No way...
I just built a working web app inside Framer using the new AI Agent. I gave it my friend's image dithering tool as a reference, with one instruction: "Don't clone it. Build your own simplified version."
In a few minutes it:
• Designed a completely new interface
• Created reusable components and styles
• Wrote the code for the core functionality
• Let me upload an image, apply the effect, and download the result (It did not supported copy paste of image but it was MVP so can't complain)
Is the design perfect?
No. And honestly, that's where we as designers still add the most value. But going from an idea to a working MVP this quickly is kind of insane.
Feels like we're spending less time asking for components to be built and more time refining the product itself.
The future of design workflows is going to be really interesting.
(Screen recording ↓)
No way...
I just built a working web app inside Framer using the new AI Agent. I gave it my friend's image dithering tool as a reference, with one instruction: "Don't clone it. Build your own simplified version."
In a few minutes it:
• Designed a completely new interface
• Created reusable components and styles
• Wrote the code for the core functionality
• Let me upload an image, apply the effect, and download the result (It did not supported copy paste of image but it was MVP so can't complain)
Is the design perfect?
No. And honestly, that's where we as designers still add the most value. But going from an idea to a working MVP this quickly is kind of insane.
Feels like we're spending less time asking for components to be built and more time refining the product itself.
The future of design workflows is going to be really interesting.
(Screen recording ↓)
