yeehaw brother

I wanted to show what Wicked would look like with hard lighting.

Doing a manual relight by hand is something I am technically capable of, but the energy and concentration required is not currently doable with my current health. I am really struggling, but writing these educational posts is really helping me keep my sanity. I have discovered a profound love of teaching when I don't have the energy to actually be creating photography or doing my photo restorations.

Here are the original frames from the movie.

I had the idea to composite the stage actresses into the scene. Stage lighting is done with spotlights, which are hard light sources. I spent about 20 minutes trying to get them to blend in. This is all image editing muscle memory stuff and does not require any problem solving or concentration.

But they still look like they are on a stage.

Then I used gen fill to create hard shadows and blended them.

And that worked perfectly. And only took about 10 minutes.

Does the fact that I'm using this as an accessibility aid and a tool for educational purposes matter?

Should I disclose how I did it?

Should I just try to make my point some other way and omit the hard light versions?

I've had a few people express concern when I talk about using AI as a disability aid. They say it gives ammo to the other side. This frustrates me because I use ChatGPT to help plan and track my health recovery. I used it as an interactive journal so I can easily recall things to tell my therapist, even if I have brain fog. I use it to help finish my thoughts when I'm struggling to concentrate. I use it as a research shortcut to help me find good sources and to direct me to just the information I need within those sources.

I don't trust it implicitly. But it sure makes it easier to find information I can trust.

And there are artistic tools that reduce tedious tasks. Automatic selections, the remove tool, generative fill, and generative expand. These are very surgical uses of the same kind of technology that powers things like MidJourney. When I select someone's eyes, it is using training from millions of photos of eyes to accurately draw a line around them. When I remove a speck of dust, it is referencing millions of images to predict what was underneath that dust and then generates brand new pixels to fill it in.

Whenever I went grocery shopping I'd always look at the little scooter carts and walk by them. It felt embarrassing and I felt like I was too young to justify using them when others might need them. But I collapsed two different times trying to push a cart through the store. When Katrina visited we went to the store and she encouraged me to use the scooter. And I made it through the entire store without getting tired.

And I will be honest, it was a miserable experience. I did not want anyone looking at me. I hated the beeping to back up because everyone turned their heads. It was a hard exercise in social anxiety.

But I didn't collapse. I didn't need 3 days to recover. Why was I abstaining from using this helpful tool?

AI has saved me hours of energy. It has improved the rate of my health recovery. It has made my therapy more effective. And it lets me create art a little faster. By being a sounding board and helping me quickly clarify my thoughts, I am able to finish writing posts that would otherwise remain unpublished and die in my drafts.

This has been so hard to talk about. And I feel all of this guilt because I've seen so many absolutist opinions on AI from people I respect and who share the same general moral philosophy. But there seems to be no place for any nuance. People are afraid to even acknowledge these tools are useful.

Michael Burns had a discussion with someone who used ChatGPT to help her navigate a legal matter that kept her from getting evicted.

He had to do a follow up livestream because the reaction was... intense.

In the past I have talked about my quadriplegic friend using generative AI to help him feel creative again. He creates images and incorporates them into stories. He was an artist before his accident and it was the first time in over a decade he was able to manifest anything from his imagination. It has helped his depression and it has been stunningly therapeutic for him.

One person said he should learn to draw with his mouth.

He struggles to use a straw, but sure.

I don't know how to talk about this. I am afraid to talk about this. And I feel ashamed because people make me feel like I am betraying artists. Even though I have fought for artists consistently for over a decade.

I have never published original writing that contained a single word written by AI. I have not published original artwork that did not have an authentic artistic workflow. And when I use AI tools, I talk about it. I show my work.

I use AI to help me make things.

I do not use it to make things for me.

If I could push a button to destroy AI, I would do it. Without hesitation. But as long as it exists, I can't justify abstaining from using a genuinely helpful tool.

So I try to create an AI code of ethics and stick to them. But this particular gray area is confounding me.

I just can't justify using the energy to do this manually for an educational example.