Courtney Ring

You can tell when someone actually cares about the work. Justin’s portfolio makes that pretty clear. There’s strong product thinking throughout. Not just visually solid work, but work that clearly understands what matters and why. No over-explaining. No unnecessary context. Just the right level of detail. But that’s only part of it. What stood out more is how everything is put together. It feels considered. There’s care in how the work is framed, how the story is told, and how the portfolio flows as a whole. Even projects that could easily feel generic end up having a bit more depth to them. You understand the intent behind them. That combination — solid thinking + thoughtful presentation — is rare. Full breakdown of what I think works well and what still could use improvement in this week’s showcase. Link in the comments.

162

12 Comments

Be honest. Do you understand that Figma is an HTML app and the area that you actually drag the boxes and transform controls and type the fonts is just a HTML5 <canvas> tag and that you are literally using agentic AI agents to input things into an HTML browser-based tool, instead of just the browser? And then my second question is: why does Claude to Figma render things in the canvas tag instead of just actually rendering things for real and then just have the mouse and keyboard-based manipulation controls, instead of editing the simulated elements in the <canvas> tag, have them manipulate the actual things in the browser, that then writes back to the code and at that point you would have the first true production what you see is what you get manipulator interface. Just kidding that wasn't a question; I know why. But using a coding tool to write to an HTML wrapper that holds a <canvas> tag that shows you a simulation of the thing you want to eventually export out back into a coding tool to then export out and write into a browser. Well that just seems... a little bit suspect. I promise that next week somebody is going to have Figma-like front-end interface manipulation that just directly edits DOM elements and writes it back to code directly, cutting out the entire simulation layer. And if that was a paradigm that I believed in, I could probably have it done by the end of the week. I was present for many technical deep dives on Mural's entire architecture from the front end to the back end that uses the <canvas> tag in the exact same way. And I'm telling you, right now, using code to draw graphic art that then you re-export out and make code has maybe a month left at best until if you really want to use your mouse to manipulate things, you'll be doing it directly in the DOM. I personally am not a fan of that paradigm, and there have been tools that could use the inspector to write back to the code that have never really caught fire. I still think that it is backwards, but at least it would be cutting out the middle layer. And as much as designers who start with visuals try to convince me that they're thinking through more than just what something looks like, "the purpose of a system is what it does." And when your system is designed around manipulating what something looks like, that is the paradigm you are forced to be in because the medium defines the message. Anyways, see you in a month when this exists. In the meantime, visit https://lnkd.in/edZE8C84 and start learning March 2nd!

155

28 Comments

This is a fascinating look at how AI is reshaping the way we design and build. Our workflows, and the relationship between design and code are evolving quickly. Jenny Wen explores what this shift means and how we can adapt to what’s ahead.

6