AI, artists, and how Bokuju actually works

A recent TikTok post got 257 likes and eight comments. Most of the comments were some variation of "AI slop."

I want to address this honestly, because I think the criticism is understandable and I also think it deserves a more nuanced response than either "you're right, I'm sorry" or "you just don't get it."

What AI slop actually means

When I hear "AI slop" I think of something churned out fast, with no real thought behind it. A content farm pumping out AI images, a faceless account flooding a marketplace with generated products. I get why that association exists. There's a lot of that stuff.

What it doesn't describe is spending time getting a battle system right. Or thinking carefully about what moves a king cobra should have versus a mountain chicken frog.

The sumi-e style is not random either. I wanted every card to feel like it came from the same world, the same hand. There are specific masters in that tradition who work at the intersection of yokai illustration and real animal anatomy, something playful and slightly otherworldly, rendered with genuine anatomical respect. That's what I'm trying to capture. The AI is executing a style direction I made. It's not generating whatever it wants.

I hear the argument about training data and what generative models are built on. That's a legitimate conversation. But before going there, think about the concept: snap a photo, get a painted card in 20 seconds. Generative AI is what makes that possible. It's a means to an end, not the end itself.

About the economics

To be clear Bokuju isn't an art exhibit. Sure you can share the things you snap online and compare collections, but the core of what Bokuju does is: you point your phone at a creature and 20 seconds later you have a card in your collection. That experience only exists because of generative AI.

That doesn't mean there's no place for artists here. I think there is.

An app that makes you wait two weeks for a hand-drawn card of the pigeon you photographed on your lunch break is an app that caters to a very specific niche. Maybe that's a good thing, maybe not. I'm not sure.

Most people who download Bokuju try their five free snaps and leave. Some of those who stay will care deeply about hand-drawn art and will pay for it. I think that's real and I want to serve it. But building the whole product around that behaviour before proving it exists is a quick way to build nothing.

Having said that, I grew up on Pokemon and Pokemon cards. I know how cool some of that art looks.

What I'm actually thinking about

Here's where I want to be constructive rather than defensive.

I can't commission one million species. There are roughly 8.7 million animal species on earth and I am one person trying to grow a product. Even if I wanted to pre-fill the catalogue with hand-drawn art, the conversation with an artist would go something like: want to be involved? Sure. Want to do a tabby cat because they're commonly photographed? Sure. Want to do it for free? No. Rev share? OK, but what's the guarantee? Honestly, none right now.

That's the real situation. I'd rather say it plainly than dress it up as a partnership program it isn't yet.

What I can build is this: if an artist wants to submit a depiction of a specific species by scientific name, I'll put it in the system. When a user captures that species, they'll be shown the artist's version alongside their AI-generated card and given the option to upgrade for 100 Daku, roughly $14.99. They see the art before they decide. If they buy it, every cent goes to the artist. I keep nothing from that transaction.

It won't make anyone rich quickly. It's a bet on the product growing. But it's real, it's built on actual transactions rather than promises, and it means an artist who genuinely cares about a particular species can put their work in front of people who just photographed that species in the wild and are already emotionally invested.

If that sounds interesting to you, reach out: [email protected]