It's Monday. Here's what changed in AI this week, and what it means for your design work:

This week I traded my desk for the south of France. Summer here pulls every French founder to the coast, somewhere between Monaco and Saint-Tropez, so I spent the week meeting entrepreneurs before I head off on holiday.

Most of them are way ahead of me, and that's exactly why it was worth it. A few conversations completely reset how I'm thinking about the next few months.

On the product side, we pushed the Limora UI forward and started building our biggest update yet - an MCP. It's a solid amount of work and easily the most fun part of the build so far. We're also rethinking Stage positioning, so expect things to sharpen soon.

AI image models just stopped being toys. This one makes product shots you can actually ship.

xAI shipped Grok Imagine Image 2.0 this week, and it jumped straight to #2 in the world for both generating and editing images. What matters isn't the ranking - it's that this is clearly built for people who make things for a living. It renders real, sharp typography (small text finally holds up), does region-only magic-wand edits so you change one part of an image without touching the rest, and removes backgrounds with clean transparency you can drop straight into Figma.

The story isn't "another image model." It's that image generation crossed into production. For the last year these tools were great for moodboards and useless the second a client needed the actual asset. Precise editing, transparent cut-outs, and readable text change that. You can now go from a rough phone photo to a clean product shot with no studio, no photographer, no afternoon in Photoshop. That's a real threat to stock photography and cheap product-shoot budgets, and a real gift to solo designers and small brands who could never afford either. The five-image reference feature is the sleeper - compositing was the one job AI kept dumping back on you, and now it doesn't.

Open Grok, switch to Quality Mode, and start with the e-commerce template. Upload a product photo, describe the set you want (lighting, surface, mood), and let it place the product properly. Use the magic wand to fix the one detail that's always slightly off, then export with the background removed and drop it into your layout.

Motion is still the part of design most people avoid - the tools are either too shallow (template video makers) or too deep (After Effects). Motionly sits in the middle. It's an AI-native motion graphics editor: give it your story, copy, timing, and assets, and it generates a real animation draft. Then you edit it visually on a timeline, so nothing's locked in a black box like most AI video tools. Export a clean MP4 when you're done. It plugs into Claude, Codex, and your dev environment, and you can feed it data to spin out variations at scale.

Claude Code builds generic UI by default. These 3 free plugins give it a designer's taste - better spacing, self-corrected layouts, and the small animations that make a product feel expensive.

💌 Know a designer who should be using AI smarter?
Forward them this email. Or just send them to logiaweb.net to join.

See you next Monday,

Say hi on X, Instagram or YouTube