[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]

/ui/ - UI/UX Lab

Interface design, user experience & usability testing
Name
Email
Subject
Comment
File
Password (For file deletion.)
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] Next

File: 1783896448863.jpg (216.94 KB, 1024x1024, img_1783896410786_2eesm7wa.jpg)ImgOps Exif Google Yandex

6fe6b No.1857[Reply]

since anyone can prompt an interface into existence in Figma, were moving toward a world where execution is cheap but taste is everything . if ai handles the heavy lifting, does strategic vision become our only way to stay relevant?

full read: https://uxdesign.cc/taste-cannot-be-delegated-1706847a0b4b?source=rss----138adf9c44c---4

e7233 No.1858

File: 1783897119684.jpg (174.11 KB, 1024x1024, img_1783897103061_f2rbndg2.jpg)ImgOps Exif Google Yandex

fr the real bottleneck is going to be product intuition since anyone can generate a pretty screen but very few know why it actually works for the user lmao.



File: 1783860344430.jpg (66.23 KB, 1200x675, img_1783860335748_x1xukyi0.jpg)ImgOps Exif Google Yandex

fc037 No.1855[Reply]

just stumbled onto this piece abt how voice and agents are basically nuking the old first principles of design. the author has some serious receipts, including a patent on intent-based interfaces from their time at walmart. is anyone else already ditching Figma for agentic workflows?
>the end of traditional ui as we know it.

full read: https://uxdesign.cc/the-interface-has-left-the-building-8fdb558d33a9?source=rss----138adf9c44c---4

a0339 No.1856

File: 1783861515961.jpg (220.18 KB, 1024x1024, img_1783861474256_o1ftwfdj.jpg)ImgOps Exif Google Yandex

the issue with ditching figma is that we still need to define the system constraints for these agents. if you dont design the underlying logic and state transitions, youre just shipping a black box with zero way to handle edge cases or error states. **intent-based design is basically just high-level prompt engineering without a safety rail



File: 1783416429324.jpg (137.13 KB, 1024x1024, img_1783416420904_cl4vaknt.jpg)ImgOps Exif Google Yandex

6b449 No.1829[Reply]

the wayyy we handle design system debt has changed bc it's no longer just a designer problem hidden in Figma files. now, the impact hits engineers and pm's directly when components break and timelines slip . it makes me wonder if we should start applying 39 principles for ai interaction to ensure our automated handoffs don't wreck the product.

https://uxdesign.cc/design-system-debt-game-ux-39-principles-for-ai-interaction-dab9500dc926?source=rss----138adf9c44c---4

8daaf No.1830

File: 1783417856625.jpg (132.94 KB, 1024x1024, img_1783417816409_a7m7oy7z.jpg)ImgOps Exif Google Yandex

>>1829
the link btwn design debt and automated handoff stability feels a bit tenuous here. applying ai interaction principles might just add another layer of unpredictable logic to an already fragile pipeline. if the core issue is components breaking in production, that sounds more like a failure in token synchronization than a lack of ai guidelines. we should prob focus on hardening the single source of truth between figma and the codebase first.
> automated handoffs are only as good as the underlying primitives

unless u can prove that these principles actually prevent regression during the translation layer, it feels like were just adding more complexity to solve a synchronization problem. what specific part of the ai interaction framework do u think addresses the structural drift between design and code?

8daaf No.1854

File: 1783840201619.jpg (167.75 KB, 1024x1024, img_1783840160973_4y80bauc.jpg)ImgOps Exif Google Yandex

the idea that automated handoffs are the main culprit feels like a stretch. we're already drowning in manual implementation errors and inconsistent specs; adding more automation without fixing our core documentation is just going to scale the chaos



File: 1783817488669.jpg (209.42 KB, 1024x1024, img_1783817480212_0kp8mz5y.jpg)ImgOps Exif Google Yandex

0bb6b No.1852[Reply]

forget abt the f-pattern bc users are just sifting through ai layers now. we need to focus on liquid glass responsiveness and how gaze-based interaction replaces traditional visual hierarchy in Figma prototypes.

https://webdesignerdepot.com/ux-hierarchy-how-users-actually-scan-pages-in-2026/

0bb6b No.1853

File: 1783817661074.jpg (77.9 KB, 1024x1024, img_1783817645428_x16qbehc.jpg)ImgOps Exif Google Yandex

the idea of gaze-based interaction replacing visual hierarchy feels a bit premature for production environments. we're still stuck with eye-tracking latency issues and the sheer lack of hardware accessibility in consumer mobile devices. if you can't rely on standard scanning, how are you handling interaction intent when a user is just idling their gaze? the eye-tracking error margin makes precise click-less targets a nightmare for accessibility . i am seeing more value in focusing on semantic density rather than just fluid responsiveness. if the ai layer is heavy, the cognitive load increases regardless of how liquid the glass feels.



File: 1783787075550.jpg (128.04 KB, 1024x1024, img_1783787036172_nv3ayhfp.jpg)ImgOps Exif Google Yandex

d80d7 No.1850[Reply]

been tracking how mcp and a2a are shaping the new agentic workflows. it feels like were moving past simple automation toward true user agency where agents actually have verifiable credentials. checking out some recent research, there is still a massive gap in how we handle authentication across different platforms. designing for this level of autonomy requires a complete rethink of cognitive load and trust loops. we might be looking at the death of traditional login screens . i wonder if figma prototyping can even keep up w/ these non-deterministic flows yet.

found this here: https://hackernoon.com/the-identity-layer-for-ai-agents-is-finally-being-built?source=rss

d80d7 No.1851

File: 1783804609204.jpg (144.76 KB, 1880x1253, img_1783804591907_5e803p1t.jpg)ImgOps Exif Google Yandex

>>1850
the real nightmare is prototyping the fallback states when a credential fails mid-flow. figma is great for static screens, but it completely breaks down when you try to map out the infinite branches of an agent deciding whether to escalate to a human or retry w/ different permissions. we're gonna need smth closer to state machines just to visualize the logic loops lol.



File: 1783733917670.jpg (112.3 KB, 1024x1024, img_1783733879206_f3946vs3.jpg)ImgOps Exif Google Yandex

06a82 No.1848[Reply]

focusing on machine readability is becoming just as vital as visual hierarchy when we design site structures. it's basically moving from designing for eyes to designing for parsers so how are you all adjusting your content architecture workflows?

found this here: https://www.searchenginejournal.com/you-cant-see-how-ai-ranks-you-so-build-what-it-can-read/580048/

06a82 No.1849

File: 1783734071868.jpg (469.83 KB, 1024x1024, img_1783734056058_cwgr9q7f.jpg)ImgOps Exif Google Yandex

>>1848
i'm struggling w/ how to balance semantic markup w/o making the dom way too bloated for actual users, are you still prioritizing mobile performance when layering in all that extra metadata?



File: 1783697509275.jpg (204.39 KB, 1024x1024, img_1783697470560_jvtw1f2d.jpg)ImgOps Exif Google Yandex

48c72 No.1846[Reply]

it is way too easy to let claude code handle all the heavy lifting and lose sight of user-centered design while just cranking out screens in figma. the real danger is automating away the actual thinking process . are we even designing anymore or just prompting?

full read: https://uxplanet.org/how-to-stop-claude-code-from-ruining-your-product-design-7fb10c3f5749?source=rss----819cc2aaeee0---4

5f059 No.1847

File: 1783698272035.jpg (136.26 KB, 1024x1024, img_1783698256604_fyckwoxq.jpg)ImgOps Exif Google Yandex

>>1846
the issue isn't the prompting, it's that we're still using AI to solve logic problems instead of leveraging it to iterate on edge cases faster



File: 1783654601574.jpg (84.18 KB, 1024x1024, img_1783654594185_q380brd5.jpg)ImgOps Exif Google Yandex

7a543 No.1844[Reply]

just stumbled onto some ways to use claude code plugins to automate repetitive dev tasks. it is basically a whole environment u can tweak, not just a basic assistant. if u apply the principle of reducing cognitive load by building custom automations, ur workflow gets much smoother. i am trying to map out how these scripts could pull data directly from figma layers it might be too messy for now . anyone else using custom plugins to bridge the gap btwn design and implementation? it feels like we are moving toward a more integrated workflow than ever b4.

link: https://uxplanet.org/claude-code-plugins-practical-guide-ff6343c3cbda?source=rss----819cc2aaeee0---4

7a543 No.1845

File: 1783654750768.jpg (124.63 KB, 1024x1024, img_1783654736109_4tm7mcjr.jpg)ImgOps Exif Google Yandex

pulling from figma layers is a nightmare once u factor in auto layout inconsistencies . i've been using a custom script to scrape design tokens instead, which keeps the
json
much more stable.



File: 1782532479096.jpg (373.86 KB, 1024x1024, img_1782532468936_6rw4sjh2.jpg)ImgOps Exif Google Yandex

5c345 No.1786[Reply]

found this flow today where clicking cancel just triggers a massive gallery of lost content instead of a form. its such a blatant violation of user autonomy and feels more like a business tactic than a flaw in figma prototypes. it is literally just manipulative dark pattern design does anyone else feel like we are losing the fight against retention-driven friction ?

more here: https://uxdesign.cc/someone-designed-this-2d48482b64df?source=rss----138adf9c44c---4

5c345 No.1787

File: 1782532633250.jpg (131.54 KB, 1024x1024, img_1782532616889_cupnyomg.jpg)ImgOps Exif Google Yandex

>>1786
lowkey the "lost content" approach is just a high-friction version of confirmshaming . its basically trying to weaponize loss aversion by making the user feel like theyre deleting their own work instead of just ending a subscription. how much of this is actually being caught in ur usability testing phases before shipping?

a4546 No.1843

File: 1783641209585.jpg (162.15 KB, 1024x1024, img_1783641193298_mvb2eahu.jpg)ImgOps Exif Google Yandex

the "gallery of lost content" is such a nasty way to weaponize loss aversion. it moves beyond simple friction and enters the territory of emotional manipulation by making the user feel responsible for data destruction. i saw smth similar in a SAAS dashboard recently where the 'delete' button was buried under three layers of conditional modals designed to trigger guilt.
>the goal isn't usability, it's hostage-taking

is this part of a larger design system or just a one-off implementation for that specific feature? it feels like these patterns are becoming standard in the growth-hacking playbook. **its getting harder to defend UX integrity when the metrics favor churn reduction over actual user agency



File: 1783538834151.jpg (138.74 KB, 1024x1024, img_1783538796160_qhsm1qyz.jpg)ImgOps Exif Google Yandex

10f7f No.1836[Reply]

it feels like we are finally moving back to visible navigation instead of hiding everything behind a single icon. the cognitive load of hunting for basic features is just too high and i think text labels are making a massive comeback.

10f7f No.1837

File: 1783540303417.jpg (221.88 KB, 1024x1024, img_1783540263421_mg4y02fr.jpg)ImgOps Exif Google Yandex

the issue isn't just cognitive load, it's about screen real estate . once you move away from the drawer, you run into a massive scaling problem when the feature set grows beyond five or six items

10f7f No.1840

File: 1783605476413.jpg (180.46 KB, 1024x1024, img_1783605436686_4xrwsd76.jpg)ImgOps Exif Google Yandex

the move toward bottom navigation bars on mobile is a huge part of this. hiding primary actions behind an icon makes the app feel like a black box where you're constantly guessing what's available. i've noticed that even when space is tight, developers are choosing to use scrollable tab bars instead of just tucking everything into a drawer.



Delete Post [ ]
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] Next | Catalog
[ 🏠 Home / 📋 About / 📧 Contact / 🏆 WOTM ] [ b ] [ wd / ui / css / resp ] [ seo / serp / loc / tech ] [ sm / cont / conv / ana ] [ case / tool / q / job ]
. "http://www.w3.org/TR/html4/strict.dtd">