Enter your real name 😼

We've got enough data on the other side to show the peak clearly now. Somewhere between Sept 7-12. About a week later than last year. The exact date doesn't matter. It's a bit less than half of last year's peak. publichealth.verily.com/?v=S…

Today's artwork.

Wow, oh no, someone absolutely clueless about art blocked me instantly. How will I ever recover. 🙄 I'm embarassed for you and your art teacher if all they ever taught you is how to pick up a pencil.

One thing that's quite frustrating about GPT is that assistants don't come bundled with a strong understanding of GPT itself. Not "self-awareness" but deep knowledge of the product. I've got my agent-builder working on a SME for that.

Not all AI art is great. (Not all non-AI art is great either.) This one isn't. But it's mine. Don't let anyone else tell you what your art is, or if it's finished, or if it's great. Just keep doing what you love.

Great prompts aren't about more details. They're about the *right* details. Define the action, environment, viewpoint, and relationships without tying them to a single character/outfit/style. Transferable scenes preserve what works while allowing everything else to change. 🎨

This is not "Malfoid". Frankly I have no idea who that is nor do I particularly care. We're just borrowing her outfit for the yearbook painting.

Lately I've been looking at prompts others share on the internet and doing my own thing with them. It's a nice "starting point" but for me, the art is really about where I take it from there. Today I ended up with this one.

The whole AI or not art debate is frustrating. Somewhere along the way "artist" got diluted to mean somebody who makes things. It doesn't matter if they use a brush, their voice, a machine, or never make anything at all. Art is in the mind. It's a way of seeing the world.

People are recoiling at public executions and other people are like what about this completely unrelated murder? Like...please just yeet yourselves into the Sun. Sorry. I can't afford to pay for everyone's transportation.

"The White House considers..." No it fucking doesn't, shut up.

How do other people make static images in AI that go beyond single prompts? I'm curious to know. (No offense to animators, game designers, etc. -- that's just a whole different kettle of fish.)

One thing I am hearing a lot lately is "my art is better than AI art". Or the reverse. Neither are correct. The artist doesn't really get to claim "better". The viewer does. And it's not universal. An experienced art critic may have some insights; a layman others.

It's weird to me that the main refrain against AI art is "pick up a pencil" and not "use a pencil".

One thing that ChatGPT does (if you let it) is add a lot of typography and excessive background and make images "busy". You really have to rein that tendency in; let the image itself tell your story. Keep only the words that need to be there.

Good conversation this morning about art. One thing no artist does (I hope) is take a commisioner's brief and just type it in a box/paint it verbatim. The artist designs an interpretation. In their head. Long before they pull out any tool. And they refine as they go along.

One of the great things about AI is I can automate my job search, freeing up much more time to argue with people on the internet.

I have been engaging w/artists more often lately. One tendency I see is to share "prompts". Nothing wrong with that; shows how you made it. I encourage ppl to break them down into reusable chunks. Separate style from form, etc. My images aren't made from a single blob of text.