Democratizing AI Through Continuous Adaptability: The Role of DevOps

Below are the abstract and slides for my con­tri­bu­tion to the TILT­ing Per­spec­tives 2024 pan­el “The mutu­al shap­ing of demo­c­ra­t­ic prac­tices & AI,” mod­er­at­ed by Mer­el Noorman.




This pre­sen­ta­tion delves into democ­ra­tiz­ing arti­fi­cial intel­li­gence (AI) sys­tems through con­testa­bil­i­ty. Con­testa­bil­i­ty refers to the abil­i­ty of AI sys­tems to remain open and respon­sive to dis­putes through­out their life­cy­cle. It approach­es AI sys­tems as are­nas where groups com­pete for pow­er over designs and outcomes. 

Autonomy, democratic agency, legitimation

We iden­ti­fy con­testa­bil­i­ty as a crit­i­cal sys­tem qual­i­ty for respect­ing peo­ple’s auton­o­my. This includes their demo­c­ra­t­ic agency: their abil­i­ty to legit­i­mate poli­cies. This includes poli­cies enact­ed by AI systems. 

For a deci­sion to be legit­i­mate, it must be demo­c­ra­t­i­cal­ly willed or rely on “nor­ma­tive author­i­ty.” The demo­c­ra­t­ic path­way should be con­strained by nor­ma­tive bounds to avoid arbi­trari­ness. The appeal to author­i­ty should meet the “access con­straint,” which ensures cit­i­zens can form beliefs about poli­cies with a suf­fi­cient degree of agency (Peter, 2020 in Rubel et al., 2021).

Con­testa­bil­i­ty is the qual­i­ty that ensures mech­a­nisms are in place for sub­jects to exer­cise their demo­c­ra­t­ic agency. In the case of an appeal to nor­ma­tive author­i­ty, con­testa­bil­i­ty mech­a­nisms are how sub­jects and their rep­re­sen­ta­tives gain access to the infor­ma­tion that will enable them to eval­u­ate its jus­ti­fi­a­bil­i­ty. In this way, con­testa­bil­i­ty sat­is­fies the access con­straint. In the case of demo­c­ra­t­ic will, con­testa­bil­i­ty-by-design prac­tices are how sys­tem devel­op­ment is democ­ra­tized. The auton­o­my account of legit­i­ma­tion adds the nor­ma­tive con­straints that should bind this demo­c­ra­t­ic pathway.

Him­mel­re­ich (2022) sim­i­lar­ly argues that only a “thick” con­cep­tion of democ­ra­cy will address some of the cur­rent short­com­ings of AI devel­op­ment. This is a path­way that not only allows for par­tic­i­pa­tion but also includes delib­er­a­tion over justifications.

The agonistic arena

Else­where, we have pro­posed the Ago­nis­tic Are­na as a metaphor for think­ing about the democ­ra­ti­za­tion of AI sys­tems (Alfrink et al., 2024). Con­testable AI embod­ies the gen­er­a­tive metaphor of the Are­na. This metaphor char­ac­ter­izes pub­lic AI as a space where inter­locu­tors embrace con­flict as pro­duc­tive. Seen through the lens of the Are­na, pub­lic AI prob­lems stem from a need for oppor­tu­ni­ties for adver­sar­i­al inter­ac­tion between stakeholders. 

This metaphor­i­cal fram­ing sug­gests pre­scrip­tions to make more con­tentious and open to dis­pute the norms and pro­ce­dures that shape:

  1. AI sys­tem design deci­sions on a glob­al lev­el, and
  2. human-AI sys­tem out­put deci­sions on a local lev­el (i.e., indi­vid­ual deci­sion out­comes), estab­lish­ing new dia­log­i­cal feed­back loops between stake­hold­ers that ensure con­tin­u­ous monitoring.

The Are­na metaphor encour­ages a design ethos of revis­abil­i­ty and reversibil­i­ty so that AI sys­tems embody the ago­nis­tic ide­al of contingency.

Post-deployment malleability, feedback-ladenness

Unlike phys­i­cal sys­tems, AI tech­nolo­gies exhib­it a unique mal­leabil­i­ty post-deployment. 

For exam­ple, LLM chat­bots opti­mize their per­for­mance based on a vari­ety of feed­back sources, includ­ing inter­ac­tions with users, as well as feed­back col­lect­ed through crowd-sourced data work.

Because of this open-end­ed­ness, demo­c­ra­t­ic con­trol and over­sight in the oper­a­tions phase of the sys­tem’s life­cy­cle become a par­tic­u­lar concern.

This is a con­cern because while AI sys­tems are dynam­ic and feed­back-laden (Gilbert et al., 2023), many of the exist­ing over­sight and con­trol mea­sures are sta­t­ic, one-off exer­cis­es that strug­gle to track sys­tems as they evolve over time.


The field of DevOps is piv­otal in this con­text. DevOps focus­es on sys­tem instru­men­ta­tion for enhanced mon­i­tor­ing and con­trol for con­tin­u­ous improve­ment. Typ­i­cal­ly, met­rics for DevOps and their machine learn­ing-spe­cif­ic MLOps off­shoot empha­size tech­ni­cal per­for­mance and busi­ness objectives.

How­ev­er, there is scope to expand these to include mat­ters of pub­lic con­cern. The mat­ters-of-con­cern per­spec­tive shifts the focus on issues such as fair­ness or dis­crim­i­na­tion, view­ing them as chal­lenges that can­not be resolved through uni­ver­sal meth­ods with absolute cer­tain­ty. Rather, it high­lights how stan­dards are local­ly nego­ti­at­ed with­in spe­cif­ic insti­tu­tion­al con­texts, empha­siz­ing that such stan­dards are nev­er guar­an­teed (Lam­p­land & Star, 2009, Geiger et al., 2023).

MLOps Metrics

In the con­text of machine learn­ing sys­tems, tech­ni­cal met­rics focus on mod­el accu­ra­cy. For exam­ple, a finan­cial ser­vices com­pa­ny might use Area Under The Curve Receiv­er Oper­at­ing Char­ac­ter­is­tics (AUC-ROC) to con­tin­u­ous­ly mon­i­tor and main­tain the per­for­mance of their fraud detec­tion mod­el in production.

Busi­ness met­rics focus on cost-ben­e­fit analy­ses. For exam­ple, a bank might use a cost-ben­e­fit matrix to bal­ance the poten­tial rev­enue from approv­ing a loan against the risk of default, ensur­ing that the over­all prof­itabil­i­ty of their loan port­fo­lio is optimized.


These met­rics can be mon­i­tored over time to detect “drift” between a mod­el and the world. Train­ing sets are sta­t­ic. Real­i­ty is dynam­ic. It changes over time. Drift occurs when the nature of new input data diverges from the data a mod­el was trained on. A change in per­for­mance met­rics may be used to alert sys­tem oper­a­tors, who can then inves­ti­gate and decide on a course of action, e.g., retrain­ing a mod­el on updat­ed data. This, in effect, cre­ates a feed­back loop between the sys­tem in use and its ongo­ing development.

An expan­sion of these prac­tices in the inter­est of con­testa­bil­i­ty would require:

  1. set­ting dif­fer­ent metrics,
  2. expos­ing these met­rics to addi­tion­al audi­ences, and
  3. estab­lish­ing feed­back loops with the process­es that gov­ern mod­els and the sys­tems they are embed­ded in.

Example 1: Camera Cars

Let’s say a city gov­ern­ment uses a cam­era-equipped vehi­cle and a com­put­er vision mod­el to detect pot­holes in pub­lic roads. In addi­tion to accu­ra­cy and a favor­able cost-ben­e­fit ratio, cit­i­zens, and road users in par­tic­u­lar, may care about the time between a detect­ed pot­hole and its fix­ing. Or, they may care about the dis­tri­b­u­tion of pot­holes across the city. Fur­ther­more, when road main­te­nance appears to be degrad­ing, this should be tak­en up with depart­ment lead­er­ship, the respon­si­ble alder­per­son, and coun­cil members.

Example 2: EV Charching

Or, let’s say the same city gov­ern­ment uses an algo­rith­mic sys­tem to opti­mize pub­lic elec­tric vehi­cle (EV) charg­ing sta­tions for green ener­gy use by adapt­ing charg­ing speeds to expect­ed sun and wind. EV dri­vers may want to know how much ener­gy has been shift­ed to green­er time win­dows and its trends. With­out such vis­i­bil­i­ty on a sys­tem’s actu­al goal achieve­ment, cit­i­zens’ abil­i­ty to legit­i­mate its use suf­fers. As I have already men­tioned, demo­c­ra­t­ic agency, when enact­ed via the appeal to author­i­ty, depends on access to “nor­ma­tive facts” that under­pin poli­cies. And final­ly, pro­fessed sys­tem func­tion­al­i­ty must be demon­strat­ed as well (Raji et al., 2022).

DevOps as sociotechnical leverage point for democratizing AI

These brief exam­ples show that the DevOps approach is a poten­tial sociotech­ni­cal lever­age point. It offers path­ways for democ­ra­tiz­ing AI sys­tem design, devel­op­ment, and operations. 

DevOps can be adapt­ed to fur­ther con­testa­bil­i­ty. It cre­ates new chan­nels between human and machine actors. One of DevOp­s’s essen­tial activ­i­ties is mon­i­tor­ing (Smith, 2020), which pre­sup­pos­es fal­li­bil­i­ty, a nec­es­sary pre­con­di­tion for con­testa­bil­i­ty. Final­ly, it requires and pro­vides infra­struc­ture for tech­ni­cal flex­i­bil­i­ty so that recov­ery from error is low-cost and con­tin­u­ous improve­ment becomes prac­ti­cal­ly feasible.

The mutual shaping of democratic practices & AI

Zoom­ing out fur­ther, let’s reflect on this pan­el’s over­all theme, pick­ing out three ele­ments: legit­i­ma­tion, rep­re­sen­ta­tion of mar­gin­al­ized groups, and deal­ing with con­flict and con­tes­ta­tion after imple­men­ta­tion and dur­ing use.

Con­testa­bil­i­ty is a lever for demand­ing jus­ti­fi­ca­tions from oper­a­tors, which is a nec­es­sary input for legit­i­ma­tion by sub­jects (Henin & Le Métay­er, 2022). Con­testa­bil­i­ty frames dif­fer­ent actors’ stances as adver­sar­i­al posi­tions on a polit­i­cal field rather than “equal­ly valid” per­spec­tives (Scott, 2023). And final­ly, rela­tions, mon­i­tor­ing, and revis­abil­i­ty are all ways to give voice to and enable respon­sive­ness to con­tes­ta­tions (Genus & Stir­ling, 2018).

And again, all of these things can be fur­thered in the post-deploy­ment phase by adapt­ing the DevOps lens.


  • Alfrink, K., Keller, I., Kortuem, G., & Doorn, N. (2022). Con­testable AI by Design: Towards a Frame­work. Minds and Machines33(4), 613–639.
  • Alfrink, K., Keller, I., Yur­ri­ta Sem­per­e­na, M., Buly­gin, D., Kortuem, G., & Doorn, N. (2024). Envi­sion­ing Con­testa­bil­i­ty Loops: Eval­u­at­ing the Ago­nis­tic Are­na as a Gen­er­a­tive Metaphor for Pub­lic AIShe Ji: The Jour­nal of Design, Eco­nom­ics, and Inno­va­tion10(1), 53–93.
  • Geiger, R. S., Tan­don, U., Gakhokidze, A., Song, L., & Irani, L. (2023). Mak­ing Algo­rithms Pub­lic: Reimag­in­ing Audit­ing From Mat­ters of Fact to Mat­ters of Con­cern. Inter­na­tion­al Jour­nal of Com­mu­ni­ca­tion18(0), Arti­cle 0.
  • Genus, A., & Stir­ling, A. (2018). Collingridge and the dilem­ma of con­trol: Towards respon­si­ble and account­able inno­va­tion. Research Pol­i­cy47(1), 61–69.
  • Gilbert, T. K., Lam­bert, N., Dean, S., Zick, T., Snoswell, A., & Mehta, S. (2023). Reward Reports for Rein­force­ment Learn­ing. Pro­ceed­ings of the 2023 AAAI/ACM Con­fer­ence on AI, Ethics, and Soci­ety, 84–130.
  • Henin, C., & Le Métay­er, D. (2022). Beyond explain­abil­i­ty: Jus­ti­fi­a­bil­i­ty and con­testa­bil­i­ty of algo­rith­mic deci­sion sys­tems. AI & SOCIETY37(4), 1397–1410.
  • Him­mel­re­ich, J. (2022). Against “Democ­ra­tiz­ing AI.” AI & SOCIETY
  • Lam­p­land, M., & Star, S. L. (Eds.). (2008). Stan­dards and Their Sto­ries: How Quan­ti­fy­ing, Clas­si­fy­ing, and For­mal­iz­ing Prac­tices Shape Every­day Life (1st edi­tion). Cor­nell Uni­ver­si­ty Press.
  • Peter, F. (2020). The Grounds of Polit­i­cal Legit­i­ma­cy. Jour­nal of the Amer­i­can Philo­soph­i­cal Asso­ci­a­tion6(3), 372–390.
  • Raji, I. D., Kumar, I. E., Horowitz, A., & Selb­st, A. (2022). The Fal­la­cy of AI Func­tion­al­i­ty. 2022 ACM Con­fer­ence on Fair­ness, Account­abil­i­ty, and Trans­paren­cy, 959–972.
  • Rubel, A., Cas­tro, C., & Pham, A. K. (2021). Algo­rithms and auton­o­my: The ethics of auto­mat­ed deci­sion sys­tems. Cam­bridge Uni­ver­si­ty Press.
  • Scott, D. (2023). Diver­si­fy­ing the Delib­er­a­tive Turn: Toward an Ago­nis­tic RRISci­ence, Tech­nol­o­gy, & Human Val­ues48(2), 295–318.
  • Smith, J. D. (2020). Oper­a­tions anti-pat­terns, DevOps solu­tions. Man­ning Publications.
  • Treveil, M. (2020). Intro­duc­ing MLOps: How to scale machine learn­ing in the enter­prise (First edi­tion). O’Reilly.

PhD update – June 2024

I am writ­ing this final PhD update as a fresh­ly mint­ed doc­tor. On Thurs­day, May 23, 2024, I suc­cess­ful­ly defend­ed my the­sis, ‘Con­testable Arti­fi­cial Intel­li­gence: Con­struc­tive Design Research for Pub­lic Arti­fi­cial Intel­li­gence Sys­tems that are Open and Respon­sive to Dispute.’

I start­ed the PhD on Sep­tem­ber 1, 2018 (read the very first update post­ed on that day here). So, that’s five years, eight months, 23 days from start to fin­ish. It has been quite the jour­ney, and I feel hap­py and relieved to have com­plet­ed it. I am proud of the work embod­ied in the the­sis. Most of all, I am thank­ful for the trans­for­ma­tive learn­ing expe­ri­ence, none of which would have been pos­si­ble with­out the sup­port of my super­vi­sors Gerd, Ianus, and Neelke.

On the day itself, I was hon­ored to have as my exter­nal com­mit­tee mem­bers pro­fes­sors Dignum, Löw­gren, van Zoo­nen, and van de Poel, pro­fes­sor Voûte as the chair, and Joost and Mireia as my paranymphs.

The the­sis PDF can be down­loaded at the TU Delft repos­i­to­ry, and a video of the pro­ceed­ings is avail­able on YouTube.

Me, with a copy of the the­sis, short­ly before start­ing the layper­son­’s talk. Pho­to: Roy Borgh­outs.

Recent events

Review­ing my notes since the last update, below are some more notable things that hap­pened in the past eight months.

  • I ran a short work­shop on AI Ped­a­gogy Through A Design Lens, togeth­er with Hosana Morales, at the TU Delft spring sym­po­sium on AI edu­ca­tion. Read the post.
  • A sto­ry about my research was pub­lished on the TU Delft indus­tri­al design engi­neer­ing web­site in the run-up to my defense on May 14, 2024. Read the sto­ry.
  • I updat­ed and ran the fifth and final iter­a­tion of the AI & Soci­ety indus­tri­al design engi­neer­ing mas­ter elec­tive course from Feb­ru­ary 28 through April 10, 2024. A pre­vi­ous ver­sion is doc­u­ment­ed here, which I plan to update some­time in the near future.
  • I gave a talk titled Con­testable AI: Design­ing for Human Auton­o­my at the Ams­ter­dam UX meet­up on Feb­ru­ary 21, 2024. Down­load the slides.
  • The out­comes of a design sprint on tools for third-par­ty scruti­ny, orga­nized by the Respon­si­ble Sens­ing Lab, which took inspi­ra­tion from my research, were pub­lished on Decem­ber 7, 2023. Read the report.
  • I was inter­viewed by Mireia Yur­ri­ta Sem­per­e­na for a DCODE pod­cast episode titled Beyond Val­ues in Algo­rith­mic Design, pub­lished Novem­ber 6, 2023. Lis­ten to the episode.
  • Togeth­er with Clau­dio Sar­ra and Mar­co Alma­da, I host­ed an online sem­i­nar titled Build­ing Con­testable Sys­tems on Octo­ber 26, 2023. Read the thread.
  • I was a pan­elist at the Design & AI Sym­po­sium 2023 on Octo­ber 18, 2023.
  • A paper I co-authored titled When ‘Doing Ethics’ Meets Pub­lic Pro­cure­ment of Smart City Tech­nol­o­gy – an Ams­ter­dam Case Study, was pre­sent­ed by first author Mike de Kreek at IASDR 2023 on Octo­ber 9–13. Read the paper.

Looking ahead

I will con­tin­ue at TU Delft as a post­doc­tor­al researcher and will stay focused on design, AI, and pol­i­tics, but I will try to evolve my research into some­thing that builds on my the­sis work but adds a new angle.

The Envi­sion­ing Con­testa­bil­i­ty Loops arti­cle men­tioned in pre­vi­ous updates is now in press with She Ji, which I am very pleased about. It should be pub­lished “soon.”

Upcom­ing appear­ances include a brief talk on par­tic­i­pa­to­ry AI at a Cities Coali­tion for Dig­i­tal Rights event and a pre­sen­ta­tion as part of a pan­el on The Mutu­al Shap­ing Of Demo­c­ra­t­ic Prac­tices And AI at TILT­ing Per­spec­tives 2024.

That’s it for this final PhD update. I will prob­a­bly con­tin­ue these posts under a new title. We’ll see.

AI pedagogy through a design lens

At a TU Delft spring sym­po­sium on AI edu­ca­tion, Hosana and I ran a short work­shop titled “AI ped­a­gogy through a design lens.” In it, we iden­ti­fied some of the chal­lenges fac­ing AI teach­ing, par­tic­u­lar­ly out­side of com­put­er sci­ence, and explored how design ped­a­gogy, par­tic­u­lar­ly the prac­tices of stu­dios and mak­ing, may help to address them. The AI & Soci­ety mas­ter elec­tive I’ve been devel­op­ing and teach­ing over the past five years served as a case study. The ses­sion was punc­tu­at­ed by brief brain­storm­ing using an adapt­ed ver­sion of the SQUID gamestorm­ing tech­nique. Below are the slides we used.

Participatory AI literature review

I’ve been think­ing alot about civic par­tic­i­pa­tion in machine learn­ing sys­tems devel­op­ment. In par­tic­u­lar, involv­ing non-experts in the poten­tial­ly val­ue-laden trans­la­tion work from spec­i­fi­ca­tions that engi­neers do when they build their mod­els. Below is a sum­ma­ry of a selec­tion of lit­er­a­ture I found on the top­ic, which may serve as a jump­ing-off point for future research.


The lit­er­a­ture on par­tic­i­pa­to­ry arti­fi­cial intel­li­gence (AI) reveals a com­plex land­scape marked by chal­lenges and evolv­ing method­olo­gies. Fef­fer et al. (2023) cri­tique the reduc­tion of par­tic­i­pa­tion to com­pu­ta­tion­al mech­a­nisms that only approx­i­mate nar­row moral val­ues. They also note that engage­ments with stake­hold­ers are often super­fi­cial and unrep­re­sen­ta­tive. Groves et al. (2023) iden­ti­fy sig­nif­i­cant bar­ri­ers in com­mer­cial AI labs, includ­ing high costs, frag­ment­ed approach­es, exploita­tion con­cerns, lack of trans­paren­cy, and con­tex­tu­al com­plex­i­ties. These bar­ri­ers lead to a piece­meal approach to par­tic­i­pa­tion with min­i­mal impact on deci­sion-mak­ing in AI labs. Del­ga­do et al. (2023) observe that par­tic­i­pa­to­ry AI involves stake­hold­ers most­ly in a con­sul­ta­tive role with­out inte­grat­ing them as active deci­sion-mak­ers through­out the AI design lifecycle.

Gerdes (2022) pro­pos­es a data-cen­tric approach to AI ethics and under­scores the need for inter­dis­ci­pli­nary bridge builders to rec­on­cile dif­fer­ent stake­hold­er per­spec­tives. Robert­son et al. (2023) explore par­tic­i­pa­to­ry algo­rithm design, empha­siz­ing the need for pref­er­ence lan­guages that bal­ance expres­sive­ness, cost, and collectivism—Sloane et al. (2020) cau­tion against “par­tic­i­pa­tion wash­ing” and the poten­tial for exploita­tive com­mu­ni­ty involve­ment. Brat­teteig & Verne (2018) high­light AI’s chal­lenges to tra­di­tion­al par­tic­i­pa­to­ry design (PD) meth­ods, includ­ing unpre­dictable tech­no­log­i­cal changes and a lack of user-ori­ent­ed eval­u­a­tion. Birhane et al. (2022) call for a clear­er under­stand­ing of mean­ing­ful par­tic­i­pa­tion, advo­cat­ing for a shift towards vibrant, con­tin­u­ous engage­ment that enhances com­mu­ni­ty knowl­edge and empow­er­ment. The lit­er­a­ture sug­gests a press­ing need for more effec­tive, inclu­sive, and empow­er­ing par­tic­i­pa­to­ry approach­es in AI development.


  1. Birhane, A., Isaac, W., Prab­hakaran, V., Diaz, M., Elish, M. C., Gabriel, I., & Mohamed, S. (2022). Pow­er to the Peo­ple? Oppor­tu­ni­ties and Chal­lenges for Par­tic­i­pa­to­ry AI. Equi­ty and Access in Algo­rithms, Mech­a­nisms, and Opti­miza­tion, 1–8.
  2. Brat­teteig, T., & Verne, G. (2018). Does AI make PD obso­lete?: Explor­ing chal­lenges from arti­fi­cial intel­li­gence to par­tic­i­pa­to­ry design. Pro­ceed­ings of the 15th Par­tic­i­pa­to­ry Design Con­fer­ence: Short Papers, Sit­u­at­ed Actions, Work­shops and Tuto­r­i­al — Vol­ume 2, 1–5.
  3. Del­ga­do, F., Yang, S., Madaio, M., & Yang, Q. (2023). The Par­tic­i­pa­to­ry Turn in AI Design: The­o­ret­i­cal Foun­da­tions and the Cur­rent State of Prac­tice. Pro­ceed­ings of the 3rd ACM Con­fer­ence on Equi­ty and Access in Algo­rithms, Mech­a­nisms, and Opti­miza­tion, 1–23.
  4. Ehsan, U., & Riedl, M. O. (2020). Human-Cen­tered Explain­able AI: Towards a Reflec­tive Sociotech­ni­cal Approach. In C. Stephani­dis, M. Kuro­su, H. Degen, & L. Rein­er­man-Jones (Eds.), HCI Inter­na­tion­al 2020—Late Break­ing Papers: Mul­ti­modal­i­ty and Intel­li­gence (pp. 449–466). Springer Inter­na­tion­al Pub­lish­ing.
  5. Fef­fer, M., Skir­pan, M., Lip­ton, Z., & Hei­dari, H. (2023). From Pref­er­ence Elic­i­ta­tion to Par­tic­i­pa­to­ry ML: A Crit­i­cal Sur­vey & Guide­lines for Future Research. Pro­ceed­ings of the 2023 AAAI/ACM Con­fer­ence on AI, Ethics, and Soci­ety, 38–48.
  6. Gerdes, A. (2022). A par­tic­i­pa­to­ry data-cen­tric approach to AI Ethics by Design. Applied Arti­fi­cial Intel­li­gence, 36(1), 2009222.
  7. Groves, L., Pep­pin, A., Strait, A., & Bren­nan, J. (2023). Going pub­lic: The role of pub­lic par­tic­i­pa­tion approach­es in com­mer­cial AI labs. Pro­ceed­ings of the 2023 ACM Con­fer­ence on Fair­ness, Account­abil­i­ty, and Trans­paren­cy, 1162–1173.
  8. Robert­son, S., Nguyen, T., Hu, C., Albis­ton, C., Nikzad, A., & Sale­hi, N. (2023). Expres­sive­ness, Cost, and Col­lec­tivism: How the Design of Pref­er­ence Lan­guages Shapes Par­tic­i­pa­tion in Algo­rith­mic Deci­sion-Mak­ing. Pro­ceed­ings of the 2023 CHI Con­fer­ence on Human Fac­tors in Com­put­ing Sys­tems, 1–16.
  9. Sloane, M., Moss, E., Awom­o­lo, O., & For­lano, L. (2020). Par­tic­i­pa­tion is not a Design Fix for Machine Learn­ing. arXiv:2007.02423 [Cs].
  10. Zytko, D., J. Wis­niews­ki, P., Guha, S., P. S. Baumer, E., & Lee, M. K. (2022). Par­tic­i­pa­to­ry Design of AI Sys­tems: Oppor­tu­ni­ties and Chal­lenges Across Diverse Users, Rela­tion­ships, and Appli­ca­tion Domains. Extend­ed Abstracts of the 2022 CHI Con­fer­ence on Human Fac­tors in Com­put­ing Sys­tems, 1–4.

PhD update – September 2023

I’m back again with anoth­er Ph.D. update. Five years after I start­ed in Delft, we are near­ing the fin­ish line on this whole thing. But before we look ahead, let’s review notable events since the pre­vi­ous update in March 2023.


  1. I pre­sent­ed our frame­work, Con­testable AI by Design, at the annu­al NWO ICT Open con­fer­ence, which, for the first time, had an entire track ded­i­cat­ed to HCI research in the Nether­lands. It was an excel­lent oppor­tu­ni­ty to meet fel­low researchers from oth­er Dutch insti­tu­tions. The slides are avail­able as PDF at
  2. I vis­it­ed Ham­burg to present our paper, Con­testable Cam­era Cars, at CHI 2023. We also received a Best Paper award, which I am, of course, very pleased with. The con­fer­ence was equal parts inspir­ing and over­whelm­ing. The best part of it was meet­ing in-per­son researchers who shared my interests.
  3. Also, at CHI, I was inter­viewed about my research by Mike Green for his pod­cast Under­stand­ing Users. You can lis­ten to it here. It is always good prac­tice to try and lay out some of my argu­ments spon­ta­neous­ly live.
  4. In June, I joined a pan­el at a BOLD Cities “talk show” to dis­cuss the design of smart city sys­tems for con­testa­bil­i­ty. It was quite an hon­or to be on the same pan­el as Eef­je Cup­pen, direc­tor of the Rathenau Insti­tute. This event was great because we had tech­no­log­i­cal, design, polit­i­cal, and pol­i­cy per­spec­tives. Sev­er­al guests argued for the need to rein­vig­o­rate rep­re­sen­ta­tive democ­ra­cy and give a more promi­nent role to elect­ed politi­cians in set­ting tech­nol­o­gy pol­i­cy. A report is avail­able here.
  5. In August, the BRIDE project had its clos­ing event. This is the NWO research project that par­tial­ly fund­ed my Ph.D. The event was an excel­lent oppor­tu­ni­ty to reflect on our work togeth­er over the past years. I took the oppor­tu­ni­ty to revis­it the work of Sask­ia Sassen on city­ness and to think through some of the impli­ca­tions of my work on con­testa­bil­i­ty for the field of smart urban­ism. The slides are avail­able at
  6. Final­ly, last week, a short opin­ion piece that lays out the argu­ment for con­testable AI in what I hope is a rea­son­ably acces­si­ble man­ner, was pub­lished on the TU Delft website.
Photo of Eefje Cuppen and I being interviewed by Inge Janse at the BOLD Cities talk show on June 22, 2023—photo by Tiffany Konings.

Eef­je Cup­pen and I being inter­viewed by Inge Janse at the BOLD Cities talk show on June 22, 2023—photo by Tiffany Konings.

Envisioning Contestability Loops

Through­out this, I have been dili­gent­ly chip­ping away at my final pub­li­ca­tion, “Envi­sion­ing Con­testa­bil­i­ty Loops: Eval­u­at­ing the Ago­nis­tic Are­na as a Gen­er­a­tive Metaphor for Pub­lic AI.” I had a great time col­lab­o­rat­ing with Leon de Korte on an info­graph­ic of part of my design framework.

We took this info­graph­ic on a tour of Dutch inter­ac­tion design agen­cies and con­duct­ed con­cept design work­shops. I enjoyed return­ing to prac­tice and shar­ing the work of the past cou­ple of years with peers in prac­tice. My friends at Eend wrote a nice blog post about it.

The analy­sis of the out­comes of these work­shops forms the basis for the arti­cle, in which I explore the degree to which the guid­ing con­cept (gen­er­a­tive metaphor) behind con­testable AI, which I have dubbed the “Ago­nis­tic Are­na” is a pro­duc­tive one for design prac­ti­tion­ers. Spoil­ers: It is, but com­pet­ing metaphors are also at play in the pub­lic AI design space.

The man­u­script is close to com­ple­tion. As usu­al, putting some­thing like this togeth­er is a heavy but grat­i­fy­ing lift. I look for­ward to shar­ing the results and the under­ly­ing info­graph­ic with the broad­er world.

Are we there yet?

Look­ing ahead, I will be on a pan­el along­side the great Julian Bleeck­er and a host of oth­ers at the annu­al TU Delft Design & AI sym­po­sium in October.

But aside from that, I will keep my head down and focus on com­plet­ing my the­sis. The aim is to hand it in by the end of Novem­ber. So, two more months on the clock. Will I make it? Let’s find out!

PhD update – March 2023

Hel­lo again, and wel­come to anoth­er update on my Ph.D. research progress. I will briefly run down the things that hap­pened since the last update, what I am cur­rent­ly work­ing on, and some notable events on the horizon.

Recent happenings

CHI 2023 paper

Stills from Con­testable Cam­era Cars con­cept video.

First off, the big news is that the paper I sub­mit­ted to CHI 2023 was accept­ed. This is a big deal for me because HCI is the core field I aim to con­tribute to, and CHI is its flag­ship conference.

Here’s the full citation:

Alfrink, K., Keller, I., Doorn, N., & Kortuem, G. (2023). Con­testable Cam­era Cars: A Spec­u­la­tive Design Explo­ration of Pub­lic AI That Is Open and Respon­sive to Dis­pute.

I have had sev­er­al papers reject­ed in the past (CHI is noto­ri­ous­ly hard to get accept­ed at), so I feel vin­di­cat­ed. The paper is already avail­able as an arX­iv preprint, as is the con­cept video that forms the core of the study I report on (many thanks to my pal Simon for col­lab­o­rat­ing on this with me). CHI 2023 hap­pens in late April. I will be rid­ing a train over there to present the paper in per­son. Very much look­ing for­ward to that.

Con­testable Cam­era Cars con­cept video.

Responsible Sensing Lab anniversary event

I briefly pre­sent­ed my research at the Respon­si­ble Sens­ing Lab anniver­sary event on Feb­ru­ary 16. The whole event was quite enjoy­able, and I got some encour­ag­ing respons­es to my ideas after­ward which is always nice. The event was record­ed in full. My appear­ance starts around the 1:47:00 mark.

It me. (Cred­it: Respon­si­ble Sens­ing Lab.)
Video of my con­tri­bu­tion. (Pakhuis de Zwi­jger / Respon­si­ble Sens­ing Lab.)

Tweeting, tooting, blogging

I have been get­ting back into the habit of tweet­ing, toot­ing, and even the occa­sion­al spot of blog­ging on this web­site again. As the end of my Ph.D. nears, I fig­ured it might be worth it to engage more active­ly with “the dis­course,” as they say. I most­ly share stuff I read that is relat­ed to my research and that I find inter­est­ing. Although, of course, posts relat­ed to my twin sons’ music taste and strug­gles with uni­ver­si­ty bureau­cra­cy always win out in the end. (Yes, I am aware my tim­ing is ter­ri­ble, see­ing as how we have basi­cal­ly final­ly con­clud­ed social media was a bad idea after all.)

Current activities

Envisioning Contestability Loops

At the moment, the major­i­ty of my time is tak­en up by con­duct­ing a final study (work­ing title: “Envi­sion­ing Con­testa­bil­i­ty Loops”). I am excit­ed about this one because I get to once again col­lab­o­rate with a pro­fes­sion­al design­er on an arti­fact, in this case, a visu­al expla­na­tion of my frame­work, and use the result as a research instru­ment to dig into, in this case, the strengths and weak­ness­es of con­testa­bil­i­ty as a gen­er­a­tive metaphor for the design of pub­lic AI.


In par­al­lel, I have begun to put togeth­er my the­sis. It is paper-based, but of course, the intro­duc­to­ry and con­clud­ing chap­ters require some thought still.

The aim is to have both the final arti­cle and the­sis fin­ished by the end of sum­mer and then begin the ardu­ous process of get­ting a date for my defense, assem­bling a com­mit­tee, etc.

Agonistic Machine Vision Development

In the mean­time, I am also men­tor­ing Lau­ra, anoth­er bril­liant mas­ter grad­u­a­tion stu­dent. Her project, titled “Ago­nis­tic Machine Vision Devel­op­ment,” builds on my pre­vi­ous research. In par­tic­u­lar, one of the chal­lenges I iden­ti­fied in Con­testable Cam­era Cars, that of the dif­fer­en­tial in infor­ma­tion posi­tion between cit­i­zens and experts when they col­lab­o­rate in par­tic­i­pa­to­ry machine learn­ing ses­sions. It’s very grat­i­fy­ing to see oth­ers do design work that push­es these ideas further.

Upcoming events

So yeah, like I already men­tioned, I will be speak­ing at CHI 2023, which takes place on 23–28 April in Ham­burg. The sched­ule says I am pre­sent­ing on April 25 as part of the ses­sion on “AI Trust, Trans­paren­cy and Fair­ness”, which includes some excel­lent-look­ing contributions.

And before that, I will be at ICT.OPEN in Utrecht on April 20 to present briefly on the Con­testable AI by Design frame­work as part of the CHI NL track. It should be fun.

That’s it for this update. Maybe, by the time the next one rolls around, I will be able to share a date for my defense. But let’s not jinx it.

Tensions in the professional field of design

I liked a pas­sage in a Kees Dorst paper on “aca­d­e­m­ic design” so much, I turned it into a lit­tle diagram.

Ten­sions in the pro­fes­sion­al field of design. (PDF)

Note that these ten­sions are inde­pen­dent of each oth­er. The dia­gram does not imply two “sides” of design. At any giv­en moment, a design activ­i­ty can be plot­ted on each axis inde­pen­dent­ly. This is also not an exhaus­tive list of ten­sions. Final­ly, Dorst claims these ten­sions are irreconcilable.

The orig­i­nal passage:

Con­tem­po­rary devel­op­ments in design can be described and under­stood in much the same way. The pro­fes­sion­al field that we so eas­i­ly label ‘design’ is com­plex, and full of inner con­tra­dic­tions. These inner ten­sions feed the dis­cus­sions in the field. To name a few: (1) the objec­tives of design and the moti­va­tion of design­ers can range from com­mer­cial suc­cess to the com­mon good. (2) The role and posi­tion of the design­er can be as an autonomous cre­ator, or as a prob­lem solver in-ser­vice to the client. (3) The dri­ve of the design­er can be ide­al­is­tic, or it can be more prag­mat­ic (4) The result­ing design can be a ‘thing’, but also imma­te­r­i­al (5) The basis for the process of design­ing can be intu­itive, or based on knowl­edge and research… Etcetera… The devel­op­ment of the design dis­ci­plines can be traced along these lines of ten­sion — with design­ers in dif­fer­ent envi­ron­ments and times chang­ing posi­tion rel­a­tive to these fun­da­men­tal para­dox­es, but nev­er resolv­ing them. Ulti­mate­ly, the real strength and coher­ence of design as a field of pro­fes­sions comes from rec­og­niz­ing these con­tra­dic­tions, and the dynam­ics of the field is a result of con­tin­u­ous exper­i­men­ta­tion along the rifts defined by them. Rather than a com­mon set of prac­tices and skills that design­ers might have [Cross, 1990] it is these inner con­tra­dic­tions in design that define its cul­ture, its men­tal­i­ty. Design research should be an active force in these dis­cus­sions, build­ing bridges between them where pos­si­ble. Not to resolve them into a mono­lith­ic Sci­ence of Design, but advanc­ing the dis­cus­sion in this dynam­i­cal­ly shift­ing set of relations.

Dorst, K. (2016, June 27). Design prac­tice and design research: Final­ly togeth­er? Pro­ceed­ings of DRS 2016. Design Research Soci­ety 50th Anniver­sary Con­fer­ence, Brighton, UK.

Citizen participation in “The End of the End of History”

Below are some choice quotes on “cit­i­zen par­tic­i­pa­tion” from chap­ter 8 of The End of the End of His­to­ry, a rec­om­mend­ed book on our recent glob­al polit­i­cal his­to­ry. I feel like many of us in the par­tic­i­pa­to­ry tech­nol­o­gy design space are com­plic­it in these prac­tices to some extent. I con­tin­ue to grap­ple with alter­na­tive mod­els of mass demo­c­ra­t­ic con­trol over technology.

The Cen­ter-Left will pro­pose a range of mea­sures designed to pro­mote “civic engage­ment” or “com­mu­ni­ty participation.”

Cit­i­zens’ sum­mits, juries and pan­els all aim at par­tic­i­pa­tion rather than pow­er, at the tech­no­crat­ic incor­po­ra­tion of the peo­ple into pol­i­tics in order to man­age away conflict.

Like­wise the pop­u­lar­i­ty of delib­er­a­tive modes of engage­ment, delib­er­a­tive stake­hold­er events or work­shops are char­ac­ter­is­tic tools of tech­no­crat­ic do-good­ers as they cre­ate the sim­u­lacrum of a demo­c­ra­t­ic process in which peo­ple are assem­bled to pro­vide an osten­si­bly col­lec­tive solu­tion to a prob­lem, but deci­sions lack a bind­ing qual­i­ty or have already been tak­en in advance.

Though unable to gain trac­tion at a transna­tion­al lev­el, the Left may find some suc­cess in munic­i­pal pol­i­tics, fol­low­ing the 2010s exam­ple of Barcelona.

Side­step­ping […] ani­mus toward Big Tech com­pa­nies, [tech solu­tion­ism (Moro­zov, 2013) and the ide­ol­o­gy of ease (Green­field, 2017)] may come to be applied to non-mar­ket activ­i­ties, such as solv­ing com­mu­ni­ty prob­lems, per­haps at the lev­el of munic­i­pal government.

Sov­er­eign, nation­al pol­i­tics – which neolib­er­al­ism was designed to defang – will remain beyond the grasp of the Left. Pro­gres­sives will pre­fer instead to oper­ate at the munic­i­pal, the every­day or the supra­na­tion­al lev­el – pre­cise­ly the are­na to which neolib­er­al­ism sought to dis­place pol­i­tics, to where it could do no harm.

Hochuli, A., Hoare, G., & Cun­liffe, P. (2021). The End of the End of His­to­ry. Zero Books.

De opkomst van de meritocratie

Thi­js Klein­paste heeft een mooie boekbe­sprek­ing van Michael Young’s De opkomst van de mer­i­to­cratie in de Ned­er­landse Boekengids. Een paar pas­sages die ik vooral sterk vond hieronder.

De grote ver­di­en­ste van Young is dat hij inzichtelijk maakt hoe onschuldige principes als ‘beloning naar ver­di­en­ste’ volkomen kun­nen ontsporen als ze wor­den ingezet bin­nen een verder onveran­derd soci­aal en economisch stelsel. Con­creet: som­mi­gen een uitverko­ren posi­tie geven in een maatschap­pelijke hiërar­chie en anderen opdra­gen om hun plek te kennen. 

Het klassen­be­lang van de mer­i­to­cratie is abstracter. Het belan­grijk­ste is om allereerst een klasse of kaste te bli­jven om zo de voorde­len daar­van te kun­nen bli­jven oog­sten. In iedere mod­erne staat wordt macht uit­geoe­fend – of mer­i­to­cratis­ch­er gezegd: moet er bestu­urd wor­den – en als er dan toch een kaste moet zijn die deze taak vervult, laat dat die van de hoogst gediplomeer­den zijn. De mer­i­to­cratie repro­duceert zichzelf door deze gedachte mee te geven aan elke nieuwe licht­ing die tot haar uitverko­ren rangen toe­treedt: dat zij de juiste, met recht geroepen groep is om de wereld te orde­nen. Niet de arbei­der­sklasse, niet de ongelei­de democ­ra­tie, niet het gekri­oel van belan­gen­groep­jes – maar zij. Alle mater­iële voorde­len van de mer­i­to­cratie vloeien voort uit het in stand houden van die uitverko­ren status.

Te vaak lijkt de gedachte te zijn dat verte­gen­wo­ordig­ing en het bedi­enen van belan­gen onprob­lema­tisch in elka­ars ver­lengde liggen. Om die zelfge­noegza­amheid te door­breken is ken­nelijk iets stel­ligers nodig, zoals de gedachte dat waar man­agers en bestu­ur­ders zijn, er ges­taakt moet kun­nen wor­den: dat waar macht wordt uit­geoe­fend en waar aan­wi­jzin­gen wor­den gegeven, zij die de aan­wi­jzin­gen moeten opvol­gen kun­nen stem­men met hun voeten. Dat con­flict omar­md wordt en niet wordt gezien als iets wat gevaar­lijk is voor de maatschap­pelijke lieve vrede, de ‘economie’, of zelfs de democ­ra­tie. Con­flict is ongetwi­jfeld gevaar­lijk voor de hege­monie van de man­ag­er en diens klasse van droomkoninkjes, en daarmee voor de soev­ere­initeit van de mer­i­to­cratis­che orde, maar dat gevaar is zow­el heilza­am als noodza­ke­lijk. Een van de lessen van het boek van Young is immers ook dat je moet kiezen: zelf een rev­o­lu­tie mak­en, of wacht­en tot die uitbreekt.

Zelf lezen:

I might start using this place more fre­quent­ly again. Most­ly for microblog­ging of inter­est­ing things I notice that are relat­ed to my research. I dunno.