𝐙𝐨𝐡𝐚𝐢𝐛 𝐀.

@DannPetty @krispuckett So cool! Rooting for your next mission

Not a bad week. → Freelancing for Stripe → Reframe won fan favorite at the Figma Makeathon → Saw another preview of RecentWork → New Studio Display arrived → Best of all, a huge offer hit inbox Things are heating up.

@krispuckett Congrats @krispuckett !

Feels a bit wild, but I'm really freaking excited and grateful to be on Deep Dive with Ridd. These new tools have changed how I create and have helped me in unbelievable ways. I get so fired up taking about AI and design. Ridd is also an incredible interviewer!

there are three ways to delegate work in 2026 to AI: problem is defined. you brainstorm with it, lock the specs, it implements. to a high-agency human: problem isn’t defined yet. you give them a direction, they brainstorm with you on a higher level, figure out the details. they write their own spec, run their own experiments, keep you posted. to a low-agency human: they ask you for detailed specs before engaging with the problem at all. so you brainstorm with AI, write the specs. and ask it to implement as well.

@robinebers claude code is down and you gotta build ui - use Gemini in cursor know exactly what to do in quick updates/explore with a superfast model - composer model and ofcourse it’s an ide. sometimes you should look under the hood

sometimes I think about leaving Cursor behind 🫠 it's still the best tool but I had so many conversations lately with people that feel the same yes, it's the best AI agent there is but the direction feels unclear, I sense a lack of vision wish I'd know what their grand plan is, or if they have one coming from an engineering and product owner background, this hits harder for me than others maybe, but I really hope they figure it out

@hemeon So many managers just throw ICs under the bus and keep their incompetence hidden

To be fair there's a lot of managers who don't really work that much, get paid a ton, and slow things down / those jobs probably should go away. So many ICs I know at FAANG companies get suppressed/compromised by poor leadership. If AI flattens orgs then I'm here for it.

@designertom done

We just launched the State of Prototyping survey. It takes 4 minutes to fill out on your phone: - Tools you're actually running - How you think about fidelity - Whether and how AI changed your process - Team context Help us make better decisions about our careers and teams by filling it out. Share it online and with a friend. Closes in 2 weeks, results by April. Thanks to our sponsors for making this possible for the community. - @MagicPathAI - @framer - @dscout - @mobbin - @magicpatterns - @dazl_dev Take it here: survey.uxtools.co

maybe earth is flat. considering all the conspiracy theories, even the wildest ones being proven true

@fjzeit We built our own versions with fewer loops and pinging humans

How are those "Ralph" loops going people? I haven't heard much recently. Did you manage to perfectly specify your needs up front? Have you one-shot your requirements? What amount of rework are you doing? What's your token consumption like? How many agent/skill assets are you managing now? How's your cognitive ownership?

@krispuckett @AnthropicAI Had a pretty unproductive day with sonnet today

Love @AnthropicAI but this feels like more common place these days.

product design with ai moves so fast that it gets hard to justify things that take time like brand illustrations, graphics, the perfect copy, the ui details. and it’s hard to quantify and box them into kpi’s to make a compelling use case as well. but those are the details that make all the difference.

@scottdwitt Not all of those calls are made by designers alone and are more collective. Designers tend to influence roadmaps and vision, but business and market reality/pmf make the final call

@HeyZohaib Good points. > Design isn't decoration > Engineering isn't just code. Seems like the high-level strategic direction is missing from your model: Who do we serve? Why? How?

Design is usually spec creation. Engineering is usually spec implementation. Design opens wide and then systems/structure emerge as you have a solution nailed. Engineering usually starts with the systems/structure that allows for a scalable solution. Roles will continue to merge but these key differences will and should stay intact. Also, design is not just UI.

@jamonholmgren i guess one big reason to let agents run when you’re not working is to distribute the token usage

@HeyZohaib I don't mind it! As long as you aren't constantly context switching.

My current agentic workflow is about 5x faster, better quality, I understand the system better, and I’m having fun again. My previous workflows have left me exhausted, overwhelmed, and feeling out of touch with the systems I was building. They also degraded quality too much. This is way better. I’m not ready to describe in detail. It’s still evolving a bit. But I’ll give you a high level here. I call this the Night Shift workflow.

@amfonte @hemeon @figma Thank you for sharing this! I used to go to photoshop > slice tool > divide by a few number to copy and paste in figma > autolayout. Lol

@hemeon @figma Hey Marc, I use this plugin all the time to resolve this! figma.com/community/plug…

@figma I have very tall screenshots from apps that I keep loading and the images are getting scrunched and reszied like crazy. please remove this limitation - let me add any size image to the canvas. Just destroys them.

@DannPetty @alex_barashkov Skipping figma for designers feels like skipping learning to code and jumping straight into vibe code. Figma, or any design tool that lets you manipulate objects easily on a canvas is important to mock things up for collaboration, reviews/approvals, or just to guide AI.

@alex_barashkov Totally fair points on where things are heading. But new designers need to get hired today, not in 2027. Figma still gets them in the room. Learn it, use it, and constantly grow and evolve with the industry from there.

Huge respect to Dann, but it does not matter whether new designers skip Figma or not. The market for junior- and mid-level positions in the digital industry is shrinking rapidly, and we’re only at the beginning. Learn AI, learn marketing, learn whatever you can get your hands on, but just don’t stand still and think your Figma skills will be enough by the end of 2026 to get a job at all.

Using @paper has made me such a skeptic (in a good way). Instead of asking AI to make the design change, i ask it to first show me how you’d do it using paper mcp. Then sometimes it’s easier to manipulate on paper’s canvas and have the AI follow that design. I felt for a long time that there was a gap with doing quick iterations, backing up versions that kind of worked, and just have the design evolution stored somewhere. Figma did it well but we’re not using figma anymore. Tools like paper, pencil, subframe the gap beautifully. No more “early iteration is on this xyz commit” or “just change the url to /v2” to see some other treatment of the layout or whatever. And if you actually have to revert, you risk losing the 10 other interactions you did. Now all design artifacts exist, in a much higher fidelity than a screenshot, or canvas based figma. You can even map all flows automatically in these tools to audit your design. Even discover/rediscover views that you forgot that existed and spot inconsistencies easily.

@onuro @DenisJeliazkov For websites, every department is focused on the success of their department only, and has a different vision of the overall org. The founders/primary owners start giving mixed signals. Ego trips pile up. They’re tests of a designer’s political acumen than design skills.

I don't think they're always ego trips. Imho it's the Designer's job to analyze the feedback and respond accordingly, without wasting self or other party's time. And no matter how stupid they sound, vague feedbacks like "I don't like it" are perfectly normal and common, and is part of the process. You can always ask the context of what's not to like, and still get "Just don't like it". And even then, you can actually still turn this dilemma into a passed result by simply backing your design with factual evidences. One thing for certain is you can not enforce people to give feedback on data or metrics. It's impossible :)

Design reviews are ego trips. "I don't like blue." "What if it was bigger?" "This doesn't feel right." Feedback without metrics is opinion. From now on: No feedback without data. No feedback without context. No feedback without solutions. That's a design review.

@charlieholtz Need a way to set up project scoped local env variables. Just like the build and run scripts

My Conductor workflow: - new workspace (⌘N) for every bug/feature - spin up the dev server to test (add a run script so ⌘R runs server for you) - if all looks good, create PR (⌘⇧P) - kick off a review (⌘⇧R), look at diffs + leave comments - merge, then archive (⌘⇧A) - start again!

@signulll good reason, but there are better reasons too. nowadays i’m watching my 1.5 year old explore the physics engine of this world. throwing the thing, biting the thing, scratching it, jumping on it. the way he does it all feels very scientific too. intentional instead of random.

if you ask most peeps why they want kids they’ll give you some npc answer. the reason why i think having children would be amazing is that you get a front row seat to consciousness booting up. like watching an os that god wrote light up for the first time. & as they grow up you get to examine the world through an innocence lens you basically shedded a long long time ago.