The new plugin by Meng To uses AI to upscale up to 8X the original resolution. Perfect for a variety of images including avatars, background images, Midjourney renders, and marketing images. (Interesting that the author heavily relied on ChatGPT 4 to build it.)
I love seeing how teams improve their workflows with internal Figma plugins. Here, Bryan Berger from Discord built a handy plugin to randomize avatar fills from their design system library. The source code is available on GitHub. (I guess something is in the air, as Miguel Solorio is building something similar for his specific use case.)
Built a handy new @figma plugin to randomize avatar fills from our design system library (growing!). Caching prevents duplicates and category filtering allows one to be intentional about the vibe. pic.twitter.com/AxLfbkM6Bo
— Bryan Berger (@bryanberger) April 6, 2023
I mentioned the FigGPT plugin in the last issue, and now Edward Chechique wrote a detailed walkthrough of using this plugin. It requires an OpenAI API key but then provides direct access to ChatGPT prompts and tools for UX writing directly from the Figma document.
“FigGPT is a tiny plugin that connects ChatGPT to Figma and helps you to compose and edit copy.”
Discord designer Daniel Destefanis built an internal plugin for generating entire conversations using ChatGTP and rendering them using their design system. This is an excellent example of automating a mundane part of day-to-day work with AI.
Built a new @figma plugin I call "Magic Messages". It generates entire conversations using ChatGPT and renders them using our design system at Discord. You can set a topic, # people talking, and # of messages. This way we can generate content for our designs more easily. pic.twitter.com/4M01yB87tZ
— Daniel Destefanis (@daniel__designs) March 29, 2023
Cool open-source proof-of-concept, related to the above post by Edward. For now, I’m skeptical of AI-generated components (explaining ideas and details may take longer than building from scratch), but bullish on offloading mundane tasks, error detection, and quality checks to the AI.
Edward Chechique shares some ideas for AI features that he hopes to see at Config 2023.
Adobe announced Firefly, a family of creative generative AI models coming to Adobe products. The integration with Creative Suite and responsible generative AI is what makes it potentially very interesting. I also like that it’s positioned (similarly to GitHub’s Copilot) as a productivity tool and not a substitute for creators. For more context, see this thread by Sudharshan and another one by Linus with mind-blowing examples.
Jordan Singer shares what Diagram learned from participating in OpenAI Converge with early access to GPT‑4 and how they’re using it in Magician and Genius plugins.
as part of @OpenAI’s Converge we’ve been building AI design tools with access to GPT-4 @diagram over the past few months
— jordan singer (@jsngr) March 14, 2023
here’s what we’ve learned and how we’re using it 👇 pic.twitter.com/sSHUympi46
A demo of a really smart FigJam widget that lets you collect data onto the canvas, fine-tune a model, and keep that tuned model directly on the canvas to generate new images: “With a few simple API endpoints (/train, /status, /imagine), I made a multiplayer-enabled (!) canvas that had live-trained ML models living in it. Many people can come together and try out the model, you can alt-drag trained models to try out explorations without losing your history, you can mark it up with pencil drawings and stickies and do anything else you’ve gotten used to in FigJam and Figma.”
Nick Stamas is making a good point that Figma is uniquely positioned to create an AI assistant for designers, but instead of replacing them, it will be more like GitHub Copilot. “There’s one place where an enormous pile of UI data exists in a way that could be used to train a large neural network: Figma.”
Spoiler alert: AI isn't going to take your design job.
— Nick Stamas (@nickstamas) February 27, 2023
But there's one company that's sitting on a hidden mountain of data that could radically change the way products are designed. The rest are likely vaporware or toys.
Here's my prediction 👇
If introducing Genius wasn’t enough, here is another upcoming AI tool for creating editable UI designs from a simple text description. As Vitaly Friedman rightfully put, “It’s about time to get used to websites looking exactly the same everywhere.”
Jordan Singer and the Diagram team share the preview of Genius, their new tool for generating user interfaces with AI. “It understands what you’re designing and makes suggestions that autocomplete your design using components from your design system. It’s important to explore lots of ideas and iterate in the design process, and Genius ideates and iterates alongside you as you design.”
Sara Brunettini shows how AI-powered Figma plugins can be used to streamline your workflow and make the design process a breeze. “Imagine AI assisting with repetitive tasks like naming layers, designing light/dark modes or even generating micro-copy for you. And the best part is these plugins integrate seamlessly with Figma.”
“A magical AI-powered design tool for Figma” is now available to the public with a 7‑day trial.
🪄 We're finally ready to release the public beta for Magician, a magical AI-powered design tool for Figma.
— Diagram (@diagram) January 24, 2023
Try for free → https://t.co/qtxyyKmhyF pic.twitter.com/41Ze81YHq2
Okay, my mind is blown. “Developing plugins for Figma using ChatGPT can be a powerful tool for automating low-level tasks and generating specific parts of code. However, it’s important to have a clear understanding of your goals and the steps necessary to achieve them in order to use ChatGPT effectively.”