Keep the tension: designing with generative AI

Slides from my keynote at the CLICKNL Design Innovation Sessions, Digital Futures theme session, Dutch Design Week, 21 October 2026, Natlab, Eindhoven.

CLICKNL invited me to give this keynote. I saw it as a great opportunity to intervene in the debate about design and generative AI. My take is informed by my past work on contestability. That work is all about approaching “AI” as an arena where various parties compete over how it is made and how it is used. My take is also informed by my commitments to practice theory and political realism. Practice theory means looking at what people and things do together, rather than at technologies in isolation. Political realism means looking at politics as it actually happens. It asks who does what to whom, for whose benefit, instead of starting from abstract moral principles.

In the talk I argue that neither wholesale rejection nor uncritical embrace of generative AI will get us very far. GenAI deskills designers and opens up new ways of designing at the same time. It is built on the past work of designers, which raises the question of whose work we build on, and who gets to call the results their own. And it automates work, but in the process work may get delegated more than replaced, transforming practices as a result. Rather than trying to resolve these tensions, I look for emerging design practices that hold them and use them as an engine for moving forward. I discuss three cases: Vera van der Burg’s Text-to-Clay, Blaze Type’s test of Monotype’s AI type tool, and The Call by Holly Herndon and Mat Dryhurst. They all go beyond consuming off-the-shelf models. They treat pipelines and workflows as an object of design. And in all three there is an awareness of the value of effort and friction. I end with an open question. If AI is not one thing, not an agent, and neither our savior nor our doom, what metaphor does your mind go to?

I hope this talk helps designers think in concrete terms about how we might “do AI” differently from the dominant ways. And I hope it helps them break free from the simplistic for-or-against debates that get us nowhere.

CLICKNL DDW 2026 DIS Digital Futures keynote slides – Kars Alfrink – Keep the tensionDownload

Works discussed

Studio Dumbar

From Text-to-Clay, Vera van der Burg

Blaze Type and Monotype’s AI type tool

The Call, Holly Herndon and Mat Dryhurst

CRTA visual identity, Bruketa&Žinić&Grey

Mud to Models and Soil to Servers workshops

  • Fieke Jansen, Sunjoo Lee, Kars Alfrink and Lena Trotereau, From Mud to Models: Regenerative Futures for Community AI (conference workshop), TH/NGS 2025, Amsterdam, 12 Dec 2025, and From Soil to Servers: Regenerative Futures for Community Computing (workshop), Creative Coding Utrecht, Utrecht, 27 Mar 2026; Kars Alfrink and Sunjoo Lee, with Advaita Jairam, From Soil to Servers: Pairing Material Practice with Speculative Design to Rethink Computing Infrastructure (conference workshop), Philosophy of Human-Technology Relations Conference (PHTR 2026), The Hague, 22 Sep 2026.
  • TU Delft, “From Mud to Models: A Workshop on Regenerative AI Futures”, 17 Dec 2025. https://www.tudelft.nl/2025/io/december/from-mud-to-models-a-workshop-on-regenerative-ai-futures
  • Kars Alfrink, Sunjoo Lee, Lena Trotereau and Fieke Jansen, “Mud and models: Does making matter for regenerative AI speculation?”, in Iskander Smit and Andrea Krajewski (eds.), The State of Responsible Technology – RESIZE<REMIX<REGEN (RIOT 2026), Stichting ThingsCon Amsterdam, June 2026, pp. 34–39. https://doi.org/10.5281/zenodo.20569440

Further reading

  • Alfrink, K. (2025). People’s compute: Design and the politics of AI infrastructures. OSF. https://doi.org/10/q8kp
  • Caplan, L. (2025). Introduction: Contesting models – art and democracy since the 1990s. In L. Caplan & K. Greaves (Eds.), Model collapse: European contemporary art in a time of democratic crisis (pp. 1–20). Manchester University Press. http://www.jstor.org/stable/jj.36233585.6
  • Geuss, R. (2008). Philosophy and real politics. Princeton University Press.
  • Kuutti, K., & Bannon, L. J. (2014). The turn to practice in HCI: Towards a research agenda. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems (pp. 3543–3552). ACM. https://doi.org/10.1145/2556288.2557111
  • Löwgren, J., & Stolterman, E. (2004). Thoughtful interaction design: A design perspective on information technology. MIT Press.
  • McGowan, T. (2016). Capitalism and desire: The psychic cost of free markets. Columbia University Press.
  • McGowan, T. (2019). Emancipation after Hegel: Achieving a contradictory revolution. Columbia University Press.
  • Morozov, E. (2024, August). AI and the techno-utopian path not taken. Le Monde diplomatique (English edition). https://mondediplo.com/2024/08/07ai-cold-war
  • Parsons, P. C. (2026). Myths and ironies of AI-assisted design. In Designing Interactive Systems Conference (DIS ’26) (pp. 523–535). ACM. https://doi.org/10.1145/3800645.3813086
  • Porfirio, D., McDermott, I., Chen, H.-M., Satake, S., Kanda, T., & LaToza, T. D. (2026). Designing for robot wranglers: A synthesis of literature and practice. In Designing Interactive Systems Conference (DIS ’26) (pp. 4284–4298). ACM. https://doi.org/10.1145/3800645.3813070
  • Retkowsky, J., Hafermalz, E., & Huysman, M. (2026). Harnessing a “spirited technology”: How working with generative AI collapses the creative process. Academy of Management Discoveries, 12(3), 363–387. https://doi.org/10.5465/amd.2024.0280
  • Retkowsky, J., Hafermalz, E., Huysman, M., & Odijk, D. (2026, September 16). Research: Gen AI is collapsing creative processes. Harvard Business Review. https://hbr.org/2026/09/research-gen-ai-is-collapsing-creative-processes
  • Salvaggio, E. (2026, August 16). The system from nowhere. Cybernetic Forests. https://mail.cyberneticforests.com/the-system-from-nowhere/
  • Sayers, S. (2011). Alienation as a critical concept. International Critical Thought, 1(3), 287–304. https://doi.org/10.1080/21598282.2011.609265
  • Shumailov, I., Shumaylov, Z., Zhao, Y., Gal, Y., Papernot, N., & Anderson, R. (2024). The curse of recursion: Training on generated data makes models forget. arXiv. https://arxiv.org/abs/2305.17493
  • Stilgoe, J., Owen, R., & Macnaghten, P. (2013). Developing a framework for responsible innovation. Research Policy, 42(9), 1568–1580. https://www.sciencedirect.com/science/article/pii/S0048733313000930
  • Suchman, L. (2023). The uncontroversial ‘thingness’ of AI. Big Data & Society, 10(2), 1–5. https://doi.org/10.1177/20539517231206794
  • Unger, G., & Narayanan, A. (2026, October 14). Steering AI beyond automation: A research agenda and plan for action. AI as Normal Technology. https://www.normaltech.ai/p/steering-ai-beyond-automation
  • van Maanen, G. (2022). AI ethics, ethics washing, and the need to politicize data ethics. Digital Society, 1(2), Article 9. https://doi.org/10.1007/s44206-022-00013-3
  • Wright, E. O. (2010). Envisioning real utopias. Verso.

Designing AI for Autonomy as a Political Project

This post was taken down on 26 October 2026. A revised version will appear at this address.

About the original

Based on a talk delivered at the AI & Algorithm Seminar hosted by the Dutch Data Protection Authority on 25 March 2026 at ECC Leiden. The original ran to about 1,449 words and 5 images, with 11 references.

Abstract

Generated by Claude, a language model, from the original text; not written by the author.

The post defines autonomy as authenticity plus agency, shows how AI threatens both, and argues that ethics- and rights-based approaches protect individuals from above but do not build collective autonomy from below: they offer guardrails, not arenas. As an example of an agonistic arena it presents the Vision Model Macroscope, a prototype built with CLEVER°FRANKE around Amsterdam’s scan-vehicle trash detection case, and findings from four focus groups with municipal workers. It closes with three moves for designing AI for autonomy: from individuals to collectives, from applications to infrastructure, and from abstract principles to actual power relations.

On AI discourse

This post was taken down on 26 October 2026. A revised version will appear at this address.

About the original

The original ran to about 296 words, with 3 references.

Abstract

Generated by Claude, a language model, from the original text; not written by the author.

A short post on the author’s distance from online AI discourse. Drawing on Lee Vinsel’s notion of hype and criti-hype and Jodi Dean’s “communicative capitalism”, it argues that online debate is not a public sphere: platforms profit from the circulation of opinions, and that circulation substitutes for political action. For someone working on AI politics the conclusion is practical rather than principled: building power happens through organising and institutions, not in threads.

Postdoc update – December 2025

Six months since my last update. The pace hasn’t slowed. Here’s what I’ve been up to and what’s on the horizon for the next six months or so. But first, a very welcome December break. Happy holidays, dear reader.

Happenings

People’s Compute at Goldsmiths: On September 18, I presented my research agenda, “People’s Compute: Design and the Politics of AI Infrastructures,” at the Politics of AI symposium at Goldsmiths, University of London. Many thanks to Dan McQuillan, Fieke Jansen, Jo Lindsay Walton, and Pat Brody for the invitation and for putting together such a thought-provoking program. Read the transcript.

Digital Autonomy Unconference: On October 3, I attended the Digital Autonomy Unconference in Amsterdam, which was organized in collaboration with Code for NL and focused on enhancing digital autonomy within Dutch public institutions. The Digital Autonomy Competence Center was also launched at this event, for which I serve as a research associate. Read the news item.

Master’s Graduation Projects: Two more of my students have graduated. Ameya Sawant completed a project about designer autonomy and GenAI (August 29, with Fernando Secomandi as chair). David Mieras completed a project about the responsible use of AI in policy preparation (October 28, with Lianne Simonse as chair).

Personal Grant: I mentioned going for a personal grant the last time around. Unfortunately, I did not advance to the final round. However, I did receive some useful feedback and will try again next year. Onwards and upwards.

Designing Responsible AI: Sara Colombo, Francesca Mauri, and I ran the second iteration of our master’s elective course, which builds on responsible research and innovation, value-sensitive design, and design fiction. See the course description here. A more detailed write-up of how the course works is forthcoming.

International Contestable AI Workshop: On November 18, I had the pleasure of hosting a delegation from Denmark and the UK for a full-day workshop about Contestable AI at TU Delft. Read the report.

Enterprise UX: On November 21, I delivered an invited talk titled “Reclaiming Autonomy: Designing AI-Enhanced Work Tools That Empower Users” at the Enterprise UX conference in Amersfoort. Thanks to Peter Boersma for the invitation. The references to Office Space and Luddism were surprisingly well-received. Read the transcript.

Your difficult design doctor, holding forth at Enterprise UX.

NIAS Workshop: I participated in a workshop at NIAS on November 26-27, exploring permacomputing, server collectives, and networks of consent. Incredibly inspiring, it has given me many new ideas for approaching my own ongoing research. Many thanks to John Boy for the invitation. View the event page.

Stop the Cuts (continued): The fight against cuts to higher education continues. On December 9 we once again went on strike and I joined the demonstration in Amsterdam with over 7,000 participants. Our far-right government may have fallen, but the cuts remain on the table. Now is the time to maintain pressure on the parties forming a government. If you work in academia and want to act, join a union (Aob or FNV) and sign up for the WOinActie newsletter.

Advisory Today, Co-Decisive Tomorrow? A paper based on a year-long participant observation of a smart city project in Amsterdam, co-authored with Mike de Kreek, Tessa Steenkamp, and Martijn de Waal (part of the Human Values for Smarter Cities project), has been accepted for the 2026 Participatory Design Conference. Very pleased about that one. A preprint will be up once we submit the final camera-ready version, I think.

ThingsCon TH/NGS: At this year’s ThingsCon conference on December 12, Fieke Jansen, Sunjoo Lee, Lena Trotereau, and I ran a workshop titled “From Mud to Models” exploring regenerative futures for community AI. Thanks to Iskander Smit for bringing us together. A report is forthcoming. Update: read the report on the TU Delft website.

Participants building clocks powered by mud batteries at our ThingsCon workshop.

Open Letter on AI Policy: I was part of the supporting team for an open letter calling on Dutch politicians to develop a national AI policy that promotes social progress. A special thanks goes out to Cristina Zaga for taking the initiative and leading the charge on this one, but also to the core team members Roel Dobbe, Iris van Rooij, Lilian de Jong, Wouter Nieuwenhuizen, Marcela Suarez, Wiep Hamstra, and Olivia Guest, and the supporting team that myself was part of Felienne Hermans, Mark D., Eelco Herder, Emile van Bergen, Siri Beerends, Nolen Gertz, Paul Peters, Gerry McGovern, Kars Alfrink, and Jelle van der Ster. Sign and share the letter here.

On deck

Looking ahead to the new year, I have several writing projects to complete: one chapter on contestability for an edited volume on the philosophy of engineering, and another chapter for an edited volume on community AI.

I will be wrapping up my duties as associate chair for the CHI 2026 design subcommittee. And I will also serve as associate chair for the DIS 2026 artifacts & systems subcommittee.

I will do the analysis and write-up of a field evaluation of the Vision Model Macroscope prototype (also part of the aforementioned Human Values for Smarter Cities project). I am also providing support on several other papers that will hopefully find their way into venues such as FAccT, DIS, and elsewhere.

Mockup of the Vision Model Macroscope prototype.

Finally, I am part of several small grant applications exploring topics that include the potential of computational argumentation techniques to enable more interactive implementations of contestable AI, as well as contestability in digital systems used for evidence management in international criminal justice.

That’s most of it, although not all of it, but this has gotten way too long already. Thanks for reading this far, if you have, and best wishes for 2026.

People’s Compute: Design and the Politics of AI Infrastructures

This post was taken down on 26 October 2026. A revised version will appear at this address.

About the original

This post is adapted from a talk given at “The Politics of AI: Governance, Resistance, Alternatives” at Goldsmiths, University of London, on 18 September 2025. The original ran to about 1,103 words and 13 images.

Abstract

Generated by Claude, a language model, from the original text; not written by the author.

The post argues that AI infrastructure (data centres, compute, platforms, APIs) constrains what designers can do at the application level, and that every smart device reinforces Big Tech’s control of that layer. It recalls Amsterdam’s De Digitale Stad as an early example of community-owned digital infrastructure and of its capture through privatisation. Drawing on the socialist republican ideal of freedom as collective self-determination, it proposes three moves for democratic AI design: from applications to infrastructures, from individuals to collectives, and from idealism to realism. It closes by positioning this agenda between technocratic management and populist anti-politics and points to the full paper, an OSF preprint.

Reclaiming Autonomy: Designing AI-Enhanced Work Tools That Empower Users

This post was taken down on 26 October 2026. A revised version will appear at this address.

About the original

Based on an invited talk delivered at Enterprise UX, on November 21, 2025 in Amersfoort, the Netherlands. The original ran to about 1,269 words and 16 images.

Abstract

Generated by Claude, a language model, from the original text; not written by the author.

The post asks what to question before adding an LLM feature to a work tool: whether it actually works, who gains or loses power, and which practices and skills it transforms. It frames these as questions of worker autonomy (authenticity and agency, after Prunkl), illustrated with scenes from Office Space, and lists ways AI threatens autonomy: manipulation, restricted choice, diminished competence. It then sketches a toolkit for designing AI for autonomy and applies three of its elements (reflexive design, stakeholder negotiation, collective autonomy support) to LegalMike, a Dutch legal AI platform, concluding that designers act as brokers of autonomy.

Designing Learning Experiences in a Post-ChatGPT World

This post was taken down on 26 October 2026. A revised version will appear at this address.

About the original

Transcript of a talk delivered at LXDCON’25 on June 12, 2025. The original ran to about 1,706 words and 32 images.

Abstract

Generated by Claude, a language model, from the original text; not written by the author.

Starting from ChatGPT’s impact on higher education, it contrasts Ethan Mollick’s adapt-and-redesign optimism with Ian Bogost’s realist diagnosis of a crisis that technology cannot fix, using Harvard’s StatGPT Slackbot as a running example. It proposes “Conscious Adaptation”, built on three strategies: institutional realism, loss-conscious innovation, and keeping solution-building and critique as distinct but mutually informing work. It ends with three imperatives for learning experience designers: protect meaningful human effort, design for real institutions rather than ideal ones, and recognise the limits of design.

On the design and regulation of technology

The following is a section from a manuscript in press on the similarities and differences in approaches to explainable and contestable AI in design and law (Schmude et al., 2025). It ended up on the cutting room floor, but it is the kind of thing I find handy to refer back to, so I chose to share it here.

The responsible design of AI, including practices that seek to make AI systems more explainable and contestable, must somehow relate to legislation and regulations. Writing about responsible research and innovation (RRI) more broadly, Stilgoe et al. (2013) assert that RRI, which we would say includes design, must be embedded in regulation. But does it really make sense to think of the relationship between design and regulation in this way? Understood abstractly, there are in fact at least four ways in which we can think about the relationship between the design and regulation of technology (Figure 1).

Figure 1: We see four possible ways that the relationship between (D) the design and (R) regulation of technology can be conceptualized: (1) design and regulation are independent spheres, (2) design and regulation partially overlap, (3) design is embedded inside of regulation, or (4) regulation is embedded inside design. In all cases, we assume an interactional relation between the two spheres.

To establish the relationship between design and regulation, we first need to establish how we should think about regulation, and related concepts such as governance and policymaking more generally. One straightforward definition would be that regulation entails formal rules and enforcement mechanisms that constrain behavior. These are backed by authority—typically state authority, but increasingly also non-stake actors. Regulation and governance are interactive and mutually constitutive. Regulation is one mechanism within governance systems. Governance frameworks establish contexts for regulation. Policymaking connects politics to governance by translating political contestation into actionable frameworks. Politics, then, influences all these domains: policymaking, governance, and regulation. And they, in turn, operate within and reciprocally shape society writ large. See Table 1 for working definitions of ‘regulation’ and associated concepts.

ConceptDefinition
RegulationFormal rules and enforcement mechanisms that constrain behavior, typically state-backed but increasingly emerging from non-state actors (industry self-regulation, transnational regulatory bodies).
GovernanceBroader arrangements for coordinating social action across public, private, and civil society spheres through both formal and informal mechanisms.
PolicymakingProcess of formulating courses of action to address public problems.
PoliticsContestation of power, interests, and values that shapes governance arrangements.
SocietyBroader context of social relations, norms, and institutions.
Table 1: Working definitions of ‘regulation’ and associated concepts.

What about design? Scholars of regulation have adopted the notion of ‘regulation by design’ (RBD) to refer to the inscribing of rules into the world through the creation and implementation of technological artifacts. Prifti (2024) identifies two prevailing approaches to RBD: The essentialist view treats RBD as policy enactments, or “rules for design.” By contrast, the functionalist view treats design as a mere instrument, or “ruling by design.” We agree with Prifti when he states that both approaches are limited. Essentialism neglects the complexity of regulatory environments, while functionalism neglects the autonomy and complexity of design as a practice.

Prifti proposes a pragmatist reconstruction that views regulation as a rule-making activity (“regulativity”) performed through social practices including design (the “rule of design”). Design is conceptualized as a contextual, situated social practice that performs changes in the environment, rather than just a tool or set of rules. Unlike the law, markets, or social norms, which rely on incentives and sanctions, design can simply disable the possibility of non-compliance, making it a uniquely powerful form of regulation. The pragmatist approach distinguishes between regulation and governance, with governance being a meta-regulative activity that steers how other practices (like design) regulate. This reconceptualization helps address legitimacy concerns by allowing for greater accountability for design practices that might bypass democratic processes.

Returning to the opening question then, out of the four basic ways in which the relationship between design and regulation can be drawn (Figure 1), if we were to adopt Prifti’s pragmatist view, Type 3 would most accurately capture the relationship, with design being one of a variety of more specific ways in which regulation (understood as regulativity) actually makes changes in the world. These other forms of regulatory practice are not depicted in the figure. This seems to align with Stilgoe et al.’s aforementioned position that responsible design must be embedded within regulation. Although there is a slight nuance to our position: Design is conceived of as a form of regulation always, regardless of active work on the part of designers to ‘embed’ their work inside regulatory practices. Stilgoe et al.’s admonition can be better understood as a normative claim: Responsible designers would do well to understand and align their design work with extant laws and regulations. Furthermore, following Prifti, design is beholden to governance and must be reflexively aware of how governance steers its practices (cf. Figure 2).

Figure 2: Conceptual model of the relationship between design, ‘classical’ regulation (i.e., law-making), and governance. Both design and law-making are forms of regulation (i.e., ‘regulativity’). Governance steers how design and law-making regulate, and design and law-making are both accountable to (and reflexively aware of) governance.

Bibliography

  • Prifti, K. (2024). The theory of ‘Regulation By Design’: Towards a pragmatist reconstruction. Technology and Regulation, 2024, 152–166. https://doi.org/10/g9dr24
  • Stilgoe, J., Owen, R., & Macnaghten, P. (2013). Developing a framework for responsible innovation. Research Policy, 42(9), 1568–1580. https://doi.org/10/f5gv8h

Postdoc update – July 2025

I am over one year into my postdoc at TU Delft. Where did the time go? By way of an annual report, here’s a rundown of my most notable outputs and activities since the previous update from June 2024. And also, some notes on what I am up to now.

Happenings

Participatory AI and ML Engineering: On 13 February 2024 at a Human Values for Smarter Cities meeting and on 11 June 2024 at a Cities Coalition for Digital Rights meeting, I presented a talk on participatory AI and ML engineering (blogged here). This has since evolved into a study I am currently running with the working title “Vision Model Macroscope.” We are designing, building, and evaluating an interface that allows municipal workers to understand and debate value-laden technical decisions made by machine learning engineers in the construction of camera vehicles. For the design, I am collaborating with CLEVER°FRANKE. The study is part of the Human Values for Smarter Cities projected headed up by the Civic Interaction Design group at AUAS.

Envisioning Contestability Loops: My article “Envisioning Contestability Loops: Evaluating the Agonistic Arena as a Generative Metaphor for Public AI” (with Ianus Keller, Mireia Yurrita Semperena, Denis Bulygin, Gerd Kortuem, and Neelke Doorn) was published in She Ji on 17 June 2024. (I had already published the infographic “Contestability Loops for Public AI,” which the article revolves around, on 17 April 2024.) Later in the year, on 5 September 2024, I ran the workshop that the study builds on as a ThingsCon Salon. And on 27 September 2024, I presented the article at Lawtomation Days in Madrid, Spain, as part of the panel “Methods in law and technology research: inter- and cross-disciplinary challenges and opportunities,” chaired by Kostina Prifti (slides). (Also, John Thackara said nice things about the article online.)

Contestability Loops for Public AI infographic
Envisioning Contestability Loops workshop at ThingsCon Salon in progress.

Democratizing AI Through Continuous Adaptability: I presented on “Democratizing AI Through Continuous Adaptability: The Role of DevOps” at the TILTing Perspectives 2024 panel “The mutual shaping of democratic practices & AI,” which was chaired and moderated by Merel Noorman on 14 July 2024. I later reprised this talk at NWO ICT.OPEN on 16 April 2025 as part of the track “Human-Computer Interaction and Societal Impact in the Netherlands,” chaired by Armağan Karahanoğlu and Max Birk (PDF of slides).

From Stem to Stern: I was part of the organizing team of the CSCW 2024 workshop “From Stem to Stern: Contestability Along AI Value Chains,” which took place as a hybrid one-day session on 9 November 2024. I blogged a summary and some takeaways of the workshop here. Shoutout to Agathe Balayn and Yulu Pi for leading this endeavor.

Contestable AI Talks: I was invited to speak on my PhD research at various meetings and events organized by studios, agencies, consultancies, schools, and public sector organizations. On 3 September 2024, at the data design agency CLEVER°FRANKE (slides). On 10 January 2025, at the University of Utrecht Computational Sociology group. On 19 February 2025, at digital ethics consultancy The Green Land (slides). On 6 March 2024, at Communication and Multimedia Design Amsterdam (slides). And on 17 March 2025, at the Advisory Board on Open Government and Information Management.

Designing Responsible AI: Over the course of 2024, Sara Colombo, Francesca Mauri, and I developed and taught for the first time a new Integrated Product Design master’s elective, “Designing Responsible AI” (course description). Later, on 28 March 2025, I was invited by my colleagues Alessandro Bozzon and Carlo van der Valk to give a single-morning interactive lecture on part of the same content at the course AI Products and Services (slides).

Books that represent the range of theory covered in the course “Designing Responsible AI.”

Stop the Cuts: On 2 July 2024, a far-right government was sworn in in the Netherlands (it has since fallen). They intended to cut funding to education by €2 billion. A coalition of researchers, teachers, students, and others organized to protest and strike in response. I was present at several of these actions: The alternative opening of the academic year in Utrecht on 2 September 2024. Local walkouts on 14 November 2024 (I participated in Utrecht). Mass demonstration in The Hague on 25 November 2024. Local actions on 11 December 2024 (I participated in Delft). And finally, for now at least, on 24 April 2025, at the Delft edition of the nationwide relay strike. If you read this, work in academia, and want to act, join a union (I am a member of the AOb), and sign up for the WOinActie newsletter.

End of the march during the 24 April 2025 strike in Delft.

Panels: Over the past months, I was a panelist at several events. On 22 October 2024, at the Design & AI Symposium as part of the panel “Evolving Perspectives on AI and Design,” together with Iohanna Nicenboim and Jesse Benjamin, moderated by Mathias Funk (blog post). On 13 December 2024 at TH/NGS as part of the panel “Rethink Design: Book Launch and Panel Discussion on Designing With AI” chaired by Roy Bendor (video). On 12 March 2025, at the panel “Inclusive AI: Approaches to Digital Inclusion,” chaired by Nazli Cila and Taylor Stone.

Slide I used during my panel contribution at the Design & AI symposium.

Design for Human Autonomy: I was part of several activities organized by the Delft Design for Values institute related to their annual theme of autonomy (led by Michael Klenk). I was a panelist on 15 October 2024 during the kick-off event (blog post). I wrote the section on designing AI for autonomy for the white paper edited by Udo Pesch (preprint). And during the closing symposium, master’s graduation student Ameya Sawant, whom I am coaching (with Fernando Secomandi acting as chair), was honored as a finalist in the thesis competition.

Master Graduation Students: Four master students that I coached during their thesis projects graduated, which between them explored technology’s role in society through AI-mediated civic engagement, generative AI implementation in public services, experimental approaches to AI trustworthiness, and urban environmental sensing—Nina te Groen (with Achilleas Psyilidis as chair), Romée Postma (with Roy Bendor), Eline Oei (with Giulia Calabretta), and Jim Blom (with Tomasz Jaskiewicz).

Architecting for Contestability: On 22 November 2025, I ran a single-day workshop about contestability for government-employed ICT architects participating in the Digital Design & Architecture course offered by the University of Twente, on invitation from Marijn Janssen (slides).

Qualitative Design Research: On 17 December 2024, I delivered a lecture on qualitative design research for the course Empirical Design Research, on invitation from my colleague Himanshu Verma (slides). Later, on 22 April 2025, I delivered a follow-up in the form of a lecture on reflexive thematic analysis for the course Product Futures Studio, coordinated by Holly McQuillan (slides).

Democratic Generative Things: On 6 June 2025 I joined the ThingsCon unconference to discuss my contribution to the RIOT report, “Embodied AI and collective power: Designing democratic generative things” (preprint). The report was edited by edited by Andrea Krajewski and Iskander Smit.

Me, holding forth during the ThingsCon RIOT unconference.

Learning Experience Design: I delivered the closing invited talk at LXDCON on 12 June 2025, reflecting on the impact of GenAI on the fields of education and design for learning (slides). Many thanks to Niels Floor for the invitation.

People’s Compute: I published a preprint of my position paper “People’s Compute: Design and the Politics of AI Infrastructures” over at OSF on 14 April 2025. I emailed it to peers and received over a dozen encouraging responses. It was also somehow picked up by Evgeny Morozov’s The Syllabus with some nice commentary attached.

On deck

So what am I up to at the moment? Keeping nice and busy.

  • I am co-authoring several articles, papers, and book chapters on topics including workplace automation, AI transparency, contestability in engineering, AI design and regulation, computational argumentation, explainable and participatory AI, and AI infrastructure politics. I do hope at least some of these will see the light of day in the coming months.
  • I am preparing a personal grant application that builds on the vision laid out in People’s Compute.
  • I will be delivering an invited talk at Enterprise UX on 21 November 2025.
  • I am acting as a scientific advisor to a center that is currently being established, which focuses on increasing digital autonomy within Dutch government institutions.
  • I will be co-teaching Designing Responsible AI again in Q1 of the next academic year.
  • I’ll serve as an associate chair on the CHI 2026 design subcommittee.
  • And I have signed up to begin our university’s teaching qualification certification.

Whew. That’s it. Thanks for reading (skimming?) if you’ve made it all the way to the end. I will try to circle back and do another update, maybe a little sooner than this one, say in six months’ time.

On how to think about large language models

How should we think about large language models (LLMs)? People commonly think and talk about them in terms of human intelligence. To the extent this metaphor does not accurately reflect the properties of the technology, this may lead to misguided diagnoses and prescriptions. It seems to me an LLM is not like a human or a human brain in so many ways. One crucial distinction for me is that LLMs lack individuality and subjectivity.

What are organisms that similarly lack these qualities? Coral polyps and Portuguese man o’ war come to mind, or slime mold colonies. Or maybe a single bacterium, like an E. coli. Each is essentially identical to its clones, responds automatically to chemical gradients (bringing to mind how LLMs respond to prompts), and doesn’t accumulate unique experiences in any meaningful way.

Considering all these examples, the meme about LLMs being like a shoggoth (an amorphous blob-like monster originating from the speculative fiction of Howard Philips Lovecraft) is surprisingly accurate. The thing about these metaphors though is that it’s about as hard to reason about such organisms as it is to reason about LLMs. So to use them as a metaphor for thinking about LLMs won’t work. A shoggoth is even less helpful because the reference will only be familiar to those who know their H.P. Lovecraft.

So perhaps we should abandon metaphorical thinking and think historically instead. LLMs are a new language technology. As with previous technologies, such as the printing press, when they are introduced, our relationship to language changes. How does this change occur?

I think the change is dialectical. First, we have a relationship to language that we recognize as our own. Then, a new technology destabilizes this relationship, alienating us from the language practice. We no longer see our own hand in it. And we experience a lack of control over language practice. Finally, we reappropriate this language use in our practices. In this process of reappropriation, language practice as a whole is transformed. And the cycle begins again.

For an example of this dialectical transformation of language practice under the influence of new technology, we can take Eisenstein’s classic account of the history of the printing press (1980). Following its introduction many things changed about how we relate to language. Our engagement with language shifted from a primarily oral one to a visual and deliberative one. Libraries became more abundantly stocked, leading to the practice of categorization and classification of works. Preservation and analysis of stable texts became a possibility. The solitary reading experience gained prominence, producing a more private and personal relationship between readers and texts. Concerns about information overload first reared its head.

All of these things were once new and alien to humans. Now we consider them part of the natural order of things. They weren’t predetermined by the technology, they emerged through this active tug of war between groups in society about what the technology would be used for, mediated by the affordances of the technology itself.

In concrete material terms, what does an LLM consist of? An LLM is just numerical values stored in computer memory. It is a neural network architecture consisting of billions of parameters in weights and biases, organized in matrices. The storage is distributed across multiple devices. System software loads these parameters and enables the calculation of inferences. This all runs in physical data centers housing computing infrastructure, power, cooling, and networking infrastructure. Whenever people start talking about LLMs having agency or being able to reason, I remind myself of these basic facts.

A printing press, although a cleverly designed, engineered, and manufactured device, is similarly banal when you break it down to its essential components. Still, the ultimate changes to how we relate to language have been profound. From these first few years of living with LLMs, I think it is not unreasonable to think they will cause similar upheavals. What is important for me is to recognize how we become alienated from language, and to see ourselves as having agency in reappropriating LLM-mediated language practice as our own.