The Feed & The Thread

We’re drowning in AI tools but starving for strategy, so we ask whether we’re using these tools to ship faster or to understand deeper. Jeff Gothelf proposes a lightweight canvas to cut through the feature-stuffing chaos, while Pavel Samsonov warns that our obsession with speed is killing real learning by bypassing the discomfort of challenging existing opinions. We also explore how design styles install mental categories in users' minds and debate the tension between high-performance visual rendering and the need to preserve the human story behind our metrics. From The Feed Under Autumn’s Spell...

We keep building for the average user, but the real friction lives at the edges where standard heuristics fail the bottom third of our audience. Patrick Neeman argues that we must respect Doherty’s threshold to keep humans in the loop long enough to catch AI errors, while Kai Wong suggests looking to farmers’ markets to design tangible, inclusive experiences for those with lower literacy. Meanwhile, the community is dissecting the subtle affordances of hover states and cross-platform iconography to bridge the gap between static interfaces and responsive, native-feeling design. From The Feed Another Saturday, Another Song...

We examine whether AI speed is actually a pressure valve that traps us in constant cognitive load, as Pedro A. Brêtas argues that saved time just fills with new demands. This tension deepens when we look at the "vibe architects" creating fragile, unmaintainable systems and the atrophy of human judgment that Adam Cutler warns against. Meanwhile, we unpack the undervaluation of research rigor and the educational gaps in foundational UX skills that are leaving new designers without the evidence to back their artifacts. From The Feed The feed I produce myself (Pedro A. Brêtas) — Save...

We ask who actually holds the power to decide in design, exploring Stacie Sheldon’s argument that empathy often masks unchecked authority and Luke Wroblewski’s shift of that control from documentation to code. We contrast this structural shift with the human cost of the industry, from a professor advising students to hide their identity to a designer pausing history for user choice. It’s a look at how we navigate the tension between systemic power and individual voice in a field that is rapidly changing. From The Feed AI beyond average: from chatting to collaborating (Sen Li...

We explore the widening gap between AI capability and human intent, asking whether precision in prompting, as argued by Theresa-Marie Rhyne, or owning the feedback loop, as Christian Crumlish suggests, is the key to bridging that distance. We also examine the emerging need for AI restraint, from a student’s prototype that chooses when to stay silent to the push for standardized compliance patterns that reduce cognitive load. Ultimately, we ask if the future of UX lies not in making AI more available, but in teaching it when to disappear. From The Feed Five-step color balancing wi...

We’re shifting from capturing attention to governing it, a move Aurélie Radom argues is necessary now that AI has made machine focus cheap and human attention the true scarce resource. This governance challenge extends to our tools and processes, from Will Schiavone’s findings on AI’s nuanced but fragmented coding of user responses to the practical headache of retiring static design assets that live on in local libraries. We explore how these shifts force us to rethink everything from organizational headcount to the statistical foundations of our research, asking whether we are truly building systems or just fi...

We explore whether AI’s speed is liberating us for strategy or simply triggering a Jevons Paradox that demands more output, drawing on Daniel Ruston’s warning about the shift from maker to orchestrator and Caitlin Steele’s call for building “defensible space” against constant crises. By examining community debates on original craft versus automation and the tension between thematic aesthetics and functional readability, we question what truly matters when execution becomes nearly instant. From The Feed I Tested Opus 5.5 at Every Effort Level for Animation Design (Nick Babich) — Higher effort settings yield nuanced animations but require care...

We’re building AI systems faster than we can define who’s responsible when they fail, creating a dangerous gap where everyone feels they’re doing their job but no one owns the outcome. Jeff Gothelf argues we must name a specific accountable human before automating decisions, while Pavel Samsonov warns that we’re inverting design by seeking solutions before defining problems. This tension is mirrored in the community, where we’re seeing hollow AI-generated conference talks and designers scrambling to save the entry-level craft that AI is eroding. From The Feed LLMs are just normal technology...

We ask who is responsible when the tools we rely on to store our work disappear, noting that a quarter of web pages from the last decade have already vanished. Dora Czerna argues that this loss forces us to take ownership of our data, while Fabricio Teixeira challenges the notion that static deliverables are the final product. We also examine a designer’s controversial shortcut of skipping manual ideation to use AI-generated drafts as a starting point, and a plea for resources to bridge the empathy gap in rigid waterfall environments. From The Feed For the De...

We examine the blurring line between browsing interfaces and direct answers, asking where trust and value reside when the tool disappears. Taylor Dykes and Luke Wroblewski argue that context and outcomes matter more than static pages, while we debate whether choosing neutral typefaces like Helvetica is now a deliberate act of restraint against AI-generated noise. From The Feed Skip the Tools, Make the Outcomes (info@lukew.com) — Users prefer direct, real-time answers over browsing interfaces, effectively making the interface disappear. Is this the line for burgers? (Hiroshi Sato) — Effective design solutions often emerge from borrowing insights and...

We ask whether AI should feel like a tool or a companion, pitting Julie Zhuo’s argument for the "mom test" of approachability against Luke Wroblewski’s push for direct, just-in-time outcomes. This tension deepens as we explore how tactile design choices, from "soft screen" interfaces to living progress bars, ground digital interactions in physical reality. It’s a look at why raw capability isn’t enough, and how we must design for the human weight of our own drive to use these technologies. From The Feed What will make personal AI go big? (Julie Zhuo) — Personal A...

We argue that the real friction in design isn't the craft itself, but the business realities and power structures that dictate who pays and what gets measured. From Scott Berkun’s warning that influence matters more than style to the blind spots in quantifying human stress, we explore how hidden costs and opaque metrics create the actual bottlenecks. This is a look at why navigating these systemic constraints is the true test of professional success. From The Feed The 5 big lessons designers must learn (Scott Berkun) — Professional success depends on understanding business realities and navigating power stru...

We examine how the UX pipeline is breaking as AI absorbs the entry-level work that used to train our judgment, a shift Patrick Neeman argues threatens the profession’s long-term viability. This tension is mirrored in community discussions where juniors are penalized for using AI tools, forcing a dangerous gap between efficiency and the perceived authenticity of manual labor. We explore whether we can maintain rigorous human insight and ethical transparency while relying on tools that increasingly bypass the foundational experience needed to design for the future. From The Feed Figma vs. Claude Design: From UI Sc...

We explore the tension between designing for machines and humans, asking why we rely on generic loading states and rigid code checks when user trust actually depends on context and explanation. Adi Leviim argues that unexplained variance is the primary trust killer in AI products, while Nick Babich shows that generative AI accelerates implementation but cannot replace the nuanced judgment required for consistent design systems. In The Thread, we connect these dots to the silence around UX credentials and the loss of design rationale, revealing that without articulating the "why," we build interfaces that feel opaque and untrustworthy to...

We’re losing the language to critique AI actions, and today we ask who is actually responsible when those systems fail. Jeff Gothelf argues we must measure human impact rather than API calls, while Pavel Samsonov warns that anthropomorphizing models causes us to surrender our own critical agency. In The Thread, we see this tension play out in the community as designers struggle to inject humanity into bland AI-generated interfaces and push back against unsustainable industry expectations. From The Feed Assigning agency to AI means surrendering your own (Pavel Samsonov) — Anthropomorphizing AI models causes professionals to stop...

We examine the friction between standardized design patterns and the growing demand for distinct, premium aesthetics, asking why we often mistake AI-generated averages for intentional design decisions. From Adir SL’s analysis of the App Store button’s disappearance to the community’s struggle to escape the "AI look," we explore how true distinctiveness now comes from knowing exactly which standards to break and why. From The Feed Whatever happened to “Download on the App Store” buttons? (Adir SL) — The disappearance of mandatory App Store buttons signals a shift from uniform platform control to flexible, brand-specific design choic...

We explore the widening gap between AI’s generative speed and the semantic clarity users actually need, asking whether we’re building intelligent products or just adding noise. From Tanner Kohler’s pushback on automation as a silver bullet to Luke Wroblewski’s fix for dense chatbot text, we examine how to balance novelty with genuine usability. We also dive into the community’s debate on whether playful loading states or rapid, untested iterations truly serve the user’s intent. From The Feed Designing AI Products and Features: Study Guide (Tanner Kohler) — Adding AI features does not automatica...

We examine the dangerous industry deflection that blames users for "prompting" AI wrong, drawing on Patrick Neeman’s parallel to Ralph Nader’s automotive critiques to argue that systemic interface flaws are not user skill gaps. This tension is tested by Nick Babich’s findings that identical prompts yield wildly different results, proving that tool defaults matter more than user intent, while Julie Zhuo reframes the conversation by arguing AI’s true value lies in expanding personal creativity rather than just automating drudgery. We also dive into the specific, grounded friction of daily design, from the burden of playful physics...

We’re asking why the heavy UX processes that once protected us are now slowing us down, as Patrick Neeman argues the Double Diamond is obsolete in an era where AI has collapsed the cost of building. This tension is mirrored in our thread, where we unpack why a beginner’s humility in seeking critique matters more than their unfinished code, and how misaligned data scales can break user trust. It’s a reminder that when the barrier to entry drops, our job shifts from preventing errors to knowing which ones to keep. From The Feed Rules...

AI has removed the friction that once forced teams to define coherent structures, leading to what Patrick Neeman calls a "Winchester Mystery House" of isolated features. We examine this velocity-versus-structure trade-off alongside Lucas Plabst’s findings that participants may withhold critical insights from AI moderators, threatening the validity of our data. Finally, we confront the dangerous shortcuts in our field, from synthesizing survey responses to using high-fidelity tools that prematurely lock in exploratory ideas. From The Feed Don’t let AI build a Winchester Mystery House (Patrick Neeman) — AI removes the friction that forced teams to define...