We sat across each other at a local bar with good beer, good sandwiches and good spirits. I wanted to talk to Nat about AI and how I use it in my more creative endeavors, where I have a choice to matter.

Nat hates AI with passion and “out of principle” (he draws by hand and writes fan fiction) like many other folks out there, but unlike them, he can separate me as a person from what I do and how I do it. Also, his opinion is way more important to me, naturally.

The big argument is that AI destroys the planet (because of the power the servers of big companies have to draw) and that it rips artists off because it creates work like they do without them getting a dime or even credit. Those are true facts and I don’t have answers to those.

Most people who use AI tools today use the Cloud and the big tech that comes with it. The more they use it, the more the big tech behind it grows and draws even more power from the grid, creating more emissions. As well, most AI for imagery creation is based on scraped information from thousands of people and artists, and it’s impossible to give them credit even if we want to, let alone compensation.

In general, AI takes an existing problem we have as a society and puts it on steroids. Many companies not only use AI as a tool, they’re starting to force its usage, directly or indirectly, by preferring it over hiring more human help instead.

These problems are not going away.

Meanwhile, there are good uses for this technology. In my case, it enhances my photos (which I took, meaning they’re my original work) and makes them better.

For example, in one of my recent “doors” photos, one of the doors had a big ugly sign on it. Removing it with Adobe’s AI took me a few seconds, and the picture looks better. The door part in the photo now has a small part (say roughly 10% of it) that is fake, but the work as a whole is still mine. Further, Adobe’s auto-fill uses my photo to “rebuild” the door, so in theory, there’s no theft here since the pixels it uses come from my photo, not someone else’s (at least not the way I understand how this particular tool is used).

In my opinion, this is an ethically acceptable use of AI. I would do it on my own without AI, but because it would take more time and effort, I wouldn’t bother with it every time. So, in this case, AI is an enabler, something that enhances my original work. I still contribute to Adobe’s AI though, and the first two problems I raised are still there.

I can’t make this argument when I create “original” art from scratch using AI. In that case, there’s very little - if at all - of my own work involved. And while I use my own desktop for it (Stable Diffusion is not based on a cloud server), as Nat pointed out, it doesn’t mean that those inspired by what I created will follow suit and use the same tools I use. While this makes it “less” of my problem (I can’t be held responsible for what other people do with their tech), I am part of it because I’m part of a movement that inspires people to do more AI-related work, not less.

This bothers me, so I picked up a “how to draw” book yesterday. It’s not my first attempt to learn to draw, and it probably won’t be the last.

I suck at drawing. I can’t even sketch. AI gave me the option to express what’s on my mind in a way that has been impossible for me before. But perhaps now that I’ve seen what I can create with help, I may have more motivation to bridge this gap, or at least attempt to.

Maybe I can learn to draw the same way I use AI in some of my photos: in small fixes. The drawing as a whole would still be AI, but something like a facial expression or something in the background would be something I draw myself—at least for starters. Approaching it like this in small steps might slowly help. After all, no one knows how to draw at birth; it’s something you got to learn.

Well, we’ll see what happens.