2 Comments
User's avatar
Alicia Bankhofer's avatar

Great piece. I was under the impression that the image generator in Canva uses Dalle 2 (or 3?).

I use Canva oder Dalle-3 in ChatGPT for images.

The way I see it, the software gods have to master 3 things before teachers can fully use 1) character consistency (for storyboards etc) 2) anatomical accuracy in body parts but also organs so that they can be correctly labeled and 3) text / writing / spelling. If they have already achieved no 3 then yay! We can expect this to be standard with all tools soon.

What do you think?

Expand full comment
Tom Daccord's avatar

Thank you. Your comments are spot-on. We're not there yet, but improvements are coming quickly.

Expand full comment