👋 Hello hello,
Figma just did something that'll make every designer's jaw drop. You can now copy any website and paste it directly into Figma as editable layers. Just... like that. Years of "inspect element, rebuild from scratch" energy, gone.
Meanwhile Meta is quietly trying to become the only app creators ever need to open again, and ElevenLabs shipped something that might make us reconsider our entire video production stack. We also have some Claude Fable use cases this week that genuinely don't feel real.
Let’s get into it.
🔥🔥🔥 Three Curated AI Updates
Figma just shipped a Chrome extension that lets you copy any website and paste it straight into Figma as fully editable layers. No more screenshotting a layout, guessing font sizes, or spending an afternoon rebuilding what's already on the internet.
This is huge for designers who use real websites as reference for their own work. You grab a site, paste it into your file, and now you have live, editable components to riff on, tear apart, or redesign.
For UX teams doing competitive analysis, this alone is probably worth a standing ovation. The extension is live now.
Meta's Edits app, which launched last year as its direct answer to CapCut, just got a significant upgrade. At an invite-only creator event in LA, Meta previewed a built-in AI assistant and confirmed a desktop version is on the way.
The AI assistant connects to your Instagram data, so it can actually tell you what's working in your content and why. It analyzes your view counts and video retention, then suggests ideas based on what your specific audience responds to. It'll even recommend trending audio to pair with your next post.
The desktop version brings Edits into serious editing territory. Creators will be able to sync their work between mobile and desktop, which closes a major gap CapCut has had over Edits since day one. Meta also launched a Beta tab today for early access to experimental features, plus deeper audience demographic insights.
The AI assistant is currently in testing with event attendees. Desktop is listed as "coming soon."
ElevenLabs, best known for its industry-leading AI voice technology, just expanded into video. Avatars inside ElevenLabs Creative lets you generate a talking video from three ingredients: a script, a voice, and an avatar. No camera, no studio, no filming.
Avatars can be human or non-human, and each one comes with a default voice you can swap out. The ones you create or favourite live in your Assets library and can be dragged into any generation. You can also use them with video models to keep a consistent on-screen presence across multiple scenes.
The part that stands out for anyone running content at scale is the new Avatar node inside Flows. You can batch-generate multiple variations of the same video by swapping out scripts, voices, hooks, or languages while keeping the same avatar across every version. For course creators, marketers, and social creators who need volume without losing consistency, that's a real shift.
We've been using ElevenLabs for voice and HeyGen for avatars separately. This might be the first real case for consolidating both under one roof.
🔥🔥 Two Fable 5 Use Cases To Know
Claude Fable 5 is Anthropic's most capable model available to the public right now, and people are finding genuinely wild use cases for it in creative production. One that caught our eye: generating cinematic 3D animations for a product launch in just a few prompts. No 3D software expertise required, no animator on retainer. You describe what you want, and Fable handles the visual storytelling.
If you have a product launch coming up or content that usually needs a production budget, this is worth experimenting with right now.
Same model, completely different use case. Someone used Fable 5 to design a QDD actuator (a type of mechanical joint used in robotics), animate the gearbox, and run collision detection as part of the validation loop. The whole thing took about 30 minutes and around 400k tokens.
This matters because it shows Fable operating in technical territory that usually requires expensive CAD software and specialist engineers. If you're building physical products or working in robotics, this is worth paying close attention to.
🔥 Things You Didn't Know You Could Do With AI
Pool is a new app that launched this week, and its premise is deceptively simple: it's an app for your screenshots. But what it's actually betting on is that your camera roll is one of the most underused personal context layers you have. Every screenshot you take is a signal about your taste, your interests, what you're paying attention to.
Here's how to start using it as an actual workflow tool:
Download Pool from the App Store and connect it to your camera roll.
Let it ingest your existing screenshots — product pages, tweets, design references, notes, anything you've saved over time.
Use Pool to search and surface relevant context when you're working on a project, brief, or decision.
When you find something relevant, use it as reference input into Claude, Figma, or whatever tool is in your workflow.
Build the habit of screenshotting intentionally — treat your camera roll like a personal knowledge base, not just a dumping ground.
The founders say this little icon got them $2M in funding and 15M organic views before launch. That's a lot of signal that people feel this problem deeply.
Don't forget to rate today's post
Until next time,
Team @PracticalyAI
