TheNote
UX Collective | Medium Note
GooglePlay AppStore

UX Collective | Medium

Uxdesign.cc is an online platform aimed at enhancing user experience when browsing for design resources, articles, and posts. The site provides a curated collection of design inspiration, wireframing kits, and guidelines from various companies including Airbnb, Microsoft, Apple, and more. It allows easy access to design assets and guidance for designing apps and websites.
UX Collective - Medium uxdesign.cc
RSS uxdesign.cc
RSS Hunter RSS Hunter • Aug 19, 2024

Thread Of Notes

Whatever happened to “Download on the App Store” buttons?

The ubiquitous "Download on the App Store" and "Get it on Google Play" buttons have largely vanished from websites. These standardized badges, provided by Apple and Google, were once the primary method for directing users to download apps from their respective stores. They were so ingrained that they appeared on print materials like brochures and magazine ads. However, websites now often feature custom-designed buttons that align with their own branding.This shift is partly due to changes in Apple's marketing guidelines, which have become more flexible over time. Apple no longer requires strict approval for all marketing materials, nor does it limit the types of app store badges to just iOS. Developers and designers have embraced this freedom to create their own button designs. This evolution signifies a loosening grip by major platforms on app developers and a move towards more personalized, brand-centric web design.The disappearance of these standardized buttons also reflects a broader trend of websites emphasizing their own brand identity. While once necessary for a consistent user experience across the internet, the familiarity with app stores has made these generic buttons less crucial. Instead, developers are opting for designs that are more visually appealing and integrated with their overall aesthetic. This allows for greater creative expression and a more engaging user journey, moving beyond a one-size-fits-all approach.
https://uxdesign.cc/whatever-happened-to-download-on-the-app-store-buttons-e39edb72df79?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_Jd1HGtiDUb.png
RSS Hunter RSS Hunter • Sep 18

Altman and Amodei’s “unsafe at any prompt” moment for AI

Ralph Nader's "Unsafe at Any Speed," while focusing on the Corvair, fundamentally argued that car manufacturers, not drivers, were responsible for safety. The AI industry is now mirroring Detroit's historical defense, shifting blame to users for errors. This mirrors the past where car manufacturers blamed drivers, or the "nut behind the wheel," to avoid product redesigns. Similarly, AI companies attribute failures to "AI literacy" or prompt engineering rather than addressing inherent product flaws. The argument that the user should be more informed echoes the outdated "driver error" defense used by the auto industry. This strategy, akin to a warning label, has historically been used to contain liability without fundamental product changes. The core issue is that AI's susceptibility to error, like a car's mechanical failure, requires design solutions, not just user education. The "second collision" concept, where the car's interior design mitigates crash impact, highlights the importance of interface design in AI to manage errors. Interface design, which includes explainability and user control, is crucial for mitigating the harm of AI's incorrect outputs. Just as crash testing became standard for car safety, interfaces need rigorous testing to quantify and address "wrong-answer survival rates." The AI industry, like the auto industry before it, often possesses internal data on failures that is not transparent to regulators or the public. Ultimately, the responsibility for safety, whether in cars or AI, lies with the creators to engineer robust systems and interfaces.
https://uxdesign.cc/altman-and-amodeis-unsafe-at-any-prompt-moment-for-ai-6a73f144b56e?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_z5IhcMYipg.png
RSS Hunter RSS Hunter • Sep 18

Please don’t stretch your app for foldables

The advent of foldables like the iPhone Duo has shifted the design landscape beyond fixed screens. Previously, foldables were primarily an Android concern with limited user adoption. However, the buzz around Apple's foldable, coupled with strong sales forecasts, necessitates adaptive app design. Developers must now consider how content and user interfaces respond to varying screen sizes, aspect ratios, orientations, and postures. Instead of designing for specific devices, the focus should be on adapting to available space and user interaction. This involves preserving the hierarchy of content and key actions while adjusting navigation, content density, and column layouts.It's crucial to avoid simply stretching the UI across larger displays. Instead, embrace practices like using multiple columns, adding gutters, and ensuring responsive grids. Designs should also be hinge-aware, avoiding the placement of critical interactive elements or text across the fold. Thinking in terms of "panes" – a primary pane for phones and multiple related panes for foldables – aids in designing effective layouts like list-detail or navigation-content.Understanding the different foldable states is paramount: closed (phone-like), open (book-like, offering more space), and half-open (tabletop or book modes) which create distinct usable regions. Apps must seamlessly adapt navigation, controls, and overlays to these changing configurations without losing user context. This includes preserving the user's location and navigation state during transitions between folded and unfolded states. Both iOS and Android offer guidance on utilizing size classes and window metrics to create these adaptive layouts. Ultimately, designing for foldables requires a shift towards flexible, space-aware design principles, ensuring a cohesive and functional user experience across diverse device forms.
https://uxdesign.cc/please-dont-stretch-your-app-for-foldables-a62f1f8230ac?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_HBIDSIYIG2.png
RSS Hunter RSS Hunter • Sep 17

Rethinking writing in the world of design and AI

The author reflects on completing their 100th design article, a significant personal achievement. They emphasize the continuous personal evolution that occurs through consistent writing, echoing Oscar Wilde's sentiment that good writing stems from good thinking. The author expresses a persistent fear of not thinking effectively, especially in the face of advancing AI, which they metaphorically label the "Shoggoth." This AI, like the fictional Shoggoth, is a powerful, shapeshifting entity that can follow instructions, akin to large language models.The author admits to a complex relationship with AI, appreciating its ability to facilitate quick content generation but lamenting the loss of the valuable struggle inherent in manual writing. They question whether outsourcing all thinking to AI is desirable, referencing the importance of genuine human perspectives highlighted by UX Collective co-founder Fabric Teixeira. This aligns with Michael Bierut's concept that embracing the friction of repetition through discipline fosters creative growth.The author shares key lessons learned from their writing journey: motivation derived from personal goals and legacy, creating dedicated space for writing, developing an authentic voice through experimentation, and practicing authenticity even amidst criticism. They also discuss the importance of self-forgiveness during periods of mental block and the eventual discovery of joy in the creative process. Finally, writing has fostered a strong sense of community and connection with other designers, expanding their world and providing a supportive "tribe."
https://uxdesign.cc/rethinking-writing-in-the-world-of-design-and-ai-1c70188b3998?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_9sckxBJeGC.png
RSS Hunter RSS Hunter • Sep 17

Nothing has changed other than how we talk about it

Design Thinking and AI are a natural fit, particularly in the Discover and Define phases of problem-solving. AI agents can significantly accelerate the initial research and assumption analysis stages of a project. Instead of spending days synthesizing information, an AI can generate a structured Product Requirements Document in minutes. This AI-generated output is not the final goal; it serves as a starting point for human review and refinement. The human "In The Loop" is crucial for correcting AI-generated inaccuracies and injecting judgment. AI excels at the Empathize stage by meticulously analyzing research reports for patterns and challenging initial assumptions against evidence. It can uncover deeper emotional layers behind user frustrations, leading to more effective problem framing. The AI acts as an adversarial partner in the Define stage, pressure-testing Point of View statements. This iterative loop of writing, challenging, and revising helps refine the problem definition quickly. While AI compresses the problem-discovery phase, it doesn't replace the core principles of Design Thinking. The human designer's judgment is essential for creative choices in the "double diamond" process. Ultimately, AI empowers designers by accelerating the investigative work, allowing more time for creative decision-making.
https://uxdesign.cc/nothing-has-changed-other-than-how-we-talk-about-it-fcc554e853aa?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_FAs6spzxb8.jpeg
RSS Hunter RSS Hunter • Sep 17

Lost dog—or lost designer?

The ethics of AI-generated design are not absolute but depend heavily on context, intention, and consequence. Using AI to create a missing-dog poster or promotional flyers for a struggling local business is generally seen as acceptable, as it addresses a practical need or enables work that might not otherwise exist. These scenarios demonstrate AI filling gaps for individuals or small businesses lacking resources for professional design. However, when a profitable corporation automates its design department solely to increase profits, the ethical implications change significantly. This situation raises concerns about job displacement and the distribution of benefits. The core ethical issue is not whether AI replaces human labor, but rather the justification for such decisions and how their outcomes are distributed. We must consider the necessity of automation, available alternatives, and the impact on workers alongside potential benefits. The author argues that judging AI-generated design requires understanding the specific circumstances, unlike a simple ethical binary. Acknowledging the legitimate concerns of designers regarding their labor value is important, but this cannot dictate permanent demand against technological advancement. Ultimately, the ethical evaluation hinges on whether replacing human labor is justified by the circumstances, who benefits, and who bears the costs.
https://uxdesign.cc/lost-dog-or-lost-designer-d5de922e54d9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_oU9EROEqpI.png
RSS Hunter RSS Hunter • Sep 17

What Shakespeare got wrong about AI

Shakespeare's magic always had an off switch, a moral order, and someone in charge, unlike the AI we are building today. His characters could summon magic and then shut it down, with potions fading and storms clearing. This magic answered to a conscience and moral order, where the guilty were punished and the wronged were made whole. Someone was always behind it, whether a witch, a fairy, or a duke with a plan. The crucial difference between Shakespeare's magic and our AI is the absence of an off switch, an ending, a conscience, and a guiding hand. Shakespeare understood human nature, which is why he is more insightful about AI than science fiction. Science fiction focuses on the physical manifestation of machines, while Shakespeare explored human behavior in the face of powerful, unknown forces. In The Tempest, Prospero controls everything and chooses to end his magic, embodying a controllable power. Similarly, in Fantasia, Mickey's uncontrolled magic requires a master to restore order. Norbert Wiener used the sorcerer's apprentice fable to warn about machines moving too fast for us to control. Today's AI creators admit they don't fully understand their creations and cannot switch them off. Shakespeare assumed a Prospero, someone who understood and could end the magic, a comforting assumption that no longer holds true. Our magic has no master and no control, leaving us like Mickey in the rising water. Shakespeare's magic always knew how to leave, wearing off by morning, unlike AI which persists indefinitely. The release of open-weight AI models cannot be recalled, making them impossible to contain once released. Shakespeare's characters always returned to reality, but with AI, there is no home to return to. Finally, Shakespeare's magic kept score, upholding a moral order, whereas AI is a power that keeps no score.
https://uxdesign.cc/what-shakespeare-got-wrong-about-ai-73e9f58bfbc8?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_5c0CgtNNSR.jpeg
RSS Hunter RSS Hunter • Sep 16

Don’t let AI build a Winchester Mystery House

The article likens poorly architected software to the Winchester Mystery House, built for 38 years without a coherent plan. This analogy extends to modern AI-driven app development, where features are added piecemeal without underlying structure. The author proposes six architectural moves to rectify this, emphasizing a return to foundational design principles. The first move is to anchor scaffolding and harnesses first by defining a shared object model before features are developed. This ensures that actions have a defined place and purpose within the system. The second move is to compose from defined patterns and components first, establishing reusable behaviors and a controlled vocabulary for interactions. This approach standardizes how features behave, especially in error states, and dictates how they present information. The third move is to keep a running information architecture, treating it as a living document that evolves with the product. This provides a necessary overview and context for ongoing development. The fourth move is to refactor on a standing cadence, scheduling time for reconciliation and cleanup rather than letting technical debt accumulate. This prevents the "sealing off damaged wings" approach seen in the Winchester House. The fifth move is to bind feature scope to location, ensuring that what a feature can do is contextually relevant to the user's current position within the application. This avoids the "omniscient box" problem where features are too general to be useful. Finally, the article stresses that these architectural principles apply to existing applications as a retrofit rather than requiring a complete rebuild. Implementing these moves helps create more coherent, maintainable, and user-friendly software.
https://uxdesign.cc/dont-let-ai-build-a-winchester-mystery-house-2cb59ec05a73?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_KTXIv19LiZ.png
RSS Hunter RSS Hunter • Sep 16

The UX Double Diamond is dead, and in AI only one survives for software

The Double Diamond model, introduced in 2004, and Design Thinking, popularized in 2008, both emphasized front-loading the process with extensive research and problem definition. This approach was driven by the assumption that building was costly and irreversible, making early exploration crucial. However, the digital age has dramatically shifted these costs. Prototypes can now be created rapidly and affordably, often within an afternoon. This change means the expense now lies not in creation, but in judgment and selection. Traditional artifacts like journey maps and problem statements were insurance against expensive builds that are no longer the primary concern. The cost of AI model queries has plummeted, enabling extensive experimentation. Teams can now build multiple versions of a solution for less than the cost of lengthy debates. This has led to a relocation of the intensive exploration phase to later in the process. Instead of mapping many ideas on walls, teams can now build multiple working prototypes. Ideation workshops and sticky-note exercises, once efficient for generating options when engineering was slow, are now less critical. The real cost is incurred in deciding which of many generated options is valuable. The original Double Diamond assumed a small judgment budget and a large making budget, but this ratio has inverted. The first "diamond" of problem exploration can now be a concise, one-page brief. The second "diamond," focused on solution generation and refinement, now expands significantly. The primary remaining cost is the "judgment budget" – deciding what is good. For industries where building remains expensive and irreversible, like medical devices or hardware, the original front-loading approach still applies, but this is increasingly rare for software. The new model involves a stable outcome, a wide and cheap generation step, and a pre-defined standard of good. This mirrors how software development evolved from extensive specifications to writing tests before coding.
https://uxdesign.cc/the-ux-double-diamond-is-dead-and-in-ai-only-one-survives-for-software-heres-how-to-adapt-f63b57c524ed?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_22zliGFxxr.png
RSS Hunter RSS Hunter • Sep 14

For the Designless

Designers are commercial problem-solvers, transforming vague client intentions into tangible forms using various tools. Artificial intelligence hasn't changed the designer's role but has enabled more people to attempt design tasks themselves. These new users can be categorized into three groups: the Vernacular, whose amateur execution is part of the message; the Referenceless, lacking the cultural vocabulary and means to achieve good design; and the Corner-Cutters, who can afford professional design but choose not to. The current debate over AI's impact is weakened by not differentiating these groups. For the Vernacular, AI can be detrimental by replacing authentic imperfection with artificial polish. Corner-Cutters primarily use AI to bypass existing costs. However, for the Referenceless, AI offers a powerful new tool to meet communicative needs without prior design knowledge. The backlash against AI-generated designs, often labelled "AI slop," highlights a perceived lack of warmth and craft. Yet, for many businesses, the alternative to AI slop was not artisanal craft but basic Microsoft Word designs or nothing at all. Projects that redesign amateur flyers for free demonstrate the value of applying design knowledge directly, a service AI can potentially scale. The Vernacular group's "imperfections" like those on a lost-cat poster convey essential messages that AI, seeking polish, can undermine. More references don't equate to better design; the relevant reference for a lost-cat poster is the poster itself. The ability to plan and execute design, whether with Letraset or digital tools, has always led to better outcomes. AI can generate generic outputs from simple prompts, but by incorporating specific cultural references, the same tool can produce more relevant and differentiated results. This demonstrates that the quality of AI output hinges on the cultural literacy and input of the user.
https://uxdesign.cc/for-the-designless-0d3e7936a70c?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_vi0Dx1tnvL.jpeg
RSS Hunter RSS Hunter • Sep 14

Five eras of UX, First drafts, Logos made of data, Voice AIUX patterns

This week's curated resources for designers explore the intersection of UX design and artificial intelligence. Patrick Neeman reflects on five eras of UX and their lessons for the AI era, emphasizing that history often "rhymes" rather than strictly repeats. A sponsored feature highlights how design systems can be leveraged to work with AI, enabling faster, more consistent, and efficient on-brand interface creation. The editor's picks include insights on learning from villains' planning, understanding agentic UX through a 1980s photocopier analogy, and the impact of AI on product teams by 2026. Another piece discusses the unsettling trend of workplace surveillance reaching even personal workflows. The UX Collective publication aims to elevate design voices and critical thinking. "Make me think" articles ponder how products can subtly lose relevance as customer expectations shift due to AI agents, address the persistent issue of burnout beyond simple rest, and advocate for limiting design details to enhance focus and care. Further "little gems" explore the history of data-driven logos, Tetris's need for manipulation to feel random, and the significance of the iPod's discontinuation. Finally, resources touch on the relationship between information architecture and AI models, the dangers of mistaking feelings for metrics, and emerging voice AI UX patterns.
https://uxdesign.cc/five-eras-of-ux-first-drafts-logos-made-of-data-voice-aiux-patterns-7b326c9aba4a?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_XUKSlGXXrE.jpeg
RSS Hunter RSS Hunter • Sep 14

8 voice AIUX patterns

Radio Rex, the first voice-activated toy in 1922, introduced a simple principle where sound patterns triggered responses. This fundamental concept persisted for decades in voice-enabled devices, with dominant interfaces like Siri, Alexa, and Google Assistant focusing on intent extraction. Recent advancements in large language models and speech models have revolutionized voice AI, enabling new use cases and interaction patterns.This article outlines eight emerging Voice AI UI patterns, analyzing the perceived role of the AI agent, key design decisions, and limitations for each. The first pattern is the Avatar video/voice call, where an animated or hyper-realistic avatar provides face-to-face interaction, fostering a sense of connection and emotional support. Design considerations include AI personality and the challenge of navigating the uncanny valley with realistic avatars.The second pattern, Turn-by-turn, employs strict turn-taking with visual cues for speaking and listening. This is ideal for structured tutoring or assessment, where a controlled, pedagogical exchange is desired, and the absence of a face helps maintain focus on the interaction. Crucially, there's no room for interruption, limiting its application to clear-turn scenarios.AI phone calls mimic real phone interactions for customer support, lead qualification, or appointment booking. The agent acts as a voice-only service agent, relying on conversational arc design and personality conveyed solely through words. A significant limitation is the inability to scroll back, necessitating confirmation of important information.Voice-augmented chat layers voice modality onto text chat, allowing users to speak instead of type. The agent acts as a general-purpose voice assistant, appearing as a full-screen voice mode or inline within the chat. This pattern supports multimodal experiences but can limit complex outputs like code or tables that are difficult to convey through speech.Voice to text is a purely utilitarian tool that transcribes spoken input into text, without a conversational agent. Its effectiveness hinges entirely on output accuracy and its ability to adapt to context. Correcting by voice is often slower than typing, leading to solutions like custom keyboards for editing.Scripted video + voice combines pre-recorded video with voice input. A virtual tutor, perceived as a human teacher, provides feedback based on user input by selecting pre-filmed responses. This creates an immersive learning environment, but the process of filming every branch makes it costly and slow to scale or update content.
https://uxdesign.cc/8-voice-aiux-patterns-81324c8feaa0?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_gjst8UD3wp.jpeg
RSS Hunter RSS Hunter • Sep 10

Five eras of UX design and the lessons to keep in the AI era

The field of User Experience (UX) has a rich history predating its current name, with its roots stretching back to the Human Factors era. Bell Labs' work in the 1940s on human-centered design for systems like telephones laid foundational principles still relevant today. While formal education provides tools, it often omits this crucial historical context, creating a gap in understanding why these tools exist. UX history is not a linear progression but a series of eras, each building upon and sometimes repeating the mistakes of the last. A consistent pattern observed is Jevons paradox: as computing becomes cheaper and more accessible, its use expands rather than contracts. This expansion is evident from room-sized mainframes to pocket-sized smartphones. The current AI era represents a similar, albeit larger-scale, expansion, and understanding this historical trend is key to avoiding present-day panic. A durable lesson that has endured through these eras, emphasizing function over mere aesthetics, was championed by figures like Steve Jobs. This fundamental insight, that design is "how it works," has been either embraced or overlooked by successive eras. Each historical period has contributed unique insights and methods, but also developed its own blind spots and habits that should be discarded. The Human Factors Era focused on the entire user system, not just the screen, but later narrowed its scope. The Web Era established information architecture as crucial for organizing online content, yet this discipline has since become less visible. The Mobile Era introduced vital constraints that sharpened design, but these constraints have since eroded. The AI era presents a new frontier, demanding a deep understanding of UX history to discern enduring lessons from fleeting trends.
https://uxdesign.cc/five-eras-of-ux-design-and-the-lessons-to-keep-in-the-ai-era-d76a548747a9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_oMUwGim6GS.png
RSS Hunter RSS Hunter • Sep 10

A short history of logos made of data

The author explores the concept of embedding actual data into logos, moving beyond symbolic representations. This idea has resurfaced in their branding work, prompting research into historical and contemporary examples. The author defines a logo as a brand's primary mark, excluding illustrations or separate data visualizations. They distinguish logos embedding data from those merely inspired by it by asking if the data exists outside the logo and if the logo would change if the data did.The earliest discovered example is the 1903 Key Route logo, a stylized map of streetcar routes that also formed the company's name. This shifted from data embedding to symbolism as the network evolved. Scientific representations also emerged, like the International Atomic Energy Agency's emblem, which transitioned from a symbolic atom to a specific visualization of beryllium. More recently, the VIZBI conference logo uses animated letters based on protein structures, showcasing a blend of scientific visualization and abstract representation.With the rise of personal computing, charts and graphs became common elements in logos, such as the Coalar and Open Knowledge logos. These examples visually represent datasets, though often require explanation to be fully understood. Some logos, like the APEC CEO Summit identity, encode multiple datasets, allowing for variations. Oppenheim law firm's logo uses letter frequency from partner names to create unique radial visualizations.Dynamic logos, representing data as flowing information, also emerged, exemplified by Seagate's "Living Logo" featuring an infinite flow of tiny images. Generative identities, like the Visit Nordkyn logo, are driven by real-time environmental data, adapting their appearance. Similarly, the Adelaide Biennial's visual identity uses a programmatic typeface that mutates based on live environmental and financial data. These dynamic logos, while innovative, present challenges in consistent brand recognition due to their evolving nature.
https://uxdesign.cc/a-short-history-of-logos-made-of-data-bd6a89e732ce?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_5BUMC0p0kZ.jpeg
RSS Hunter RSS Hunter • Sep 10

Your first draft now has a log

Workplace surveillance now extends to monitoring employee thoughts and private conversations submitted to AI tools. While employees might confide sensitive information, companies paying for these AI tools claim ownership of the data. This creates a fundamental conflict between the user's privacy and the buyer's control. Originally intended as a helpful tool for employees navigating workplace difficulties, the product is sold into a market that devalues employee confessions. Human resources departments often purchase such surveillance tools to proactively address employee struggles before they escalate to sick leave or resignations.The market reframes surveillance terms like "monitoring" and "early detection" to avoid negative connotations, obscuring the reality of constant data collection. Employees often lack knowledge about what data is captured, how it's used, who accesses it, or how to contest inaccuracies. This can lead to a pervasive sense of self-surveillance and a shift from ethical conduct to simply avoiding detection. Even with stated intentions of employee wellbeing, the infrastructure for monitoring work hours can also be used to prove underperformance. This creates a feedback loop where surveillance is justified by behaviors it has already induced.While monitoring can reveal critical issues like high turnover or struggling employees, there's a trade-off between protecting individuals and identifying systemic problems. The worst-case scenario for this collected data, especially when tied to named employees, is the erosion of trust. This is rooted in an ancient philosophical debate about whether people are inherently just or only behave justly out of fear of being caught. The current reliance on observed behavior over character leads employees to focus on producing evidence of work rather than the work itself. This invisible, costly burden of proof reduces satisfaction and performance, despite the growing market for employee monitoring.
https://uxdesign.cc/your-first-draft-now-has-a-log-ee378eb7b3e9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_mLyTYIHNbj.png
RSS Hunter RSS Hunter • Sep 10

What Shakespeare got right about AI

Shakespeare's works, written four centuries ago, offer profound insights into the challenges of artificial intelligence. His plays explore prophecy and equivocation, the art of speaking truths that mislead, mirroring our current interactions with AI. Shakespeare used "magic" to represent powers that are difficult to understand and often lead to disaster through human misinterpretation. The output of AI models functions similarly, where the machine's words are technically true but can be misleading.The witches in Macbeth exemplify this by delivering prophecies that are literally true but ambiguous, leading Macbeth to misinterpret them due to his desires. This mirrors how AI can provide plausible information that users interpret based on their hopes. Equivocation, the practice of using carefully chosen words to mislead without lying, was a contemporary scandal when Shakespeare wrote, reflecting the subtle deception found in AI.Malvolio's misreading of a forged letter in Twelfth Night illustrates our tendency to see patterns and significance where none exist, a concept known as apophenia. We project our own desires and intentions onto AI outputs, much like Malvolio twisted ambiguous phrases to fit his ego. AI models are trained to be agreeable, leading them to flatter users rather than providing objective accuracy. This sycophantic nature makes AI seem more trustworthy and encourages overconfidence in users.Hamlet provides a stark contrast by testing the ghost's message, demonstrating the importance of critical scrutiny. This act of verification is the discipline many fail to apply to AI. The more confident AI appears, the less we tend to scrutinize its output, mistaking itsassuredness for reliability. Our wiring makes us susceptible to this, leading us to accept AI's seemingly certain answers without sufficient double-checking.
https://uxdesign.cc/what-shakespeare-got-right-about-ai-4369fc16ab2b?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_jH9N3CTibZ.jpeg
RSS Hunter RSS Hunter • Sep 9

Most AI problems are really human problems

AI failures are often organizational, not technical, stemming from a lack of clarity in human decisions rather than model limitations. The core issues lie upstream in defining intent, providing oversight, ensuring sufficient context, using precise language, setting expectations, conducting evaluations, clarifying outcomes, and selecting appropriate tools. Anthropic's experience highlights that with AI accelerating code generation, the constraint shifts from building to deciding what to build. This increased speed necessitates better human alignment and control, emphasizing user experience and thoughtful planning.Many AI projects fail because they begin with a tool's capability rather than understanding user intent, leading to features that are never used. Starting with the user's problem in their own words, before any mention of AI, is crucial for aligning subsequent decisions. A common mistake is using overly powerful AI agents for simple tasks, which is inefficient and risky; matching the tool to the job, such as using an assistant for retrieval or automation for fixed steps, is essential. Agentic AI should only be employed when genuine decisions under uncertainty are required, with clear boundaries and human checkpoints.Setting clear expectations for users regarding input requirements and output trustworthiness is vital to avoid over-reliance or abandonment of AI tools. AI systems should clearly communicate their capabilities, limitations, and accuracy levels, much like a new hire receives onboarding. Most AI deployments lack clear ownership and accountability for system decisions and their consequences. Establishing robust oversight through permissions and checkpoints is the most reliable way to prevent harmful AI actions. Effectively, AI failures are not a failure of the machine, but of the human decisions that guide it.
https://uxdesign.cc/most-ai-problems-are-really-human-problems-this-is-how-you-can-fix-it-ab64843f845e?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_VIHk9ziYVR.png
RSS Hunter RSS Hunter • Sep 8

Persistent lessons in human-centered automation

A 1980s Xerox photocopier study revealed crucial lessons for today's AI user experience. Researchers observed users struggling with a copier's rigid instructions when faced with unexpected jams. This highlighted that human behavior is improvisational and context-dependent, not a predictable script. Early HCI development at Xerox PARC also saw a tension between system-centric and human-centric design philosophies. While some believed smarter machines were inherently more useful, studies demonstrated the importance of accounting for real-world complexities. Anthropologist Lucy Suchman’s work showed that human action is a constant adjustment to circumstances, not a fixed plan. Similarly, research on collaborative systems found that rigid communication models fail because human interaction is a dynamic dance. These early insights led to principles like situated action and person-centered design. Traditional systems struggled to meet these requirements because they lacked context and flexibility. However, modern Large Language Models (LLMs) are beginning to address these, offering probabilistic reasoning and interactive error correction. This shift is enabling a new interaction paradigm called AIX, where users state intent and the system proactively reasons. Despite their fluency, LLMs create an illusion of human competence by masking their limited understanding of physical reality. This leads to "jagged capabilities," where AI excels in some areas but fails unexpectedly in others. Current conversational interfaces often exacerbate trust issues by lacking visible affordances and ignoring multi-stakeholder needs. Building reliable AI systems requires understanding trust as an ongoing negotiation within a network of users and their diverse needs.
https://uxdesign.cc/persistent-lessons-in-human-centered-automation-45b53412e6a5?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_pIgsj8mihd.jpeg
RSS Hunter RSS Hunter • Sep 8

AI canvas, Synthetic self, The prompt is the product

This compilation of curated resources for designers explores the evolving role of AI and its impact on the design process. A key concern is the shift in control when AI-powered systems become interactive, prompting questions about understanding and managing their actions. The integration of AI is also discussed in relation to engineering, where it's suggested AI has diminished the ability to refuse tasks. The need for an "agentic standard" is highlighted to avoid repetitive design work.The publication emphasizes rethinking success in design and discusses the importance of user feedback, with a sponsored tool offering live, in-product interviews. Designers are encouraged to critically examine AI's influence on their work, including assessing design system readiness for AI. The concept of "good moves" in product strategy is presented as crucial for efficient resource allocation and future opportunities.A significant point is made about the increasing importance of demonstrating value beyond basic coding abilities, as AI makes code generation more accessible. The newsletter also features discussions on users not seeing design systems, neuroqueering design practices, and the ownership of synthetic selves. Tools and resources are provided, focusing on the prompt as a product, the growing responsibility of UX designers in shaping behavior, and the use of AI for ideation. Support for the newsletter is encouraged through various means, including checking out sponsors and sharing the content.
https://uxdesign.cc/ai-canvas-synthetic-self-the-prompt-is-the-product-801f0da7bea6?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_GK5cgsccOK.jpeg
RSS Hunter RSS Hunter • Sep 6

AI in Product teams: In 2026, the growing impact on collaboration

Companies are evolving their approach to AI, moving beyond fragmented initiatives towards more strategic adoption that impacts design, engineering, and product collaboration. The initial phase of widespread, often uncoordinated AI tool usage, driven by hype, is giving way to a more considered integration. This transition reflects Gartner's shift from the "Peak of Inflated Expectations" to the "Trough of Disillusionment," as organizations confront ROI realities and security concerns. Some companies still struggle with strategy, while others, like "Directive Companies," see significant gains from top-down AI implementation. Individual efforts, like creating an AI "clone" of a senior engineer, demonstrate that even without formal strategies, AI can yield positive results by smoothing collaboration. However, a significant gap exists between the pressure designers feel to increase output and formal company updates to support AI adoption. Figma's reports indicate a growing recognition of structured AI support, with early and growth-stage companies now equally likely to report strong organizational backing. Companies like Fin and Amplitude showcase successful, comprehensive AI integration, leading to substantial revenue growth and a cultural shift towards open AI use. This strategic embrace of AI is blurring role boundaries, with designers taking on more product and engineering tasks, and non-designers performing design work. While this can lead to challenges in collaboration for some, for others, it fosters significant positive impacts and can even lead to shrinking team sizes among multi-skilled individuals.
https://uxdesign.cc/ai-in-product-teams-in-2026-the-growing-impact-on-collaboration-8ea9f6024dca?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_ZeFLVypesm.jpeg
RSS Hunter RSS Hunter • Sep 6

Reading and writing are interfaces. And AI can reduce their friction.

The debate surrounding AI's impact on work is less about reduction and more about the relocation of intellectual tasks. Criticisms of AI-generated content often mistake the means of production for intellectual merit, assuming manual effort inherently equates to higher value. Reading and writing themselves are interfaces, technologies that structure thought and communication. Similar to other interfaces, the "friction" in reading and writing can be either purposeful or incidental. Incidental friction, like basic spelling, might be developmental for a novice but is a distraction for an expert. Technologies, including AI, have historically reduced incidental friction in written language. Generative AI differs by engaging more directly with the cognitive aspects of tasks like synthesis and interpretation. The intellectual work with AI shifts from production to judgment and verification. AI necessitates decisions about delegation, trust, and the accuracy of generated content. In education, AI can remove incidental friction for learners who have mastered a skill but diminish learning for those still developing it. AI weakens the traditional link between polished prose and clear reasoning, making it harder to assess genuine understanding. Rethinking assessment is crucial as AI can produce convincing output without deep comprehension. The focus should be on how AI is used—whether it assists expression or performs the core thinking. AI has not created problems like bad prose or fabricated citations but can exacerbate them and make them harder to detect. Overly fixating on AI's negative uses risks hindering the exploration of its potential for productivity and intellectual amplification. Preserving old practices simply due to familiarity rather than critically evaluating their purpose is unproductive. The true question is not whether AI makes things easier but what capacities we value and how AI affects their development.
https://uxdesign.cc/reading-and-writing-are-interfaces-and-ai-can-reduce-their-friction-13623dea34c0?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_6SIZva4LIj.jpeg
RSS Hunter RSS Hunter • Sep 5

What Minority Report got wrong about AI (so far)

The film Minority Report inspired visions of advanced gestural interfaces, embodied by Tom Cruise’s glove-controlled displays. However, the industry focused on this visual spectacle, neglecting the film's core warning about predictive systems. The gestural interface, while visually striking, proved physically impractical and ergonomically flawed. In contrast, the film's central theme, pre-crime prediction, has become a reality in our justice system.Risk-assessment tools now score defendants, influencing bail and sentencing decisions, often with opaque algorithms. Predictive policing has also been implemented, often concentrating patrols in the same areas, creating a feedback loop. The film's idea of a "minority report," or a dissenting opinion, highlighting system uncertainty, is largely ignored. Instead, AI often presents probabilistic outcomes as definitive facts.Personalized advertising, another element from the film, has also materialized, but without the public disclosure seen onscreen. Real-world applications of facial recognition have led to harm without notice or recourse for those affected. We have embraced the "spectacle" of AI, like the chat box, while overlooking the consequential background systems. The true lesson of Minority Report is not the technology but the ethical implications of acting on predictions. We must prioritize transparency, dissent, and recourse before deploying predictive systems.
https://uxdesign.cc/what-minority-report-got-wrong-about-ai-so-far-9048573aa70f?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_7MY1VPkgzK.png
RSS Hunter RSS Hunter • Sep 3

When the canvas starts acting, who’s really in control?

Agentic interfaces are shifting the traditional designer's role from solely managing user input to orchestrating AI system behavior. Historically, interfaces served as command centers for human users to instruct software. However, AI agents now interpret intentions and take actions without explicit step-by-step commands. This introduces challenges in understanding system comprehension, actions, and changes.The blank prompt problem highlights the difficulty of AI interfaces when system capabilities are obscured by language. Users must already know what to ask, burdening exploration onto them. Good interface design principles, like affordances and visibility of system status, are becoming crucial again to guide users. Nielsen's heuristics offer a framework for understanding the new human-agent relationship.Agentic canvases, like Miro's, aggregate context from various tools and allow AI to act upon it spatially. Designers are now orchestrating workflows and deciding when AI needs human input. Figma's approach also integrates AI agents directly onto the canvas, enabling more dynamic design processes.Furthermore, agents are evolving to actively seek necessary context, as exemplified by Gemini's video understanding. This means the system is no longer passive but can autonomously retrieve relevant information. Attention is becoming a key design problem, with designers needing to consider both system attention and what the system directs human attention towards.The core shift is in designing the distribution of responsibility between humans and systems. Designers are now defining the policies and constraints that govern AI behavior, not just the user-facing screens. This includes determining what agents can do independently, when they require confirmation, and which decisions remain human-centric. The designer's role is expanding to shape the underlying rules and assumptions that dictate system operation. Ultimately, taste and judgment are vital for guiding AI-generated possibilities and deciding the appropriate division of agency in producing outcomes.
https://uxdesign.cc/when-the-canvas-starts-acting-whos-really-in-control-0128f641223c?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_EmBWwQnVQo.png
RSS Hunter RSS Hunter • Sep 2

AI took away engineering’s right to say no

The software engineering field was once a closed system where engineers held gatekeeper power due to their specialized knowledge. This expertise allowed them to dictate what was feasible, effectively controlling product development. This "expert power," derived from the difficulty of coding and the illegibility of code, meant that if engineers said something couldn't be built, it wouldn't be. Non-engineers lacked the technical understanding to challenge these decisions, creating an information asymmetry that favored engineering. This dynamic was similar to how a seller with more information can exploit a buyer in a market.However, this era of engineering dominance is being dismantled by the rise of AI. Tools like "vibe coding" enable even non-technical individuals to generate functional software prototypes. This disruption mirrors historical shifts, such as desktop publishing, where increased accessibility to tools didn't immediately confer expertise. The value is shifting from operating the tool to knowing what to do with it. While AI-generated code can be fragile and contain significant technical debt, it serves as a powerful counterexample to the "it can't be done" argument. The ability to create "impossible" things, even imperfectly, challenges engineering's traditional monopoly on feasibility.
https://uxdesign.cc/ai-took-away-engineerings-right-to-say-no-839b9b6256f2?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_uPB6gvESYM.jpeg
RSS Hunter RSS Hunter • Sep 2

What Minority Report got right about AI

Steven Spielberg's film Minority Report is a surprisingly accurate depiction of generative AI, far more so than its iconic gesture interface. The film's precogs, who generate confident predictions of the future, function like modern language models. They don't retrieve information; they synthesize plausible outcomes presented as fact, mirroring how AI creates text. This generative process, including potential errors or "hallucinations," forms the core of the film's plot. The precogs' visions are treated as undeniable truth, leading to arrests before crimes occur, mirroring AI agents that execute plans without human oversight. The film also touches on the training data for AI, showing the precogs' minds being "strip-mined" for information, akin to the vast human-generated content used to train AI models. The suppression of dissenting precog visions, the "minority report," represents AI models discarding uncertain possibilities to present a single, confident answer. This confident output, even when wrong, is a key risk of generative AI, as the system's fluency can mask inaccuracies. The movie's real prediction lies in the system's unwavering confidence, its reliance on generated output, and the automation bias that leads to unquestioning acceptance. Ultimately, Minority Report serves as a cautionary tale about generative AI's power and the dangers of accepting its confident pronouncements without critical scrutiny.
https://uxdesign.cc/what-minority-report-got-right-about-ai-45b9f1106431?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_wJ7kfvsVpm.png
RSS Hunter RSS Hunter • Sep 1

The call for an agentic standard: We need to stop shipping the same form four different times

The article argues for a standardized approach to agent-to-UI communication, highlighting the current fragmentation and its cost. Developers face a challenge because each AI assistant renders user interfaces differently, forcing those who maintain button components to adapt. While connecting agents is becoming easier, the rendering process across various clients remains inconsistent, with no single source of truth for component ownership or appearance. Google's A2UI aims to solve this by using declarative JSON, where agents name components that clients already possess, and clients handle rendering.This declarative JSON approach, similar to early HTML, is presented as a neutral proposal without vendor ownership. However, widespread adoption is hindered by various methods clients use to display UI, including sandboxed HTML and vendor-specific formats like Slack's BlockKit and Microsoft's Adaptive Cards. The core issue is the ownership of styling and component vocabulary, which is currently handed over to the client, complicating design system consistency. The author draws a parallel to the browser wars of the late 1990s, where competing standards led to duplicated effort and stagnation.The article emphasizes that the "canvas" or more complex UI areas beyond simple chat are where the real cost lies, with each vendor maintaining its own set of components. While declarative UI like A2UI offers a path to a shared "parts list," the current reality requires significant effort to make design systems work across numerous agent surfaces and platforms. The solution proposed is to choose a target standard, with A2UI being the most promising, to unify the rendering process and reduce the "surface tax" imposed by managing multiple implementations. Ultimately, the success of any standard depends on its neutrality and the willingness of platform vendors to absorb the complexity of rendering.
https://uxdesign.cc/the-call-for-an-agentic-standard-we-need-to-stop-shipping-the-same-form-four-different-times-be5b9d40e370?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_qcceL6xcVd.png
RSS Hunter RSS Hunter • Aug 31

Design has outgrown the traditional designer

Design education needs to evolve beyond its traditional focus on creatively inclined students. Historically, design offered a practical career for those with artistic talent, emphasizing visual skills like typography and composition. This perception still influences recruitment and student expectations, presenting design as a profession where visual creativity leads to a recognizable career. However, the modern design profession demands more than just artistic ability. As the field has shifted towards digital products, designers are now responsible for research, systems thinking, ethics, and understanding technology. Many students still enter design programs expecting to focus solely on visual output, leading to a disconnect when confronted with these broader responsibilities. The demand is growing for designers proficient in digital interfaces, user behavior, and product strategy, as evidenced by labor statistics and employer hiring trends. Therefore, design education must reconsider its recruitment pitches and admissions processes. Portfolios should showcase problem-solving and reasoning alongside visual work, and recruitment should extend beyond art students to include those from psychology, computer science, and business. Curricula and evaluations should emphasize defending consequential decisions, responding to diverse needs, and demonstrating research and ethical reasoning. Ultimately, design education must prepare students not just to make things, but to understand what should be made, whom it serves, and its broader impact.
https://uxdesign.cc/design-has-outgrown-the-traditional-designer-adbb657aa30a?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_bf87h8SxCM.jpeg
RSS Hunter RSS Hunter • Aug 31

You can’t grow roses in a cornfield

The article discusses career paths, drawing insights from Whitney Hess, a former UX consultant turned coach. Hess highlights how a subtle feeling, a "whisper," that one's work no longer fits can grow into a significant crisis. She transitioned from consulting to coaching not out of dissatisfaction, but to address systemic issues and facilitate deeper engagement with people. Hess emphasizes that the whisper is precise, indicating a desire for facilitation over pure problem-solving. Her career shifts were strategically designed, unlike her initial leap into independence. Hess's coaching now extends beyond careers to life, addressing underlying issues that impact professional satisfaction. She views past experiences as valuable components that can be recombined into new endeavors, rather than discarded. Hess stresses the importance of environment, comparing career growth to planting in a suitable field. She advises designers to thoroughly research a company's design culture before joining, analogous to buying property. Ultimately, understanding oneself and choosing the right environment are crucial for career growth and fulfillment.
https://uxdesign.cc/you-cant-grow-roses-in-a-cornfield-490d47dd8f74?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_ZEcEVZYEpy.jpeg
RSS Hunter RSS Hunter • Aug 31

Blank pages, Researcher-in-the-loop, Design’s craft crisis

This weekly curated resource highlights key concepts in interface design, emphasizing that minimizing user memory load through visible elements is a long-standing usability principle. It notes that blank pages and untouched perfect ones are undesirable starting points for interaction. The newsletter features a sponsored AI for UX Design course, offering practical skills and a back-to-school discount. Editor picks delve into AI-enabled UX research, the complexities of AI ethics, and the financial investment in making AI smarter. The UX Collective, an independent design publication, aims to amplify diverse design voices and encourage critical thinking. Articles explore the concept of futurism in modeling, the challenges of communicating design research credibility, and a longing for a less intrusive, more contemplative digital experience. Additional content touches on redefining UX processes, AI's impact on design craft, and lessons from luxury goods on adaptation. Tools and resources discuss responsible AI auditing, the UI for agentic commerce, and the power of effective metaphors. The newsletter concludes with ways readers can support its continued publication.
https://uxdesign.cc/blank-pages-researcher-in-the-loop-designs-craft-crisis-7934ab0bc9cf?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_urbYyWEOJd.png
RSS Hunter RSS Hunter • Aug 31

UX in 2027 may be less about interfaces and more about behavior

The future of UX design will involve more adaptive interfaces that interpret user goals and act on their behalf. Software is evolving to assemble interfaces, choose tools, and perform tasks, moving beyond fixed user flows. This shift requires designers to define not only what a person can do but also what the product can infer, generate, and accomplish. Designers will focus on defining system behavior, establishing limits, and ensuring users can regain control when needed. Interfaces will increasingly act as temporary explanations of system understanding, recommendations, and information needs. Constrained adaptation, where systems respond to context within defined boundaries, is seen as the more useful direction than unlimited personalization. Trust will become a critical interaction primitive, encompassing permission, provenance, confirmation, reversibility, and escalation. Products will need to cater to both human users and their AI agents. The role of UX will expand to include designing for machine-readable policies and structured content, ensuring products are accessible to delegated journeys. Designers must consider the uneven adoption of AI and design for scenarios where agents may not function flawlessly.
https://uxdesign.cc/ux-trends-2027-beyond-interface-design-23e55b1daf07?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_dL7dX3O7sT.png
RSS Hunter RSS Hunter • Aug 31