Design: Weekly Summary (October 13-19, 2025)

There’s a funny little tug-of-war happening in design writing this week. On one side you’ve got people talking about the tactile, the old-school, the smell-of-ink sort of craft. On the other side you’ve got hot takes about interfaces, invisible tech, and AI doing things that used to be quiet human chores. And running through many of the posts is a shared concern — how to keep tools helpful instead of hungry, how to preserve the human bit while we reach for speed or beauty or novelty.

I would describe the tone as a mix of nostalgia and impatience. To me, it feels like folks are trying to steady the steering wheel while the car’s getting a software update. If that sounds like a jumble, good — that’s kind of the point. Read these pieces and you get fragments of a conversation, not a single sermon. If you want more, each author’s post is worth a click. There’s more detail there, and you’ll spot the parts I only hinted at.

Making and memory: craft, print, and strange benches

A few writers this week went back to the physicality of making. It’s an easy place to linger. Some of these pieces feel like sitting with someone in a workshop while they point out a nick in the wood and tell you about the time it happened.

Kate Bingaman-Burt shares notes from teaching at PSUGD and her steady practice of drawing and collecting found photos. Her cadence is gentle but steady. I’d say the post reads like a studio visit — pick up a sketchbook, flip a few pages, and someone hands you a coffee. The practical parts — a reading list, a holiday calendar of local artists — are small scaffolds for daily work. It’s the kind of thing that quietly insists: do a little bit, look closely, repeat.

Peter Rukavina writes about printing coffee bags for a local roastery. Letterpress talk is never high-gloss. It’s about timings, plates, and the tiny stakes that make an edition feel like something. His bag for ROW142 (now Receiver Coffee) is like a postcard for a move — sentimental, and also stubbornly particular. You can almost feel the platen clunking.

Then there’s Christopher Schwarz with a bench story from Holland that quietly breaks the rules about wood movement. He’s practical, the sort of person who will shrug and say, “it works,” while dust settles on a design rulebook. That bench is a reminder: rules are useful, but lived experience sometimes says otherwise. It’s like learning to cook — sure, follow the recipe, but when the pan’s screaming you make a call.

And if you like toys that think they’re alive, Jay Springett gives a nice compare-and-contrast of Sony’s AIBO and the Furby. The two little guys are design essays in miniature: AIBO reaching for emotional depth through autonomy, Furby doing a lot with cost-effective tricks and charm. The piece pulls back the curtain on intention: some designs aim to be wondrous; others aim to be affordable and clever. Both leave fingerprints on how we imagine companionship with machines.

These posts share a feeling: design is a craft that holds memory. Materials store stories. Mistakes teach more than manuals. And a good day in a shop looks a lot like a good conversation.

Liquid Glass and the fight over content vs controls

If there is a controversy this week, it’s the Liquid Glass debate. Several posts keep circling the same sore spots: transparency that hides things it shouldn’t, flouncy animations that add whimsy rather than clarity, and shrinking touch targets that make older fingers curse. This feels less like punditry and more like people saying out loud what they bump into every day.

Nick Heer and Michael J. Tsai both push at the idea that design should prioritize content at the expense of controls. I would describe their argument as plain: tools should be ready when you are. Heer uses cooking analogies — which I liked — likening interface controls to stove knobs. You wouldn’t hide the knob when you need the heat, right? Tsai’s “Liquid Glass Is Cracked” essay is more of a catalog of annoyances: transparency that obscures, animations that distract, search that’s less practical than it should be. Both posts are angry but practical. They don’t just complain; they point to specific interactions that feel worse now than they did last year.

Chris touches on this too, but with a different tilt. He mixes thoughts on iOS security with design gripes and some coders’ humility. It’s a little bit of a tangent — in a good way. It reads like a kitchen-table conversation where someone switches the topic from UI to faith to a weird bug they chased all week. That mix is human. It reminds you that design decisions land on people who also worry about other stuff.

There’s a chorus here: design choices are not neutral. Tiny visual choices — a shadow, a fade, a touch area — change how people use a product. It’s not just about pretty glass. It’s about whether you can get a thing done without asking for mercy.

Slow AI, ADRS, and the new tools doing the heavy lifting

Two threads about AI showed up that feel like bookends. On one side, there’s the careful design thinking around long-running AI tasks. On the other, there’s full-on automation for systems research.

Jakob Nielsen has a piece on “Slow AI” and the design around long jobs — things that run for hours or days. The practical advice is gold if you ever run a machine-learning experiment or a render job. Talk of checkpoints, explicit run contracts, progress updates, and the odd breadcrumb for users feels like common sense made explicit. Nielsen draws a line back to batch processing from decades ago, which is useful. It’s like reminding someone that slow cooking still needs a timer and a lid. You don’t just throw ingredients in and leave. You check on it. He’s pushing for observability and user control rather than opaque magic.

Then there’s Grigory Sapunov with ADRS — AI-Driven Research for Systems. This is the one that leans into the dramatic: teaching LLM ensembles to generate and evolve algorithms, then verifying them with the same kind of rigorous testing humans use. The pitch is bold. It’s not small automation; it’s a re-think of who sits at the bench. Sapunov and co. show examples where the machine outperforms human designs. That’s equal parts thrilling and unsettling.

Those two posts together make for a neat tension. Nielsen says: design for humans, especially when tasks take time. Sapunov says: give more autonomy to AI when it can reliably test and verify results. It’s like deciding whether to use a sous-vide cooker with a digital timer (you still check the meat) or to let the robot chef do the whole dinner party.

Interfaces as extensions of the body

Vanessa Chang wrote about how technology has become an invisible extension of our bodies — how interfaces have moved from hands to voice to eyes. That essay felt quiet and wide at once. It’s historical when it needs to be and personal when it needs to be. The core worry is familiar: if we let corporate metrics guide interface design, the result will be something that nudges people toward engagement, not wellbeing.

Chang argues for a design philosophy that foregrounds human needs. Her point about interface evolution is worth repeating: each change reshapes the way we move, speak, and think. It’s small, cumulative edits to human behavior. To me, it feels a bit like how a city changes when a new train line opens — at first it’s just one extra stop, later it’s a whole neighborhood that rearranges itself.

This ties back to the Liquid Glass posts. When design choices privilege surface or novelty over legibility and control, our bodies adapt in awkward ways. We stretch to tap a smaller target. We learn new steps to get to the old content. Chang’s essay frames those changes as cultural, not merely technical.

Finding and curating content — the indie web corner

There’s also a mini-thread on how writers and readers find one another. Creativerly rails a bit about the limits of Substack and suggests alternatives. The post reads like a travel guide for blog hunters: a list of paths less traveled and a few apps that make the trip easier. Meco, Vivaldi, Webstudio — these are tools that try to make the internet feel more like a bookstore and less like a shopping mall.

That dovetails with small updates like Mastodon’s Packs, which aims to help folks curate their social spaces. It’s a small nudge toward decentralized discovery. The theme here is obvious if you’ve spent a few too many hours on feed roulette: people want tools that lead to depth, not dopamine. That’s an argument that keeps coming back in design circles. Make it easy to find the good stuff, not impossible.

Systems thinking, people problems, and the agony of “just do it”

A different kind of design story appears in the productivity realm. The Product Picnic digs into the hazards of a “just do things” mindset. The argument is less about speed and more about coordination. Finish a ticket, ship a quick fix — sure. But if nobody thinks about the long game, systems decay.

I’d say this is a people-problem essay disguised as a product one. It leans on frameworks and examples to show how short-term outputs can hollow a system out. There’s a quiet indictment of hero culture: the person who saves the day with a midnight patch is different from the person who prevents the disaster through design. The post wants leaders to favor stability and care over the shiny rush of immediate delivery.

This resonates with Nielsen’s “Slow AI” thinking. Both pieces value deliberate work. Both push against the cult of immediacy. If you have a team that’s always in triage mode, you’ll see both poor UX and brittle systems.

Design’s small pleasures and historical details

A few pieces simply celebrate small things. Scott Sidler dives into lunette windows — their history from basilicas to front doors. It’s a micro-lesson in architectural detail. The post is a little like walking down a row of old houses and pointing out the little half-moons above doors. There’s a curious pleasure in specificity. It’s the opposite of a headline. It’s the kind of knowledge you tuck away for a future moment when you’re fixing a trim or arguing with a contractor.

Scott Boms in Through Lines offers a mix of cultural pickings — music, design for infrastructure, a LEGO magazine, and a nudge about AI’s impact on creativity. It’s a grab bag, and it’s delightful because it reminds you design shows up everywhere. It’s not only about screens; it’s also about songs and bricks and how we run offices.

Small machine stories and design legacies

The juxtaposition of Furby and AIBO also feeds into a larger conversation about agency and illusion in design. The story Jay Springett tells is, quietly, about promises. AIBO promised companionship through complex systems. Furby promised a kind of life through clever scripting and economy. Both shaped expectations of what future robots should be. This is where nostalgia is useful: it helps trace design lineage. Modern social robots and companion devices still borrow ideas from those two little guys.

And then there’s the Sapunov paper, which hints at a future where machines can meaningfully redesign parts of their own field. That’s not a toy story. It’s a fundamental shift in how research might be done. If systems research becomes more about guiding AI and less about elbow-grease tuning, then design roles shift toward governance, validation, and ethics. That’s a big change. It’s also a natural result of tooling getting smarter.

Small grievances, big implications

A running pattern this week is how small design choices become big user experiences. Liquid Glass debates are a clear example: a tweak to opacity or animation timings ripples through everyday use. Slow AI posts show that a slip in clarity for long tasks produces anxiety and mistrust. The ADRS work reveals that automation without transparent verification changes who we count as the expert.

In plain terms: design is the thing people argue about when the stakes are both tiny and huge. Tap targets, a jar that’s hard to open, a search bar that suddenly hides the query — these irritants are small but they stack up. Stack enough of them and you get a different habit, a different expectation.

Little detours and kitchen-table thoughts

There were a few small tangents worth noting, because they tell you about the blogosphere’s shape. Nick Heer returns to the cooking analogy more than once. It’s helpful and sometimes comfortingly repetitive — like someone repeating a recipe so you don’t forget the salt. Chris does that kitchen-table switcheroo I mentioned: security, UI, and spiritual reading all in one post. It’s human and messy, and that’s fine.

Also, Mastodon’s Packs and Creativerly’s push for discovery feel like minor technical fixes with social implications. They’re small changes in how content circulates, but design is often about how things move through networks, not only pixels.

If a theme has a shadow side this week, it’s the risk of design being mistaken for style. When aesthetics lead and function follows, you get Liquid Glass cracks. When automation is celebrated without thinking who benefits, you get ADRS-like disruptions with little guardrails. It’s not a condemnation of newness; it’s a reminder to ask: who’s being served, and who’s being made to change?

These posts — the practical, the nostalgic, the speculative — feel like parts of a single conversation about what design should hold on to and what it should let go. They don’t resolve anything. They don’t pretend to. But they do point, repeatedly, to a few small ideas: respect materials, make controls obvious, design long tasks with people in mind, and treat automation as a partner, not a replacement.

If you want to wander deeper into any of these lanes, follow the links to the original pieces. They each have that extra detail and that extra voice — the kinds of specifics that make the argument stick. Read a craft post, then a UI rant, then a systems piece. It’s like having a walk across town: each neighborhood has its own smell and its own furniture. You’ll see the connections if you look for them.

And if one more thing can be said in the kind of blunt way you hear from someone folding laundry while they talk: design gets interesting when it remembers people. That’s simple and a little obvious, but it keeps coming up for a reason.