Design: Weekly Summary (November 17-23, 2025)

I would describe this week in design as a messy, interesting stew. Bits of public space, furniture, and brand strategy rubbed up against heated arguments about who really knows what UX is. There was worry about sameness and a side of shiny new AI tools that promise to draw the picture for you. To me, it feels like the field is trying to be two things at once: careful craft and fast-moving software. The tension shows up in a lot of the posts, in different ways.

A bridge, four proposals, and the feel of a place

The simplest story is the one about Clapham. Ian Mansfield writes about a design competition to re-clad the Clapham High Street railway bridge. Four shortlisted ideas. An exhibition to come. You know the sort of thing: civic-minded, small-scale, and oddly revealing. I’d say projects like this show how much people still care about local character. A bridge is not just steel and paint. It’s a threshold. It’s the first hello to a street. Treat it like a museum piece and you miss the bus. Treat it like a billboard and you lose a neighbourhood.

I would describe the proposals as conversations more than statements. Some lean on art and cladding, others try to nudge pedestrian flow and public life. The post hints at who’s thinking about what: tactile surfaces, light at night, and how a structure can read as either civic or corporate. To me, it feels like London afternoons — a bit damp, a bit opinionated, people shouting over cups of tea. If the exhibition shows the models and drawings, it’ll be worth a wander. There’s an itch here: how much hand should design have in a public place versus how much should be left to everyday use. Folks will argue. They always do.

The end of the unicorn myth — UX is a craft, not a costume

This one cut sharp. Jason Clauss takes apart the idea of the ‘unicorn’ designer — the person who claims to be master of UX, UI, product, visual art, interaction, and probably pottery on the side. The claim: many of these unicorns never learned what real UX work is. They leaned on pretty mockups and intuition, not research and measurable outcomes. Now, with AI rolling up its sleeves, those gaps are getting exposed.

I’d say the tone is a bit accusatory, but it’s also helpful. The argument isn’t just gatekeeping. It’s practical. When someone treats design like art for art’s sake, the product can look good and still be useless. The post argues for humility, for method, for treating UX like a social science in parts. Not poetry. Not pure craft. Practically speaking, a button that looks beautiful but confuses people is a failure. The piece predicts a reckoning.

It’s a bit like claiming you can fix your car because you once changed a tyre. That metaphor keeps popping into my head. You might be able to change the look of the dashboard, but if the engine is on its knees, no amount of leather will save you.

Sameness: design systems and the slow disappearance of weirdness

Then there’s the piece by Itay Dreyfus, which grumbles — in a good way — about the steady drift toward sameness. Design systems and templates are great shorthand. They make teams faster. But there’s a cost: the internet starts to look like a shopping mall designed by committee. Dreyfus worries that the data-driven approach rewards what is measurable and punish what is peculiar.

The ask here is simple: cultivate taste. That’s a loaded word. But taste, as the post means it, is a personal filter. It’s the small choices that say who made the thing and why. The advice: stop letting metrics erase personality. Find a tiny oddity and guard it. I’d say this is less about throwing out systems and more about curating them. Like having a favourite mug on your office shelf while the rest of the shelves are IKEA uniform. You don’t need to rebel every day. But don’t give the mug away.

Furniture, form, and the graceful toleration of imperfection

Christopher Schwarz gives a softer, quieter take in his meditation on a Gustav Stickley plant stand. There’s a question he was asked: would you prioritise perfect joinery or beautiful lines? The post leans toward lines. He worked to fix some flaws but ultimately accepted some imperfections in service of the overall form.

This is the old argument between craft and design. Sometimes the joinery is immaculate and the thing looks fussy. Other times the joinery shows a hand, a small wobble, and the object breathes. I liked how the anecdote held both positions. It’s not a sermon for sloppy work, just a note that good design can forgive. And that’s a relief. We don’t always need museum-perfect micro-detail. Often we need something that fits the room, the routine, the life.

It’s a bit like choosing a pair of shoes. Some are beautifully made and unforgiving. Some are slightly imperfect but you want to wear them every day. Comfort counts.

Back to simplicity — a personal tech pivot that rings true for makers

In a short, candid piece Eddie Dale writes about moving from Linux to Mac to cut friction. He admires Buffett and Munger for their self-reflection and talks about a renewed love for small web apps and simple tools. There’s a lament about Norway’s design market, too — less client work, more noise.

I’d say this was one of those posts that feels like a note passed between coworkers. The main point: tools should disappear. They should let you get on with making. That matters whether you’re a solo maker or a studio. The cultural shout-out to Buffett and Munger was slightly odd, but it landed: they’re cited for the idea of leaning into what works, not chasing novelty. Which, again, is a kind of design philosophy.

Visual design matters — yes, really

Elijah Potter gets practical. After a round of bug fixes in his software Harper, users pointed to visual roughness. Fewer bugs, but it looked second-rate. The conclusion? Quality requires visual design. Potter talks about building a design system, choosing palette and type, and then iterating on the landing page.

This is the boring-but-true beat: aesthetics affect trust. You can have perfect behavior and still feel cheap. Potter’s work is a reminder that some things are surface, yes, but surfaces are part of the product. A good metaphor is packaging at a bakery. The croissant inside might be top-tier, but a grease-stained box makes people hesitate.

AI that draws diagrams and the coming change in visual work

Now, the most headline-grabbing piece: Nate on Google’s Nano Banana Pro. The model transforms documents — product requirements, spreadsheet rows, research papers — into finished visual artifacts: diagrams, dashboards, infographics. The pitch is that the model preserves relationships and produces usable visuals.

I would describe this as both thrilling and unnerving. Thrilling because it knocks down the tedious part of turning ideas into visuals. Unnerving because it also accelerates the exposure of sloppy thinking. If you can instantly generate a polished diagram, you can also instantly show a bad one to twenty stakeholders. The post comes with 30 prompts — which is the sort of practical cheat sheet teams will love.

This ties back to the unicorn argument. If a lot of people have been skating by on pretty comps, AI will either make them look better (temporarily) or reveal that they don’t know the principles behind what they’re generating. In any case, the idea that visuals will be easier to produce changes the power balance in meetings. Expect faster feedback loops, and also faster wrong turns.

Branding like software — a move toward iteration and testability

Paul Jun argues brands can’t be monoliths anymore. The metaphor he uses — brands as software — stuck with me. It’s a neat pivot: treat identity as something you ship, test, and update. Embed values in behaviour, not just in a logo.

This is the cultural speed point. Brands used to be carefully rolled out and then ossified. Jun says the world moves too fast for that. You need to test brand elements in the wild, learn from interactions, and then evolve. It’s like building a feature: ship small, measure, adjust. There’s an invitation in there to stop protecting a mythical consistency and to start tuning for relevance.

There’s also a cautionary note. If everything is mutable, you risk losing coherence. The middle path is to have core principles and flexible surface systems. Think of it like a public radio station: the mission stays the same even if the playlists change.

The cultural leftovers: Apple, subscriptions, space debris, and other curiosities

The Ephemeral Scrapbook from Numeric Citizen Space drafts a flurry of tech notes — mockups for rumored Apple Home gear, subscription service worries, photography fixes on iPhone, and a few AI oddities. It’s the kind of piece that’s equal parts link farm and diary. Worth a skim if you enjoy catching up on little sparks.

Then there's the odd but evocative short from WARREN ELLIS LTD. A delayed astronaut return because of micro-cracks in a viewport window — likely caused by space debris — sits next to sculptural speakers made from old rocket fuel tanks at Designart Tokyo. Thematically, it’s about debris and reuse. Space junk is a cold, practical problem. Turning a fuel tank into a speaker is a slightly romantic counterpoint. I would describe that pairing as an uneasy poem: danger and resourcefulness in the same breath.

Recurring themes that kept popping up

A few threads kept showing up across posts. They’re not tidy, but they’re noticeable.

• Craft vs. speed: Furniture joinery, public bridge proposals, and brand-as-software all wrestle with whether to slow down for craft or speed up for adaptability. It’s the old trade-off, again and again.

• Visuals are trust signals: Harper’s visual issues and Nano Banana Pro’s polished outputs both show how looks influence confidence. People believe things that look right. That’s not cynical — it’s human.

• Systems and sameness: Design systems help teams move faster but can flatten distinctness. The same toolkit that saves time also risks a bland internet. There’s a pushback thread saying: guard taste.

• AI as a force multiplier and magnifier: AI can accelerate good design and reveal bad design. It’s not a magic fix for lacking fundamentals.

• Brand and behaviour: A consistent brand is less about a single identity document and more about repeatable actions. Brands that act like platforms will age better.

These threads loop into each other. Systems encourage speed, speed invites AI, AI reveals gaps, gaps force a choice about craft. And on and on.

Where people agree — and where they don't

There’s surprising agreement on one narrow point: design matters. Not everyone says it with the same urgency, but posts from Elijah Potter, Itay Dreyfus, and Christopher Schwarz all come back to the idea that choices matter. The quarrel is over which choices to prioritise.

Disagreement tends to be sharper on the role of beauty versus utility. Jason Clauss is blunt: UX is a science and art alone won’t save you. Schwarz is softer: design can forgive a little roughness. Both are right in their corners. It depends on the problem you’re solving. If you’re building life-saving software, clarity outranks charm. If you’re making a chair, charm might win.

There’s also an argument over whether systems kill creativity. Some say systems are scaffolding. Some say scaffolding becomes a cage. That’s a debate as old as mixing templates with craft. The posts don’t resolve it — they sketch the trade-offs.

Little practical things that jumped out

• If you’re shipping product, invest in visual polish early. It’s not vanity; it’s trust. That’s what Elijah Potter showed with Harper.

• If your team leans on templates, pick one detail that won’t be templated. Give it personality. That’s Itay Dreyfus’s ask. Even a small flourish helps.

• The Nano Banana Pro piece shows that prompting matters. Having a list of prompts shifts the work from “make pretty” to “say what you mean clearly.” If you’re trying to onboard a team to AI visuals, start with intentions, not tools.

• For public projects like the Clapham bridge, think of the structure as both art and infrastructure. It must survive weather and be legible to people who walk past with prams and shopping bags.

• If you’re a maker feeling squeezed by market changes (I’m thinking of Eddie Dale’s note from Norway), maybe check your workflow for simplicity. Less friction often helps when demand is soft.

Tiny digressions that mattered to me — and might to you

One small tangent: the Warren Ellis piece made me picture a junkyard of satellites and old rockets, and then a neat little speaker carved out of something that once burned with purpose. It’s a human trick — making beauty from leftovers. That’ll probably become a motif in design more often, because resources get tight and stories sell.

Another stray thought about the Clapham proposals. People often forget that a bridge touches more senses than sight: sound, touch, shadow. A cladding choice might brighten the street, but if it clanks at night or collects pigeon droppings badly, people will notice. So small decisions have outsized consequences. Like picking socks for a wedding. Nobody remembers the socks if they match. They remember them if they don’t.

If you want to dig deeper

Each post has a different payoff. Some are practical, some are meditative, and some are speculative. If you like process and tool talk, Nate will be the one to read next. If you want to grit your teeth on professional ethics of UX, go to Jason Clauss. If you like the smell of wood and the slow logic of furniture, Christopher Schwarz will pin you down in the best way.

And if you want the neighborhood feel — local politics and a real-life design show — the Clapham bridge models from Ian Mansfield are exactly the kind of thing you can wander by with a coffee. That’s an invitation as much as a pointer.

There are minor posts that are worth keeping in the bookmarks too. Numeric Citizen Space gathers a lot of small tech crumbs into one place. WARREN ELLIS LTD gives a slice of weird that stays with you.

I’d say the week’s conversation is not tidy. But it’s alive. The threads — craft versus speed, systems versus taste, AI versus fundamentals — keep tangling and untangling. That’s fine. Design likes knots. It’s how you work them that shows whether the result is a bad tangle or a good sweater.