Spent $15 in DALL·E 2 credits creating this AI image

It's fun to play around with it, but like the author found, what you get is often strange or useless. I also find 1k images too small to do much with but I realize making 4k images would be cost prohibitive. I also wish it could generate vector images as well as pixel images. That would be fun to use.

“In working with DALL·E 2, it’s important to be specific about what you want without over-stuffing or adding redundant words.”

I found this to be the most important point from this piece. Often people don't really know what they really want when it comes to creative work, let alone to some omniscient algorithm. In spite of that, it's a delight to see something you love from an unspecific prompt that you won't find with anything you receive from a human.

Dall.E 2 never ceases to amaze me.

For anyone interested in learning about what Dall.E 2 can do, the author also links to the Dall.E 2 prompt book (discussed in this post https://news.ycombinator.com/item?id=32322329).

I was curious to compare results with Craiyon.ai

Here is "llama in a jersey dunking a basketball like Michael Jordan, shot from below, tilted frame, 35°, Dutch angle, extreme long shot, high detail, dramatic backlighting, epic, digital art": https://imgur.com/a/7LoAtRx

Here is "Llama in a jersey dunking a basketball like Michael Jordan, screenshots from the Miyazaki anime movie", much worst: https://imgur.com/a/g99G7Bn

I want to see a few iterations of describing an image with AI, generating it, describing it again, generating it... Like when passing a piece of text through Google translate back and forth.

My current move is creating initial versions of images with Midjourney, which seems to be a bit more "free-spirited" (read: less _literal_, more flexible) and then using DALL-E's replace tool to fill in the weird looking bits. It works pretty well, but it's a multi-step process and requires you have pay for Midjourney and DALL-E.