Love when the devs post, it feels like golden information. Using CAPS for specific words helps me a lot, and draws attention really well for these models. Also, words like prioritize, always, never, etc help a TON.
Heck, I just tell it to “give my prompt to DALL-E verbatim” and that works every time. Literally just the one word ‘verbatim’ does the trick.
It’s like Open AI is working as hard as it can to undermine it’s value proposition. It’s already challenging enough to get value out of AI without having to wonder about or navigate someones “trying to help” or “good intentions.”
Here is a novel idea, just leave my prompt the hell alone, if it triggers some safety thing just refuse to generate the image. Good grief.
Your old wild west photo then won’t include the required number of people in wheelchairs and wearing hijabs…
Thanks to everyone for the suggestions here, i just wanted to add that i took all the prompting suggestions, tested them and built them into a very simple, open-source, DALL-E 3 web-ui which is available on GitHub, just look for ‘simple-dalle-ui’. I hope this gives everybody else looking into this a headstart. To note, none of the methods suggested here ever gave me full prompt control but each had a strong impact on the result.
Warning: the above has a pattern also used for scams: “open source”… and then a demo page you can’t inspect that wants your API key.
“Full prompt control” I’m keeping to myself…
nice, seems Dall-E will replace photoshop