The Field Dispatch

Nila Rezaei's keynote made the social case. Her team once designed away the friction in a shared-house chore system — and her verdict on the result: "by removing that friction, we've made one person's life easier, but we engineered everyone else out of that interaction" Rezaei ▸ ≈28:05 — "Wearing the costume of good design, easily frictionless, but potentially alone." Her design instruction: work out "which friction is actually connection in disguise" Rezaei ▸ ≈30:12 and protect it. The wider argument: "the moment you're only just consuming something, you stopped being part of it. A community of users isn't really a community, it's just a customer list." Rezaei ▸ ≈32:34

Emma Buckland made the risk case from seventeen years in banking and superannuation UX. Real mobile use arrives in three-to-four-second bursts amid interruption — research documented twenty years ago, on her slides via Jan Chipchase: "Unless you have something like a pacemaker, there really isn't such a thing as a 24/7 user." Buckland ▸ slide Entertainment survives that; consequential tasks don't: "When TikTok loses our attention, we lose a video. When a consequential interaction loses our attention, we might send the wrong amount, pay the wrong person, miss the deadline, incur a fee" Buckland ▸ ≈10:51 — and "You can't break a $50,000 transfer in entertaining twenty second chunks." Buckland ▸ ≈09:54 Her design targets are clarity, continuity and control — where control includes reversibility: "Does control have to end the moment you press Submit?" Buckland ▸ slide

Patima Tantiprasut's case study was PetRescue, where a pet with up to a hundred applicants produces ninety-nine people who miss out. Two slides carried the reframe: "A rejected purchase is a failed transaction." and "A rejected adoption is a rejected offer of help." Tantiprasut ▸ slides The 2018 mass-adoption campaign showed what demand marketing without whole-journey design does: "Instead of zeroing out the shelters and getting every single pet adopted, we ended up with thousands of really disheartened disappointed people who wanted to participate, who wanted to help, and they weren't successful in adoptions." Tantiprasut ▸ ≈09:30 Rejection has downstream consequences — "this categorically became another set of emotional injury and it fueled the next decision" Tantiprasut ▸ ≈16:52 — often toward backyard breeders. The redesign added deliberate friction (an enquiry wall), adoption-policy levels, two-way messaging and bulk outcome notifications; "no reply" complaints fell from 50% to 0.007%.

Cornel Steyn and Jess Witt named the same move from consulting practice — "So just temporarily, sometimes friction can be added for a better goal" Steyn & Witt ▸ ≈17:49 — with Instagram's nudity-protection feature as the test: blurring works, but "60% are still viewed. So where does that line of responsibility lie?" Steyn & Witt ▸ ≈17:36 Sakib Jalil supplied a measured example: a bank login deliberately slowed, because trust reads speed as carelessness — "A bank service where we increased the perception of safety with 12 versions of delay, the sweet spot was 2-3 seconds." Jalil ▸ slide And Mina Giang marked the boundary from healthcare: form simplification is not free — "when it comes to health care, that minimum information collected can cause massive delays and impacts downstream." Giang ▸ ≈16:45

Ben Shelton and Cordelia Prangley brought the theory. Their opening slide: "Interfaces compete for cognition" Shelton & Prangley ▸ slide — and the diagnosis: "Most interfaces today are designed for functionality and engagement. Far fewer are designed around human cognitive limits." Shelton ▸ ≈03:13 Their definitional correction matters, because the term has gone loose in UX: "Cognitive Load Theory is specifically about the limits of working memory" Shelton & Prangley ▸ slide — not a general word for annoyance or visual mess, and not all bad: intrinsic, extraneous and germane load differ. The design conclusion: "So good design protects cognitive capacity. And great interfaces reduce unnecessary demand on working memory. They help users stay orientated, focused, and confident." Prangley ▸ ≈11:52 And it is measurable — self-report instruments like NASA-TLX, physiological measures, dual-task performance testing — which they applied to a University of Newcastle degree planner now in front of 23,000 students.

Fatemeh Aminpour's day-two keynote moved the unit of analysis from the screen to the journey. From walking interviews with fifteen neurodivergent adults in Sydney's public spaces: "They hit the the wall fast, and the hardest, but the wall is there for everyone." Aminpour ▸ ≈07:28 Her mechanism for sensory load: "if it cannot be predicted, it cannot be backgrounded. And if it cannot be backgrounded, it has to be processed." Aminpour ▸ ≈25:48 The costs are concrete — "That could actually cost us two days of recovery. From getting lost. I'm not exaggerating, I'm just quoting participants." Aminpour ▸ ≈13:48 — and so are the remedies: predictability, sensory clarity, and recovery points. "So a bench is not just a bench, it's the thing that makes the next twenty minutes possible." Aminpour ▸ ≈29:13 Her audience question bridged city to product: "Where in your product are you asking users to be vigilant instead of present?" Aminpour ▸ slide

Buckland's contribution to this theme is the empirical floor: attention in the field fragments into bursts of seconds, so "Mobile interaction cannot always rely on continuous attention" Buckland ▸ slide — and her reframe follows: "What happens if we stop treating attention as something the user owes the interface". Buckland ▸ slide Her clarity principle: "Because clarity isn't about capturing more of our attention, It's about needing less of it." Buckland ▸ ≈15:34

Josephine Abbatangelo added the variability the other three imply: "People don't have the same capacity every day. If a system only works when someone is at their best, it's not really working for everyone." Abbatangelo ▸ ≈26:42 Her examples are the flows that spend capacity where there is least — "These sorts of flows ask the most of us at the moment we have least to give" Abbatangelo ▸ ≈18:34 — and the ones that budget it well: Airbnb's listing flow, where "The complexity didn't go anywhere. It was just broken into pieces small enough to carry and the product remembers where you got to." Abbatangelo ▸ ≈23:12

Maria Sereno's talk built the frame. "Exclusion is our default." Sereno ▸ ≈04:40 Not personal malice — systems: "Design doesn't occur in a vacuum. Our ideas and we ourselves are products of an environment that excludes by default." Sereno ▸ ≈07:34 Her counterfactual: in a world that didn't exclude, "inclusive design wouldn't exist. It wouldn't be necessary. It would just be design. Accessibility would just be usability." Sereno ▸ ≈05:18 And her boundary: "inclusive design is not an accessibility compliance checklist. Just with a few more items about gender added on. It's not one solution fits all. And it's not optional." Sereno ▸ ≈03:37 Her toolkit is deliberately startable this week: examine assumptions, incorporate intersectionality — citing Kimberlé Crenshaw — and centre lived experience.

Josephine Abbatangelo gave the argument its economics. One in five Australians live with a disability, most invisible, and 56% fear career consequences from disclosing — so products meet these users constantly and never know it. Her framing slide: "Every product you've designed has either charged this tax or waived it. Most of us don't know which." Abbatangelo ▸ slide The tax is the compensating labour — "It's carrying the body you're in and doing the work in. And it's not passive, it's labor." Abbatangelo ▸ ≈13:35 She named her own position: "I'm a senior product designer. And I live with HEDS and ADHD. I've spent years building products and years quietly compensating for the ones that weren't built for me." Abbatangelo ▸ ≈10:32 The design choice: "So you have two options. Wait for them to disclose or design and recruit as if they're already here." Abbatangelo ▸ ≈16:52

Sam Alderton-Johnson took the power question head-on, in conversation on the day-one closing stage. On the label: "co design became co opted into every second project being labeled a co design journey when almost it was anything but." Alderton-Johnson ▸ in conversation On the line that separates the real thing: "that's where the code design becomes harmful. Is when we keep calling it co design, but we're not relinquishing any power in decision making." Alderton-Johnson ▸ in conversation His training exercise makes the point without argument — "put your hands up if you've ever designed a project plan alongside the communities that you're intending to do this work with" Alderton-Johnson ▸ in conversation — no hands go up. The stakes of his practice areas are not abstract: "Particularly in New South Wales, we have more kids removed from their families than at the height of the stolen generation." Alderton-Johnson ▸ in conversation And the precondition is relational: "We have to like each other. Have to trust each other before we go on that journey."

Nadine Raydan and Emily Smith showed what sharing power looks like inside a service. ReachOut's PeerChat is staffed by a peer workforce whose lived experience is the qualification — "Peer Work is a recognised profession in the mental health sector that acknowledges lived/living experience as a qualification and type of expertise" Raydan & Smith ▸ slide — and that workforce co-designed the service's rebuild, its safety mechanisms and its data models. The re-architecture from booked sessions to on-demand followed young people's actual help-seeking: "Young people engage with on-demand at more than twice the rate of the booked model." Raydan & Smith ▸ slide Even their taxonomy work is inclusion work: "So things like trolling really hides underlying needs and it's not inclusive, it's not strength based. So we moved away from that altogether." Raydan & Smith ▸ ≈18:54

Aminpour's keynote belongs to this theme as well as the last: her neurodivergent participants are what she calls indicator species — the people who reveal a system's assumptions first, on behalf of everyone. And Tantiprasut's ninety-nine unsuccessful applicants are the same structural insight applied to a journey: the people your design doesn't see are still in it.

MC Monsalve led the ABC's thirty-person design team through the arrival of AI, and rejected the role-collapse narrative directly: coders becoming designers and designers becoming coders is "not true", a reductive simplification — because "when almost anything can be made, the real question becomes, what should we make and why?" Monsalve ▸ ≈04:09 Her leadership choice is on one slide: instead of mandating tools, creating top-down rules and forcing change — "Create space. Create focus time. Let the team decide." Monsalve ▸ slide The outcome she reported: "it hasn't been the prototypes, it hasn't been the tools. It was creating the conditions for the team to explore together" Monsalve ▸ ≈20:34 — and an AI prototype hackathon that built digital twins of three flagship products in days rather than months.

Jessie Pahng stated the caution most directly, from enterprise observability. Her paired slides: "AI reduces the cost of execution." then "It also reduces the cost of executing the wrong assumption." Pahng ▸ slides — the second shown over a polished AI-generated dashboard, with the spoken warning "Polished doesn't mean right." Pahng ▸ ≈13:21 Her rule calibrates speed to consequence: "But if consequences are high, slow down, and validate." Pahng ▸ ≈15:08 She also reported the trust behaviour in the field: users copy AI incident analysis into other tools to confirm it before acting.

Peter Barber's talk was the theme's structural argument. From work where "Our consequences were measured in people and not pixels" Barber ▸ ≈07:44: "Because we don't trust quality, we trust accountability." Barber ▸ ≈09:50 The slide that anchors it: "AI CAN NOT CARRY ACCOUNTABILITY" Barber ▸ slide — and his account of what AI does to judgement: "AI doesn't remove human choice. It freezes it. It repeats it. It hands it back to you disguised as quality." Barber ▸ ≈20:11 The question he wants asked instead of "what will AI take from design": "What unintended agency are we surrendering to our systems?" Barber ▸ ≈18:13 His practice: three questions before any consequential decision ships — "What are we trying to make true? / Who will have to live with it? / Am I prepared to stand behind what follows?" Barber ▸ slide

The others marked edges. Sakib Jalil's bet: "I believe the next big breakthrough will be enjoyment because efficiency is no longer something we need to do anymore because we have AI to do that" Jalil ▸ ≈07:11 — with a warning from the field, startup founders telling researchers "what are you going to find out that AI cannot find out?" Jalil ▸ ≈14:20 Cornel Steyn and Jess Witt put the hidden costs on the table: "A lot of our work these days happens through a single chat interface. That interface is so simple. Too simple. It hides the true cost of all the models that's been trained before" Steyn & Witt ▸ ≈13:34 — and on the industry's energy self-reporting: "The companies reporting on this are the ones financially gaining from this. So I believe the numbers reported seem very low." Steyn & Witt ▸ ≈14:18 Maria Sereno reported generative AI reproducing the bias in its training data in her own tests. And in ReachOut's service, the first thing many young people ask is whether they're talking to a machine: "For some young people, the very first need that they have when starting a conversation is to verify that they're actually speaking to a human and not AI." Raydan & Smith ▸ ≈03:54

Nathan Shedroff's closing keynote named the territory: "I'm here to tell you there's a third diamond. And it's invisible to most people" Shedroff ▸ ≈03:17 — the corporate strategy phase that runs before the double diamond, where the brief is decided. "Well, that decision got made way over here. By the time you get it in development, it's too late." Shedroff ▸ ≈05:18 His assessment of what happens in that phase: "I truly believe that traditional strategy, which is the same most strategy, is terrible. It's slow. It's incredibly sloppy" Shedroff ▸ ≈09:19 — with McKinsey's own research on his slide: "70% of all global companies are "zombie companies" (meaning that they create little value and barely service their debt)". Shedroff ▸ slide The opportunity he sees: designers already hold the qualitative insight strategy lacks — "Qualitative Value Is Invisible to Traditional Business Tools!" Shedroff ▸ slide — and his slide-sized conclusion: "EVERYONE SHOULD BE INVOLVED IN STRATEGY". Shedroff ▸ slide

Mina Giang's two healthcare case studies are the same argument with regulatory specifics. Australia's Active Script List reached her product with its consent flows fixed by conformance requirements — so when the experience failed, "I didn't design this. I inherited it." Giang ▸ ≈04:48 Her slide: "When something doesn't work, they blame the product in front of them." Giang ▸ slide She found gaps the rules hadn't considered — "I asked what happens when a relationship changes?" The answer: "There wasn't a way to revoke access." Giang ▸ ≈09:09 — and her response is the practice this theme recommends: document the risk, write impact statements, join the standards working groups (CSIRO's Sparked FHIR accelerator takes non-technical contributors), and "If you're not there, someone else is deciding the experience your users will inherit." Giang ▸ slide

Elise Loeuille's three years on unplanned disruption at Transport for NSW relocated the problem inside the organisation. The visible brief was customer communication during disruption; the finding was on her slide: "It is not a lack of effort / It is a lack of shared visibility." Loeuille ▸ slide — "But behind this problem, you have the hidden problem, which is the coordination." Loeuille ▸ ≈07:31 Where it breaks: "Coordination breaks between teams, not within them." Loeuille ▸ slide And what design did about it: "the first thing we redesigned was not the service. Even if I'm called service designer. We started to redesign the conversation." Loeuille ▸ ≈11:23 Ninety-plus stakeholders, paired customer and employee journey maps, one co-created disruption lifecycle — and her summary of the deliverable: "The deliverable was not the artifact." Loeuille ▸ ≈15:34 A participant's line from the shared-language slide: "I never realised my decision affected someone six steps later." Loeuille ▸ slide

Amir Ansari asked what survives once you get upstream. From three first-hand cases of mature design practices dissolved by acquisitions and strategy resets — "And the work that I had done over the two years disappeared overnight. Literally." Ansari ▸ ≈11:25 — his slide-sized conclusion: "Important does not mean permanent". Ansari ▸ slide What endures is people: "Job of a leader is not to make every decision. It's to teach their team how to think." Ansari ▸ slide — and "Do not create followers. Create people who multiply judgement, craft and care." Ansari ▸ slide Sarah Fennelly and Rachael Wilson worked the same seam at Deakin University Library, building a playbook so 140 non-designer staff could do UX work — and reported the honest result: "But what we learned also was that awareness doesn't automatically translate into adoption." Fennelly & Wilson ▸ ≈22:54 Their closing slide formula: "Culture = Behaviours + Reinforcement + Leadership Example". Fennelly & Wilson ▸ slide