way way way too censored. Ugly is banned. Hairy is banned. "Hot Gravy" banned. Absolutely ridiculous. The pricing is becoming too expensive by the day when you consider free options like stable diffusion

You're not preventing any lawsuits by crippling its ability to generate an exact face. My Larry David prompts are just some old bald guy with glasses but it's never actually Larry. I really don't think that's an actual thing to worry about.

The way the model doesn't ask for variables it needs to perform an operation When the service is down like yesterday When the model keeps repeating the answer it gave me or the question

I hope this help: 1. GPT3 playground needs a responsiveness update, it works awful on the phone... DALLE lab works better but still has bugs (zoom gesture for example). 2. The playground should generate new text where you left your cursor (like GitHub Copilot). 🧵 ->

a node-based interface for Dalle prompting could be a better way to give visual feedback on prompting. or more prompt modifiers to help understand or improve the weight of certain keywords

I've always wanted to send a direct feedback so that action can actually be taken and this may not be related to api but the open ai bot in 1v1 mid in Dota 2 was terrifying when it was released but in a 5v5 match the bots are terrible.

You prevent africans from using it.

When experimenting with DALLE-2, my account got de-activated perhaps from me being too experimental/exploratory. I sent an email to see if I could get re-activated, but got no reply. I found that disappointing and annoying.

Where's my agi ?