That’s a really cool trick! I’m going to have to see if I can apply it to other things at work as well. Maybe a image to ascii and then a diff would be useful…
“The same screenshot of an iOS app, but the Subscribe button is clear blue and more prominent”
(Although it wouldn’t work directly, since it looks like git runs “text dump” on both independently and then compares the text naively).
But if one feeds an image description from chatGPT to Dall-e and back in a loop, how many steps does it take to revert to pure noise? (surely this has been tried? but I couldn't find it)
> instead of a list of pixels
We don’t store lists of pixels. Not even lossless formats like PNG does that. Good ole JPEG has 1:10 - 1:20 compression ratio, ballpark.