ChatGPT Vision can take in screenshots from Figma and generate code. Building with AI is getting wild.

This is a cool follow up demo of using the mobile app’s “draw on image” feature to edit the component we just generated.

Ooo, the docs step is really nice Been trying to figure out a good flow for components like this in cursor RAG patterns and libraries will be huge Open interpreter uses a similar trick, does semantic search from code dataset for examples Openv0 & v0dev is also on this path

The step by step instructions are fantastic! Vision on ChaptGPT will open up many opportunities and this is only going to get better 🤯 I cannot imagine how the world is going to look in a few years…

People suggest many tools as alternatives to ChatGPT, but I believe no tool can replace ChatGPT. A notable drawback is that all these features are for Plus users only. What's your take?

I want to build something with this so bad but they won't give it to me 😭

I know a lot of people are frustrated they don't have it yet, but it sounds like it's rolling out over the next 2 weeks. I honestly just got lucky being in an early wave.

Did you do anything to get access to ChatGPT Vision so early? I'm still waiting

My first thought when I saw the OpenAI vision demo was that it would become exceedingly simple to clone any UI. Some dev experience with next level prompting could likely make it trivial. Those are the skills if you’re looking for a moat. Things are different

RAG + GPT-4V on API is gonna unlock some magical capabilities, really excited to see what's going to be possible. Existing RAG apps like @cursor_ai & @vercel's v0 can make great use of it.

What you are have been doing is wild! I use Figma so this will probably come in a wee bit handy! 🙌