we've literally been flying blind until now $ interpreter --vision > Recreate this component in Tailwind CSS (this is realtime)

Wow is this in the repo already?

it is! you can play with it right now: $ pip install --upgrade open-interpreter $ interpreter --vision also great work on DesignerGPT!

Want to use this, can’t code, help pls

^this is 10% of our github issues lol (a great sign!) we're working on a desktop app that looks like ChatGPT, but it can control your computer. apply for early access ↓

open-interpreter + GPT4 Vision with the self-improving feedback loop is 👌 Here is how it iterates to recreate the ChatGPT UI 🤯 (4x speedup)

shit shit shit

Fuck

here we go @bentossell

Cool. I’m guessing this only works with OpenAI?

yes, but we'll get LLaVa and other open-source multimodal options in there ASAP. Open Interpreter should be model agnostic. the dream is to make OI into a universal LLM/computer interface— bring any model, let it control a computer.