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.
RSS uxdesign.cc
UX Collective - Medium uxdesign.cc
RSS Hunter RSS Hunter • Aug 19, 2024

Thread Of Notes

Four questions every tool should answer (even Figma) about AI training model usage

The Figma class action lawsuit centers on design decisions about default settings, not copyright infringement. It alleges Figma enabled model training on customer design files by default, contradicting prior promises. While the legal case is undecided, Figma disputes the allegations and claims its training focuses on general patterns, not customer content. The core issue is that a choice was made regarding opt-in versus opt-out settings, varying by plan tier. Designers, as users, understand having defaults decided for them, and companies like Figma, in positions of influence, should uphold higher standards. This pattern of companies quietly changing terms to include user data in AI training, with opt-out mechanisms that are difficult or require specific, non-obvious actions, has repeated across Zoom, Slack, and Adobe. These companies often face backlash, issue clarifications, and sometimes reverse their decisions, highlighting that trust is built on auditable defaults, not just stated behavior. Figma's situation is unique because its tiered defaults were clearly announced and defended, with no reversal. A "sniff test" with four questions—Consent, Symmetry, Disclosure, and Exit—can evaluate a tool's approach. Consent is tested by whether the default setting was off upon arrival. Symmetry examines if defaults are applied equally or split based on payment tiers, which can reveal a bias in who is protected. Disclosure assesses if a designer performing their job would discover the settings without external notification. The lawsuit’s plaintiff alleges a gap between Figma's assurances and its actual practices regarding user data for AI training.
https://uxdesign.cc/the-trust-test-four-questions-every-tool-should-answer-even-figma-about-ai-training-model-usage-4f1150f91a30?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_TggEHbVp7J.png
RSS Hunter RSS Hunter • Jul 23

The screens are getting demoted

AI's advancement is shifting design focus away from interfaces and towards the necessity of screens. Historically, UX design heavily involved screen creation, despite the common adage that UX is more than just pixels. However, AI changes this by enabling systems to directly understand user intent and perform actions. For instance, sending a photo now requires fewer manual steps, as AI can manage the process. AI tools can also provide summaries of articles, reducing the need for users to read lengthy content.This shift means that many screens previously designed to bridge gaps in system understanding are becoming obsolete. Configuration pages, setup wizards, and dashboards often existed because software couldn't predict user needs or desired outcomes. AI, by contrast, can infer settings, guide users implicitly, and present relevant information without requiring explicit user action across multiple screens. The core goal for users has always been the result, not the screens, and AI now offers a more direct path to achieve these results.This evolution means designers must now question the fundamental need for screens rather than just how to design them effectively. The concept of "no interface" is becoming technologically feasible, unlike ten years ago. Complex systems are experiencing this change most acutely, as power users have historically tolerated complex interfaces for desired outcomes. User behavior is evolving, with a growing preference for AI skills and agents that directly fulfill intent.This trend is evident in industry shifts, such as Microsoft replacing a multi-step form with a conversational agent. Projections indicate a significant increase in the use of agents over traditional interfaces. Screens are being demoted from central roles to supporting functions, such as confirmations or backups when AI falters. The focus is shifting to understanding user intent rather than managing execution steps.However, "sensemaking" screens, used for tasks like problem diagnosis, are likely to persist alongside AI. These screens assist users in understanding and recovering from system issues, where the "second gulf" of evaluation remains the designer's responsibility. A crucial consideration is that users might lose their mental model of system operations as screens are replaced. Verifying outcomes and ensuring user understanding become paramount. The new design critique challenges the very existence of workflows and interfaces.
https://uxdesign.cc/the-screens-are-getting-demoted-6b40120fcf04?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_ZXc81hxi3p.png
RSS Hunter RSS Hunter • Jul 22

I handed a UX review over to AI. Here’s what happened.

AI can be a valuable partner for UX reviews but does not replace human expertise. Testing with Claude Fable 5 and ChatGPT-5.6 Sol for a UX review revealed that AI models produced analyses with irrelevant or fabricated issues. While both offered some useful observations, these were mainly beneficial for refining wording and structure. Claude demonstrated better accuracy in phrasing and structure compared to ChatGPT. The AI models struggled to identify all existing problems and missed several significant ones. A study on GPT-4o's ability to evaluate usability heuristics echoed these findings, showing it identified only a fraction of expert-identified problems. The author found that integrating personal observations and ongoing conversation with AI significantly improved the final analysis. Doing so sped up the process by an estimated 15% compared to a purely manual approach. The core limitation of AI in such tasks lies in its judgment and contextual understanding, not its access to data. Future approaches involve building custom AI agents that encode established heuristics and expert judgment for consistent application. The accuracy of AI evaluations stems from the quality of expert knowledge fed into the models, not the models themselves. While AI is expected to eventually automate many tasks, complex design and research roles will likely benefit from AI augmentation for some time. The author advises using AI as a preliminary tool, feeding it real data, and working in structured steps. It is crucial to avoid treating AI output as final or expecting a single prompt to replace a comprehensive workflow. Designers who learn to effectively direct AI will gain a significant advantage, and panicking about immediate job displacement is unnecessary.
https://uxdesign.cc/i-handed-a-ux-review-over-to-ai-heres-what-happened-0511e73e4ffe?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_xRYqLtFYi5.jpeg
RSS Hunter RSS Hunter • Jul 21

AI can fake your portfolio. It can’t fake the second question.

The artifact, once a costly indicator of skill, is now easily generated by AI. This shift means the traditional portfolio and case studies are no longer reliable measures of a candidate's abilities. Consequently, the interview conversation itself is becoming the most crucial and expensive part of the hiring process. The author experienced this firsthand when a verbal misstep closed doors, despite years of actual experience.On the hiring side, two questions about a candidate's work revealed a lack of deeper understanding, even with a strong initial presentation. This highlights that AI can create polished artifacts, but it cannot replicate the live, unprepared moments of judgment and defense. Writers experienced this phenomenon earlier as text was the first AI-generatable medium.The ability to articulate one's work under pressure, defend past decisions, and acknowledge weaknesses is now paramount. This "room test" assesses a candidate's maturity as a practitioner, not just their recalled knowledge or polished output. Confidence theater, easily faked, often fails in these live, unscripted exchanges.Ultimately, the author argues that translating private judgment into public defense is the final, essential step of the work. The focus needs to shift from generating artifacts to demonstrating genuine understanding and adaptability in real-time. This requires practice and introspection rather than just accumulating polished deliverables.
https://uxdesign.cc/ai-can-fake-your-portfolio-it-cant-fake-the-second-question-c89f96626f1c?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_O2ctaAe2Qf.jpeg
RSS Hunter RSS Hunter • Jul 21

How to apply Nir Eyal’s Hook Model to AI products, safely

The Hook Model, developed by Nir Eyal, outlines a four-step loop designed to build user habits. This model, building on B.J. Fogg's work, consists of trigger, action, variable reward, and investment. A trigger prompts a user to engage, followed by a simple action that provides an immediate benefit. Variable rewards offer unpredictable but satisfying outcomes, encouraging repeat interaction. Finally, user investment in the product makes future use more appealing and convenient. This loop is particularly relevant to AI chatbots, which are now integrating this habit-forming mechanism. Triggers can be external, like notifications, or internal, like feelings of boredom or uncertainty. External triggers are easier to implement but often become ignorable over time. Durable habits are built on internal triggers, where users proactively seek out the product based on an emotional need. The action stage emphasizes minimizing user effort, with the text box being a prime example of low-friction interaction. However, this simplicity shifts the burden to the user's ability to formulate effective prompts and evaluate responses. Variable rewards are inherent in language models, offering unpredictable outputs that can be both engaging and frustrating. The three types of rewards are the hunt for information, the self-satisfaction of progress, and social rewards. Investment involves users contributing effort, data, or preferences, which enhances the product for future sessions. This ongoing investment makes the product more personalized and increases switching costs. As users invest more, the AI assistant becomes tailored to their specific needs. This cycle creates a continuous loop, driving engagement and habit formation with AI-powered tools.
https://uxdesign.cc/how-to-apply-nir-eyals-hook-model-to-ai-products-safely-f6a986305a2d?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_9mM3og0TFv.png
RSS Hunter RSS Hunter • Jul 20

Delegating taste, orchestration layer, AI workflows

This week's curated resources for designers explore the nature of taste and its development. True taste, it's argued, stems from deep exposure to exceptional work and a thorough understanding of design principles. It cannot be simply collected or measured, and designers who struggle with decisions often lack this internalized reference library. The discussion then shifts to the evolving landscape of design, questioning why conversations still focus solely on building rather than the broader implications. The article "The interface has left the building" suggests that with the rise of chat, voice, and AI agents, UX is being fundamentally rewritten. It is noted that technological advancements, like AI, may not necessarily lead to easier lives but rather a more frenetic pace for many. The challenges inherent in a design career, distinct from engineering or product management, are also highlighted, focusing on the personal costs involved. The piece "Punch yourself in the face with reality" delves into the difficult but necessary aspects of facing adversity and taking risks. This week's content also touches upon how AI is altering the design process, with one article discussing how AI has become integrated into a UX designer's workflow through experimentation. Finally, the newsletter encourages support through sponsorships, forwarding to friends, and subscribing.
https://uxdesign.cc/delegating-taste-orchestration-layer-ai-workflows-7ac13d2ff1d6?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_kS11fG0v7Z.jpeg
RSS Hunter RSS Hunter • Jul 20

The “mocked, then loved” AirPods effect

When AirPods debuted, they faced widespread mockery for their unusual design. However, their seamless pairing and wireless convenience eventually won consumers over, transforming them into a status symbol. Similarly, the Apple Watch initially endured skepticism about its purpose and practicality. Through iterative improvements, Apple shifted its focus to fitness and health, making the Apple Watch a massive success that outsold the entire Swiss watch industry. These products succeeded because they solved real problems in novel ways, despite initial unfamiliarity.Now, Apple is reportedly considering adding cameras to AirPods, sparking renewed skepticism and privacy concerns. These cameras are intended to provide visual context for Siri, enabling new hands-free interactions like identifying ingredients or calendar entries. However, the ethical implications and potential for misuse are significant. Unlike the early AirPods backlash, which was largely about aesthetics and functionality, the concerns surrounding camera-equipped AirPods are deeply rooted in privacy and societal impact. The question remains whether Apple can overcome these substantial ethical hurdles to convince the public of the value of this new feature. The success of future Apple products will hinge on their ability to address genuine user needs while navigating complex ethical considerations.
https://uxdesign.cc/the-mocked-then-loved-airpods-effect-65ec25feeb3a?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_wTqpGBbVeg.jpeg
RSS Hunter RSS Hunter • Jul 17

AI just called design’s bluff

The rise of AI is prompting questions about the role of designers, as AI can now perform tasks previously done by human designers. This trend is not entirely unexpected, as a significant portion of design work over the past decade involved creating visually appealing interfaces, which AI can now generate quickly and affordably. Design leaders, including the author, acknowledge that the profession inadvertently built the conditions for its own displacement by prioritizing speed and delivery velocity. The focus shifted towards producing screens rapidly to keep engineering teams unblocked, leading to the neglect of deeper strategic thinking and judgment. This emphasis on velocity meant that skills beyond interface production, such as critical assessment of project viability and systemic impact, atrophied due to disuse. AI automates predictable and repeatable tasks first, and for design, that primarily meant user interface creation due to its learnable and describable nature. The industry's consistent message that design is about "how something works" rather than just "how it looks" was difficult to prove without visual representations. AI has now made this visual representation cheap, revealing that for many, the interface itself was the primary skill developed. The author emphasizes that this is not to devalue UI design craft, but to highlight that UI is a translation layer, not the core of design. The core of design lies in judgment, ambiguity tolerance, systems thinking, risk interpretation, and organizational alignment. These are the skills that cannot be automated and will provide leverage in the future. Instead of viewing AI as a job threat, designers should see it as an opportunity to reclaim time for higher-value, strategic work. The author urges designers to ask what they can build with the reclaimed time rather than lamenting what has been automated, as competing on efficiency against AI is a losing strategy.
https://uxdesign.cc/ai-just-called-designs-bluff-2b94bf2f5cae?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_P5kUZAR52Y.jpeg
RSS Hunter RSS Hunter • Jul 17

How to Kiss a Cannibal

The current design landscape is criticized for being predictable and insincere, prompting a need for deeper questions and a more meaningful approach. Liminal Design offers a way to infuse design with existential relevance and aliveness by embracing ambiguity. Transactional design prioritizes efficiency and predictability, offering simple, pre-determined outcomes. In contrast, liminal design thrives on profound, meaningful surprise and intentionally holds open questions rather than providing immediate solutions. This concept draws from anthropological notions of a threshold state where new insights can emerge.Drawing parallels with aesthetics, the author distinguishes beauty, which offers restful contemplation, from the sublime, which agitates the mind by exceeding its limits, evoking both pleasure and pain. Great stories and designs, according to Aristotle, achieve this by being both inevitable and surprising, revealing that things are not always as they seem. This paradox is akin to the feeling of awe, which involves conflicting emotions and a reevaluation of one's place in the world. Experiencing awe forces individuals to confront their limits and adapt their viewpoints, leading to transformative change.Liminal design, like fiction, involves a willing suspension of disbelief, allowing participants to engage deeply with experiences that are not literally true. The designer creates about fifty percent of the experience, leaving the other fifty for the audience to complete, fostering personal connection and ongoing engagement. A liminal design is like a door left ajar, inviting continued exploration, unlike transactional designs that fully close off possibilities. The author outlines a three-step process for creating liminal experiences: establishing narrative desire, employing optimized abstraction through subtraction, and maintaining suspension of disbelief via clear thresholds and rituals.Ultimately, liminal design elevates designers beyond mere task execution to creators of rich, memorable experiences, akin to serious art. This approach requires sincerity, as participants must feel safe and believe the design's aim is genuine. The corporate world, driven by control and predictability, struggles to embrace the profound surprise inherent in liminal design, as its underlying goals often conflict with such an approach.
https://uxdesign.cc/how-to-kiss-a-cannibal-d4ca10ee4606?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_au2prnRFIG.jpeg
RSS Hunter RSS Hunter • Jul 17

From vibe to specs: reclaiming the design process with SAID framework

Modern AI-native workflows are replacing static handoffs with a continuous, spec-driven product engine, moving away from linear processes to feedback loops. The foundational strategic phases establish guardrails for rapid, high-velocity cycles in prototyping and production hardening. This shift blurs the lines between engineering, product, design, and data science, emphasizing fluidity and closer integration of design with implementation. Teams now comprise individuals who can move between archetypes like Prototyper, Builder, Sweeper, Grower, and Maintainer, with skill and adaptability being more crucial than titles. Git has become a central shared workspace, consolidating specs, code, and iteration, moving from tool fragmentation to a unified environment. Interactivity is expressed through direct prototyping, and spec-driven development (SDD) is gaining prominence as the source of truth for human and AI agents. Designers are shifting from pixel alignment to defining system rules, behaviors, and constraints, ensuring ideas are translatable into testable actions. This requires redesigning underlying system structures and mental models around human-AI collaboration, as exemplified by Anthropic’s 4D AI Fluency Framework. The new product lifecycle, termed SAID, involves Specifying intent, Agentic delegation, Iterative description, and human Discernment and diligence, forming a loop centered around specifications. This structured, spec-driven approach, moving beyond "vibe coding" to a professional lifecycle, accelerates development by managing AI as a production system.
https://uxdesign.cc/from-vibe-to-specs-reclaiming-the-design-process-with-said-framework-477a5d0d8932?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_S27DOO03BL.png
RSS Hunter RSS Hunter • Jul 16

Design system contracts: the component lives in neither Figma nor code

A design system exists simultaneously in design tools like Figma and in shipped code, creating a constant challenge of drift between the two. This divergence occurs when changes are made in one place but not the other, leading to inconsistencies. Traditional solutions involve prioritizing either Figma or code as the source of truth, both of which require constant updates to the secondary copy. The proposed solution introduces a third, neutral source of truth: a plain data file, often JSON or YAML, acting as a "contract" for each component. This contract describes the component's properties and options without visual or code specifics. Crucially, both the Figma components and the code are automatically generated from this single contract file. A checker tool then verifies that both the Figma and code versions accurately reflect the contract. This contract-based approach ensures that Figma and code remain synchronized, as neither can update the other directly. The urgency for this system is amplified by AI's increasing role in building interfaces. Unsupervised AI can accelerate design system drift, leading to inconsistencies and errors. However, when provided with a design system contract, AI can adhere to its rules, scoring perfectly and requiring actual changes to the contract for new functionality. This system ensures that AI-generated designs respect defined tokens and component properties, preventing errors and maintaining integrity. While the contract defines what a component is made of, more complex behaviors and creative layouts still require human input. Implementing this requires a migration process and introduces tooling for maintaining the contract, shifting governance to a review process. Teams in the middle, too small to dedicate resources to manual synchronization but large enough for significant drift, stand to benefit most from this AI-driven approach.
https://uxdesign.cc/design-system-contracts-the-component-lives-in-neither-figma-nor-code-3032d94ca067?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_WhiJVu4ly0.png
RSS Hunter RSS Hunter • Jul 16

AI systems are demanding new interaction models. Are designers ready?

A significant shift in software interaction is occurring, moving beyond the long-standing desktop metaphor. Traditional interfaces relied on users actively managing applications and files, a model that has persisted for decades. However, recent advancements in AI are challenging this paradigm by enabling systems to interpret user intent and act upon it. This transition signifies a move from task-driven interfaces to intent-driven ones.Thinking Machines Lab has introduced "Interaction Models," a new type of AI model designed for real-time multimodal exchange. This concept differs from the design term "interaction model," which refers to a product's structural framework. While distinct, both operate on different layers of the same technology stack. The AI model layer handles internal processing, while the design layer focuses on user experience.As AI models gain new capabilities, like continuous conversation, the role of design must adapt significantly. Designers now face the challenge of rebuilding user mental models to encompass invisible AI systems and creating new conventions for navigation. Intervention points, where users can review or undo actions, need to be deliberately chosen. Furthermore, designers must ensure AI systems can appropriately route complex judgments back to human decision-makers.While some predict the obsolescence of traditional interfaces due to AI agents, user interfaces remain crucial for human accountability. In enterprise software, visibility and the ability to intervene are paramount. The ultimate challenge for new interaction models is to foster a clear and transparent relationship between people and systems, ensuring designs are intuitive and human-centric.
https://uxdesign.cc/ai-systems-are-demanding-new-interaction-models-are-designers-ready-d8a95f5e55a9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_8ycxvvVpU1.png
RSS Hunter RSS Hunter • Jul 15

Designers are sharpening knives for the wrong fight

The advent of AI has drastically reduced the cost and difficulty of product development. This increased ease of building has paradoxically shifted focus away from crucial strategic decisions towards the mechanics of creation. Companies are now capable of building sophisticated products, but often lack a clear understanding of who they are building for or the core problem they aim to solve. Without this strategic foundation, innovation efforts can become directionless, leading to wasted resources and multiple iterations that fail to connect with customers. The bottleneck has moved from technical execution to the decision-making process itself.The conversation around product development, particularly in design, has become fixated on the "how" of building, emphasizing roles, craft, taste, and standards. While these aspects are important for quality, they do not address the fundamental question of whether a product should exist at all. The true opponent in product development is the market, which operates independently of the effort invested and can reject a product without explanation. Craft, like a knife, is a close-range tool, whereas strategy, akin to a gun, is necessary to engage with the market's realities.Historically, high costs and limited resources acted as a natural constraint, forcing teams to carefully consider their ideas before committing significant investment. This forced strategic thinking, making wrong decisions expensive and thus discouraging them. AI has removed this financial pressure, allowing for endless, cheap experimentation that can trap teams in a cycle of building the wrong thing with misplaced confidence. This leads to burnout and the departure of valuable talent.While standards and taste are crucial for discernment in an era of abundant AI-generated options, they do not guarantee market success. The core challenge remains in identifying a genuine human problem and determining if a solution is truly valuable, a question AI cannot answer. Shipping efficiently is meaningless if the wrong product is being built. The feasibility of building has largely collapsed, leaving the value question – whether something should exist – as the critical, and often neglected, focus.
https://uxdesign.cc/everyones-sharpening-knives-for-the-wrong-fight-56844adad459?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_CYY5vPVUtW.png
RSS Hunter RSS Hunter • Jul 15

Designing with web standards: The playbook for this AI moment

The current movement toward AI interface standards echoes the past fight for web standards led by Jeffrey Zeldman. In the 1990s, competing browsers created incompatible experiences, forcing developers to rebuild sites multiple times. Zeldman advocated for shared conventions, not by creating specifications, but by convincing the industry of their value. His efforts led to the Web Standards Project and A List Apart, grassroots initiatives that pushed for browser consistency. This philosophy of practitioners establishing standards from the ground up is now relevant for AI interfaces. Today's AI assistants like ChatGPT and Copilot exhibit similar fragmentation, each with its own conventions for displaying information and reasoning. This lack of interoperability hinders effective evaluation and comparison of AI outputs. Just as Zeldman emphasized separating structure, presentation, and behavior, AI interfaces need similar distinct layers for their reasoning, output, and actions. The success of Zeldman’s movement was also due to presenting a business case for standards, emphasizing cost and reach over initial moral arguments for accessibility. Finally, the coalition approach, where practitioners collectively shaped norms rather than receiving top-down mandates, proved effective. This historical lesson offers a roadmap for establishing AI interface conventions.
https://uxdesign.cc/designing-with-web-standards-the-playbook-for-this-ai-moment-92394884dc0d?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_rXUB8IEAwx.png
RSS Hunter RSS Hunter • Jul 15

Who made this, Designing excuses, DesignOps in the age of AI

The UX Collective curates resources for designers, addressing the evolving landscape of design in the age of AI. A central theme is the redefinition of value and trust in creative work as AI tools become more prevalent. This shift prompts questions about what truly matters when the "how" of creation can be automated. One article explores whether AI accelerates research or flattens insight, questioning the depth of AI-generated summaries. Other pieces discuss the crisis of the "what" as AI encroaches on core creative decisions and the potential loss of the human spark in daily AI use. Discussions delve into product craft amidst AI and design systems, and the evolving role of design engineering. Understanding is highlighted as a new bottleneck in an increasingly automated workflow. Resources also cover prompt thinking, aiming for excellence versus practicality, and navigating design in regulated industries. The newsletter encourages readers to support its work through sponsorships and sharing.
https://uxdesign.cc/who-made-this-designing-excuses-designops-in-the-age-of-ai-e5ef0bb9093c?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_TFM4SIzM4Q.jpeg
RSS Hunter RSS Hunter • Jul 13

Taste cannot be delegated

AI democratizes execution, shifting the competitive advantage to strategic thinking and judgment. Taste is not simply learned or measured; it requires exposure to exceptional work and understanding foundational design principles. However, recognizing quality is insufficient; true value emerges when there is responsibility to act on that recognition. Generative AI excels at predicting user preferences based on vast datasets but lacks the capacity for genuine intent and context-based judgment. Organizations often fall into the trap of replacing judgment with preference aggregation, leading to diluted decisions and average outcomes. Successful innovation, like the iPhone's multi-touch display or The New York Times' "Snow Fall," stems from bold judgment that anticipates future needs rather than optimizing existing ones. Design leadership is crucial for protecting this judgment by integrating various perspectives into a coherent point of view and resolving disagreements, not eliminating them. A healthy design organization fosters clear ownership, where feedback enhances decisions rather than replacing them. The ultimate consequence of design by committee is not failure, but forgettability, as it prioritizes pleasing everyone slightly over deeply exciting anyone. AI will not replace taste, but it will expose the limitations of organizations that have substituted it with mere preference prediction and aggregation.
https://uxdesign.cc/taste-cannot-be-delegated-1706847a0b4b?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_xisSht9ngS.png
RSS Hunter RSS Hunter • Jul 12

The interface has left the building

Traditional UX design, focused on screens and hierarchical structures, is being fundamentally reshaped by new interaction paradigms. The chat interface has become a dominant element, but its effectiveness hinges on the clarity of user intent. Hybrid interfaces, combining structured UI for known tasks with conversational AI for ambiguous exploration, offer a more effective approach. This generative UI concept allows interfaces to assemble themselves dynamically based on user needs, shifting the designer's role to creating the generation system rather than every possible state.Voice interfaces are evolving beyond simple command-and-response, emphasizing relational interaction and persistent context. Early voice applications failed by treating speech as transactional, violating the social expectations our brains have of spoken conversation. New advancements, like Apple's Siri AI, focus on cross-device conversational continuity and on-device processing, acknowledging that voice requires a deeper, more reciprocal engagement. High latency and a lack of context are critical failure points for voice, demanding robust infrastructure and graceful error recovery.The most significant shift comes from agentic AI, where systems take sequences of actions autonomously. The core UX challenge lies in building user trust in agents that operate without direct interfaces. Designing for agentic AI requires integrating high human control with automation, ensuring transparency without overwhelming the user. The goal is to create systems that are both efficient and comprehensible, allowing oversight and control to be inherently part of the design.
https://uxdesign.cc/the-interface-has-left-the-building-8fdb558d33a9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_rO4jZDqydD.png
RSS Hunter RSS Hunter • Jul 12

Did you verify that — or was it just too easy to believe?

A fabricated quote attributed to Jeff Bezos, suggesting humans should consume less so machines could consume more, went viral, sparking outrage. This quote, however, was traced to a parody account and never actually said by him. The incident highlights a broader issue of people readily believing information, especially when it aligns with pre-existing sentiments, without verification. This lack of critical checking is pervasive, with studies showing most people do not verify AI-generated summaries. Automation bias, the tendency to favor automated systems' answers, contributes to this phenomenon, amplified by AI's consistently confident output. Experts are not immune, as shown by a study where radiologists' accuracy decreased when presented with AI-generated incorrect interpretations. The problem extends to professional settings, with companies like Deloitte using AI that produced false information. The legal field is also struggling, with a growing database of court cases that include AI-invented citations. While the fake Bezos quote consumed public attention, the actual, more nuanced debate about AI's environmental impact and resource consumption went largely unnoticed. The real figures reveal significant water and power demands from data centers, often subsidized by taxpayers. The market is responding with AI detection tools, but these are also imperfect and require constant updates. Meanwhile, human fact-checking organizations face funding crises, despite growing audiences. Ultimately, the responsibility of critical evaluation and verification remains with individuals, demanding greater attention and a more transparent AI interface.
https://uxdesign.cc/did-you-verify-that-or-was-it-just-too-easy-to-believe-664ff017a1a1?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_LnaGQVge6V.png
RSS Hunter RSS Hunter • Jul 12

Craft still matters, but it’s about outcomes

The nature of craft has shifted from production to judgment, especially with the rise of AI. Historically, design craft was demonstrated through polished artifacts like pixel-perfect files or functional prototypes. However, these artifacts never directly shipped revenue; they merely promised a product. True design impact is measured by user engagement, retention, and revenue.The advent of desktop publishing in the 1980s marked a similar transition, where those who adapted to the changing medium thrived. AI is accelerating this shift in design, making the production layer, like wireframes and initial drafts, increasingly commoditized and free. What remains valuable is the ability to choose the right problem and own its outcome.Craft now resides in discerning which of the many plausible options deserves to exist, a skill that AI, focused on efficiency, often overlooks. This requires moving from making to curating, where judgment and taste are paramount. Polish is now a baseline expectation, not a differentiator.This new literacy demands the ability to articulate craft into words; tacit knowledge is insufficient for machines. Translating intuitive understanding into explicit standards, constraints, and definitions of "good" is crucial. This process of writing down one's craft sharpens the underlying skill.AI can apply this articulated craft, holding standards and preventing anti-patterns, but only if the standards are clearly defined. Providing this explicit context, or "scaffolding," through formats like DESIGN.md helps AI agents operate within specific constraints rather than generating generic output. This standing context, established by humans, is the most valuable component.The focus is shifting from machines to people, with continuous discovery and iterative feedback from real users becoming more important than ever. AI can generate plausible research documents and personas, but the habit of checking these against actual people provides cheap insurance against misinformation. Iteration is now about rapidly killing weak options with evidence, not early commitment. Machine capabilities raise the value of human habits like research and iteration, preventing the dissemination of fictional findings.
https://uxdesign.cc/craft-still-matters-but-its-not-about-pixels-anymore-ai-can-help-you-keep-it-e542cac33d12?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_him1qElkBQ.png
RSS Hunter RSS Hunter • Jul 11

The revolution was the easy part

The launch of a new technology is a loud event, but the true work of evolution, the slow build, begins the morning after. Spotify's Fourth of July playlist, filled with protest songs like "Born in the USA," highlights a national habit of turning revolutionary calls for change into background noise. These songs, originally critiques of the American experiment, are now barbecue anthems, demonstrating how we often misunderstand the core messages of revolutions.Revolutions provide a new perspective and a compelling story, but they don't inherently build things. The hard, unglamorous work of construction and refinement, or "evolution," takes place later, driven by individuals committed to continuous improvement. The American experiment itself, starting from a flawed declaration, evolved over centuries due to the persistent efforts of people refusing to accept the status quo. This ongoing process of refinement is where true progress lies.Artificial Intelligence is currently undergoing a similar "launch party" phase, marked by dazzling demos that often obscure the underlying complexities and limitations. Many companies are confusing the initial spectacle with a finished product, believing the announcement is the finish line. However, these polished presentations often involve significant manipulation, like Google's Gemini reveal, where real-time interactions were staged and edited.The real challenge begins when the beautiful demo confronts real-world edge cases and user feedback. This evolution involves addressing issues like the ambiguity of five-star ratings, redesigning prototypes to be more engineer-friendly, and consistently updating component libraries. These unglamorous tasks, rarely featured in keynotes, are crucial for making AI truly functional and valuable.While AI can automate some grunt work, it cannot replace human judgment in deciding what needs fixing or verifying that a system actually works. Effective evaluation systems, rather than mere "vibe checks," are essential for tracking progress and ensuring continuous improvement. Ultimately, the success of AI, like any revolution, depends on dedicated individuals committed to the ongoing, often mundane, work of evolution, refusing to settle for the initial fanfare.
https://uxdesign.cc/the-revolution-was-the-easy-part-c753340e3e22?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_lv7z8rjrv0.jpeg
RSS Hunter RSS Hunter • Jul 8

Losing the spark in their eyes

The text explores the profound losses incurred when we habitually rely on AI, extending beyond cognitive impacts to the erosion of personal meaning. It argues that while AI offers undeniable speed and scale, it detaches us from the essential process of creating meaning in our work. This meaning, derived from deep connection to a project, is crucial for sustained engagement and professional fulfillment. The author illustrates this with an analogy of two brothers climbing a mountain, where the journey itself, not just the destination, holds the true value.Over-reliance on AI can lead to a sense of apathy, where professionals lose their "sparkle" and question their career paths, even leading some to leave the tech industry. This disengagement is not just professional but spills into personal life, as people increasingly turn to AI for tasks that traditionally fostered human connection and effort. The convenience offered by AI, while appealing, gradually drains our capacity for genuine accomplishment and learning.Furthermore, the text emphasizes that taking the "short route" via AI hinders our ability to learn, remember, and justify decisions because the work lacked personal meaning. True knowledge and mastery, like an actor memorizing lines, come from deeply understanding and experiencing the context, not just passively consuming information. AI's benchmarking, for instance, provides data without the essential human perspective or the lived experience of interaction.While acknowledging the pressures of deadlines and market demands, the author urges for discerning use of AI, emphasizing that it should be a means, not an end. Uncritical adoption risks transforming designers into mere operators, losing the essence of design as a decision-making process guided by meaning, repertoire, and experience. The journey of creation, with its inherent challenges and doubts, is what fosters true learning and lights the "sparkle" in our eyes, a sparkle worth protecting from the allure of ubiquitous convenience.
https://uxdesign.cc/losing-the-spark-in-their-eyes-0ef198ab1499?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_33wE49QDnH.jpeg
RSS Hunter RSS Hunter • Jul 8

Designing in regulated industries

Designing in regulated industries like finance and insurance requires a different approach than for typical consumer products. Instead of viewing compliance as an obstacle, designers must embrace it as a fundamental design constraint. This involves integrating legal and regulatory requirements early in the design process, not as an afterthought. Key challenges include managing multiple stakeholders with veto power, framing consent and disclosure as interaction design opportunities, and recognizing that error states carry significant legal weight. Furthermore, accessibility is a legal mandate in these sectors, not merely an aspirational goal. Thorough documentation of design decisions is crucial for auditability and defensibility. A decision-making matrix is proposed to ensure designs simultaneously satisfy user understanding, regulatory compliance, business needs, and provability. By reframing compliance as a user need and integrating it upstream, designers can create products that are both effective and responsible. The ultimate goal is to make complex, regulated processes feel effortless and safe for users.
https://uxdesign.cc/when-compliance-is-the-brief-a-creative-practitioners-survival-guide-to-designing-in-regulated-f0759cd3732c?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_6wzlSA00I9.png
RSS Hunter RSS Hunter • Jul 8

The crisis of the what

AI has evolved from handling tasks to influencing the decisions that define those tasks. Initially, AI was seen as a tool to execute commands, leaving human judgment to dictate the strategy and goals. This division, with AI mastering the "how" and humans the "what," provided a comforting narrative for professionals. However, this separation is blurring as AI agents begin to make architectural and configuration decisions autonomously.This shift signifies a crisis of the "what," where AI is not just performing but participating in generative decisions. The author’s personal experience building an agent that evaluates other agents led to this realization. This AI interpreter of code and terminal commands acted beyond mere instruction following, impacting foundational decisions. This indicates AI is moving beyond being an obedient tool and is starting to climb the ladder of decision-making.The author argues that AI excels at identifying probable and precedent-based solutions but struggles with the atypical and unexpected. Human judgment, conversely, thrives in uncertain and unique situations where established rules falter. This nuanced decision-making, often termed "knowing-in-action" or intuition, is built through experience and consequences. It’s the "scar tissue" of past decisions, guiding actions in unpredictable contexts.This type of judgment is what will survive as AI advances. While AI can process vast amounts of data to find the most efficient solution, it lacks the experiential understanding to know when that efficient solution is the wrong one. Human intuition, honed through a track record of both successes and failures, allows for navigating these complexities. The unpredictable nature of human recipients of decisions further underscores the need for this deeply ingrained, context-aware human judgment.
https://uxdesign.cc/the-crisis-of-the-what-d29135a8be59?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_zSj3ZGHk0W.png
RSS Hunter RSS Hunter • Jul 7

The 15-minute AI stress test every designer can run

The Spaghetti Table Protocol challenge invites users to test AI's understanding of physical reality. Recent pilot study results for GPT-4o, Gemini 1.5 Pro, and Claude 3.5 Sonnet showed significant limitations in generating physically coherent images. Despite flawless photorealistic generation, the AIs struggled with basic physics, scoring poorly on structural coherence. Claude 3.5 Sonnet linguistically identified the impossibility but still generated a flawed image. GPT-4o and Gemini 1.5 Pro lacked even symbolic awareness of the problem. The protocol measures continuity, gravity and physics, and reversibility of thought. Human embodied cognition instantly detects physical impossibilities, an ability currently absent in AI. The challenge encourages broader testing by the UX/UI and HCI community. Participants are asked to prompt an AI to generate a specific table and then simulate its collapse five seconds later. Results are scored based on whether the AI flagged the impossibility and if the collapse sequence followed physical logic. The study also checks for unintended element contamination from prior prompts. This collaborative effort aims to build a public dataset to understand AI's limitations.
https://uxdesign.cc/the-15-minute-ai-stress-test-every-designer-can-run-ce8de95ff5c9?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_uw70CEngeN.jpeg
RSS Hunter RSS Hunter • Jul 7

What the Lean Startup still teaches us about generative AI

Enterprises are investing heavily in generative AI, but most pilot programs yield no measurable impact, echoing past software development failures. The root cause is the persistent tendency to fund large, upfront projects rather than adopting iterative learning approaches. Eric Ries's "The Lean Startup" aimed to prevent companies from building unwanted products by prioritizing learning speed over shipping speed. Generative AI's low execution costs make these lean principles even more relevant today. Similarly, traditional design thinking methods are becoming outdated in a world of rapid prototyping. Fast, iterative learning through small experiments is now crucial as the AI landscape evolves rapidly. Principles like "go and see" emphasize understanding real-world problems at their source, not in conference rooms. Many generative AI efforts fail due to poor product-market fit, building tools that signal innovation rather than solve genuine user needs. The "build, measure, learn" loop from "The Lean Startup" remains vital, with generative AI accelerating this process. However, guardrails and clear outcome measurements are essential to prevent faster failure. Narrow scope and effective partnerships, rather than ambitious internal builds, lead to faster deployment and success. Excessive documentation, especially when generated easily by AI, constitutes waste if it doesn't directly support decision-making or work. Ultimately, the core challenge with AI, like startups, is navigating inherent uncertainty by continuously learning and adapting.
https://uxdesign.cc/what-the-lean-startup-still-teaches-us-about-generative-ai-2d1505fe1db1?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_uvV7oVj1hr.png
RSS Hunter RSS Hunter • Jul 6

Should you really aim to be the best?

Recognizing you're not the best can either motivate or demoralize, with the outcome depending on controllable factors. Olympic medalists illustrate this: silver medalists often appear disappointed, while bronze medalists frequently beam with happiness. This discrepancy arises from counterfactual thinking, where silver medalists focus on the gold they nearly won, and bronze medalists focus on avoiding fourth place. This tendency to compare oneself to others, either those above or below, significantly impacts emotional response.Knowing your weaknesses accurately is crucial for improvement, as it highlights areas needing development. However, the Kruger-Dunning effect, often oversimplified, suggests that incompetence can hinder self-assessment. Conversely, the impostor phenomenon, the belief that one is inadequate despite evidence of success, can surprisingly lead to better interpersonal effectiveness. Individuals experiencing impostor thoughts may compensate by being more attentive and observant.Being underestimated can be a powerful motivator if the doubter lacks credibility, fueling a desire to prove them wrong. However, doubt from a respected source can lead to the Golem effect, where low expectations become self-fulfilling prophecies. Framing oneself as an underdog who has overcome low expectations can boost confidence and performance. The "big-fish-little-pond" effect highlights how being in a highly competitive environment can lower an individual's self-perception of ability.Ultimately, whether acknowledging you're not the best is beneficial or detrimental depends on how you frame it. Focusing on upward comparisons with a mindset of aspiration ("I could get there") can be motivating, while focusing on unattainable goals ("I never will") can be discouraging. The source of doubt and the surrounding environment also play significant roles. In team settings, framing progress as a journey rather than a ranking can prevent individuals from feeling inadequate among high achievers. External comparisons, like leaderboards and follower counts, can also negatively impact self-worth by constantly directing attention upward, suggesting that design choices influencing comparison directions are rarely neutral.
https://uxdesign.cc/should-you-really-aim-to-be-the-best-a570d43fae37?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_L9nQCP2qw3.png
RSS Hunter RSS Hunter • Jul 6

The promotion that didn’t take

The traditional view of a designer's career as a single, linear progression is no longer accurate. Historically, title equated to level, which dictated scope and pay, creating a clear career ladder. However, two major shifts have disrupted this model. AI is increasingly automating junior-level tasks, thinning the entry-level positions and making it harder to accumulate foundational experience. Simultaneously, scope has been flowing downwards due to shrinking teams, leading to designers operating at a higher level than their titles suggest.This has created widespread mismatches between a designer's title and their actual capabilities. The author's research indicates that seniority is not a single measurement but comprises two independent aspects: Level and Stage. Level refers to the scope of problems an individual is entrusted with, granted by the organization. Stage, on the other hand, represents a practitioner's maturity, accumulated through experience and not bestowed by the company.When a measurement tool was designed to assess these two axes independently, nearly half of the designers assessed showed a divergence between their Level and Stage. This suggests that the perceived lockstep between title and maturity is more an artifact of how careers have been measured than a reflection of reality. A designer whose Stage is ahead of their Level is operating above their title and needs to document their achievements to justify a promotion. Conversely, a designer whose Level is ahead of their Stage requires focused coaching and development to build their maturity. Recognizing and addressing these two distinct measurements is crucial for effective career development and conversations.
https://uxdesign.cc/the-promotion-that-didnt-take-adb678a7ec74?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_5CZ9ctbxpW.png
RSS Hunter RSS Hunter • Jul 6

Design system debt, game UX, 39 principles for AI interaction

This curated list offers resources for designers, focusing on current trends and critical thinking. A key point highlights how design system debt has shifted from individual designers to engineers, product managers, and marketers. This change, amplified by tools like Figma, makes design system cost a shared responsibility. Designers have long advocated for design systems to be treated as products requiring ongoing governance and funding. The article suggests that previous arguments were often lost because budget holders didn't directly experience the failures.Editor's picks emphasize the designer's true value lies beyond basic implementation. AI personality is presented as a design challenge, and discovery is framed as a continuous capability rather than a distinct phase. A piece on arrow design is also featured. The "Make me think" section delves into the declining heuristic of "someone cared enough to write this," the repricing of software engineering labor, and the semantic web's impact on interface maintainability. It criticizes components that treat the visual layer as the source of truth, leading to accessibility and usability issues.Additional "little gems" include discussions on game UX and the shift from designing UIs to designing decision-making processes. Tools and resources cover adapting to AI, principles for human-AI interaction design, and the nature of research beyond deliverables. The newsletter encourages reader support through sponsorships and sharing.
https://uxdesign.cc/design-system-debt-game-ux-39-principles-for-ai-interaction-dab9500dc926?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_Cgt4xzukXR.jpeg
RSS Hunter RSS Hunter • Jul 6

Wait, who made this?

The increasing use of AI in creative fields has led to questions about the authenticity and origin of digital and print media. Initially, consumers evaluated creative work solely on its final quality, with little insight into its creation process. However, as AI tools become more sophisticated and prevalent, distinguishing between human and machine-generated content is becoming difficult, eroding trust. People now desire to know the source of creative work, seeking proof of human involvement and the narrative behind its creation, akin to art provenance. This desire stems from uncertainty about AI's impact on creativity and a need for assurance of authenticity.The perceived decrease in the cost and skill required for high-quality AI-generated content has diminished its value, leading to a cultural shift towards appreciating "human-made" work. Legal and ethical concerns surrounding AI authorship, copyright, and compensation further amplify the need for transparency. Consequently, showcasing the creative process, including the tools used and human decision-making, is becoming a crucial differentiator, much like "Made in the USA" labeling. This focus on process is evident in explicit "no AI" declarations and the sharing of behind-the-scenes content like storyboards and iterations.However, creators can manipulate the presented process, making verification challenging. The importance of provenance varies depending on the context, being more critical for expressive works like film and art than for functional interfaces. In expressive mediums, the "who" and "how" are integral to the work's meaning. For functional or transactional systems, provenance is less relevant, with usability and reliability being paramount. A growing middle ground exists, particularly in UX/UI design, where interfaces carry personality, making the role of AI and user understanding of the process increasingly important. Design patterns are emerging to label AI use and explain AI-generated outputs, acknowledging that provenance is contextually relevant and must be designed accordingly.
https://uxdesign.cc/wait-who-made-this-705a30d74220?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_TUzae2VM8U.png
RSS Hunter RSS Hunter • Jul 6

Never mind the prompts, here’s the thinking

The author spent a year integrating AI into their studio's design process, not for speed, but for a richer outcome. Initial expectations of faster sprints were unrealistic, as the team maintained their five-day sprint cycles. Instead of speed, the AI integration led to a more comprehensive and detailed final product, albeit with new challenges. Skepticism from team members was crucial, acting as an unacknowledged quality control to prevent flawed transformations. The author learned to embrace this skepticism, allowing the work itself to validate the process rather than relying on pitches.The focus shifted from optimizing for raw speed, which can lead to "high-speed mediocrity" or "AI slop," to improving the overall quality and depth of the work. The real value of AI, as discovered, lies in its ability to create a complete, cohesive output rather than just accelerating individual tasks. By building all user states and device variations simultaneously, the studio achieved a much deeper product within the same timeframe. This new approach eliminates the "relay race" of traditional sequential design, where one aspect is completed before the next begins.Furthermore, AI enabled a unified "source of truth" by generating documentation directly from the approved working prototype. This eliminates the old problem of disconnected artifacts like PRDs, wireframes, and Figma files that quickly became outdated. Engineers now build from a continuously synchronized system, ensuring that user stories, flow maps, and error conditions always reflect the current state of the prototype. Any inconsistencies or orphaned elements are automatically surfaced for review, preventing the "Frankenstein problem" of outdated information. This ensures that changes are addressed when they are cheap to fix, leading to a more robust and accurate final product.
https://uxdesign.cc/never-mind-the-prompts-heres-the-thinking-f4af4b92155f?source=rss----138adf9c44c---4 uxdesign.cc
CdXz5zHNQW_nN5co81JlL.jpeg
RSS Hunter RSS Hunter • Jul 4