet learning communities remain crucial. They facilitate people meeting, inspiring, supporting each other. Residential communities are a great way to approach this. You are in one this year. Let's start 20 more like it!
Related topics to start from or research:
- Fractal NYC and the Fractal Campus course
- Chautauqua
- Radical Routes
- Feÿtopia
- Mondragon University
Just because data does not always identify people, does not mean it cannot be used in nefarious ways. Even non-personal data requires careful governance. For instance, consider the case of farm data, such as data about soil conditions, crop production etc. In the hands of the farmer, this data could be used to improve crop yields. In the hands of for instance a commodity trader, this type of data (especially in aggregate) could be used to manipulate the commodity market.
Given that non-personal data is not protected by privacy rights, how do we go about governing it? Who decides where this data ends up and who has the right to access this data? Of course, good governance of non-personal data has only become more urgent with the invention of LLMs.
Background reading:
There are defined cryptographic notions of transparency, and there are cryptographic primitives that get used to build systems that provide strong transparency guarantees. There are also successfully deployed technical projects like Certificate Transparency (which all major browsers use every time you visit a website), key transparency (deployed in WhatsApp), and software transparency (Sigstore).
While transparency thus can be a necessary tool for achieving fairness and accountability in complex multi-party systems, it is far from sufficient, both for these more technological systems but especially for more human-oriented systems (e.g., using transparency for governance, as in a DAO).
Whatsapp* has become the default† messaging tool for grassroots organising, whether it's party political canvassing, a local campaign or forming a tenants or workplace union. And Whatsapp works pretty well at the basics: you can be confident that your message is going to be delivered to everyone in your group, and because most people already have it, it's quite easy to start a new one and get people into it. But it's not adapted for the task of running an active democratic collective.
That's why I focus on building tools that work on Whatsapp, not because I like Meta particularly but because that's where people already are. Users are the most important part of any tool, because they can take it from a good idea to something that works really well.
Whatsapp isn't an open source technology and its API is quite expensive. There are tools that connect to it though, using the Whatsapp for Web websocket. That's how the Chat Hackers project that I lead connects our tools to Whatsapp. There's also the Baileys npm package for writing a quick Whatsapp tool.
If you want to build something using Whatsapp, I am very excited to help you.
* I recommend not using branded capitalisation for products
† This is starting to change to Signal in some communities, thank god
I’m fascinated with the AI advancements of the post-ChatGPT world. The world of software engineering is completely revolutionised because of these innovations. Open source has been transcended. I’m excited about these new tools and use them as core tools for my work albeit, indeed, the future is generally substantially unclear.
I've been exploring the Chinese open weights models (Deepseek, Kimi, Minimax, GLM) and local inference AI tools (llama.cpp, ollama, DwarfStar, LM Studio, etc). I find this ecosystem particularly interesting because it currently represents the only meaningful counterweight to the centralized AI future dominated by the three main corporate players (OpenAI, Anthropic, Google).
Different objectives of voting systems, cardinal and ordinal ranking systems, the challenge of aggregating preferences across a group into a single preference, pros and cons of different voting approaches across various contexts. Applications of voting systems for preference aggregation outside of traditional elections, e.g. business decisions on budget allocation between divisions, aggregating diverging outputs across AI agents into a single answer, problems the cohort is thinking about, etc.
Political systems can be understood through a set of mathematical and mechanistic ideas that help make sense of how they behave over time. Concepts such as homeostasis, autopoiesis, growth, equilibrium, feedback, and phase change give us ways of describing how political processes stabilise, amplify, or transform. Rather than treating politics as purely contingent or narrative-driven, these tools allow us to see recurring structural patterns: how actions generate responses, how those responses feed back into the system, and how certain configurations persist while others break down or reorganise.
Institutions can be understood in similar terms. Some behave in a homeostatic way, actively counteracting disturbances in order to maintain a stable state. Others exhibit growth dynamics, expanding their reach, resources, or influence until they encounter constraints or diminishing returns. In some cases, organisations take on autopoietic characteristics, reproducing the internal rules, roles, and cultures that sustain them, even when their original purpose has weakened or disappeared. These are not just abstract or playful analogies: they point to the fact that political systems have strong quantitative foundations, where votes, resources, and participation are all measurable. Thinking in these terms helps build a more precise understanding of why political systems sometimes appear stable, sometimes become unstable, and sometimes shift abruptly into new forms. In this topic, we will use guided reading and writing to explore how these mechanistic and systems-based ideas can be applied to better understand real political systems.
I have spent much of my career building mathematical and computational models of biological and social systems. We can learn the fundamentals of systems thinking by mapping political systems and experimenting with simple models of their behaviour. The aim is to develop a practical vocabulary for recognising feedback, stability and structural change—and a set of tools that can be applied to political questions.
Core reading
- Donella H. Meadows, Thinking in Systems: A Primer (Chelsea Green, 2008).
- Mitchel Resnick, Turtles, Termites, and Traffic Jams: Explorations in Massively Parallel Microworlds (MIT Press, 1994).
One of the more popular ideas to emerge from research on group decision-making (especially under the brand "collective intelligence") is known as the "wisdom of crowds." This is a mathematical guarantee that the average belief in a group of people is more accurate than the belief of an average individual (really).
Beliefs in this context mean estimates such as "how many people currently support our campaign. and forecasts, such as "how many people will vote for us". Beliefs might also be categorical, such as "which candidate will win this campaign."
Here are some more examples of things a group might estimate or forecast:
- How many jellybeans are in this jar?
- How long will this project take to complete?
- How many units of this product will sell next year?
- How many volunteers will we be able to recruit?
- Which candidate will win the election?
- Will oil prices exceed $200/barrel before end of 2026?
- Will Andy Burnham be prime minister on Dec 31, 2027?
Honestly, you can apply this practice without any further instruction from me: simply ask a lot of people the same question, and take an average. A fair bit of evidence suggests that the average will be pretty accurate. This is known as a "prediction poll."
You can also see this principle widely applied in prediction markets, also known as betting markets. In these markets, people place bets on future outcomes and earn money if they are correct. They are called "markets" because they operate on a similar principle to the stock market.
Prediction markets and prediction polls are methods for crowdsourcing accuracy — asking people outside your group to provide input, so you can come up with accurate beliefs.
Sometimes, however, not only care about generating a specific forecast or estimate, but instead might care about supporting the formation of accurate beliefs within your own group. Indeed, if you have a group making a decision together, it is inherently desirable that their decision be based on an accurate understanding of the world!
Unfortunately, scientific research shows that lots of things happen in groups that can lead outcomes to depart from the simple average. When we talk to each other around a table, dominant voices can undermine group accuracy, leading to the "wisdom of the few" instead of the wisdom of the crowds.
By thinking about the ways in which groups can form accurate beliefs and the ways in which groups can go wrong, you can better design your facilitation practice to support the wisdom of crowds.