Junkyard King, the arcade game. 🕹️ In the spirit of the 80s/90s, I started experimenting with building a web-based Junkyard King game with GPT-6. There’s a TON of room for improvement. Movement is stiff, lighting needs work, assets need work… but as a prototype, it works. Mostly, I wanted to see how far I could push Junkyard King beyond the series and into something you can actually play. I’ll get back to this little rabbit hole later. 👀

WHITE RABBIT - FRAGMENT #3# In a world where the Queen doesn’t just control her people. She steals what makes them who they are. Their memories. Their skills. Their lives. Alice carries what was taken from them inside her. Hunted by a totalitarian regime, she must survive long enough to return what was stolen and bring the Queen down. Armed with other people’s memories, her smarts, courage, and a headstrong temper that gets her into almost as much trouble as it gets her out of, Alice might be Wonderland’s last chance. An @thisisoddkid dystopian reimagining of Alice in Wonderland about identity, memory, and what remains of us when everything is taken away.

Currently stuck between “I should probably avoid blatant copyright infringement” and “f*** it, I need to make a short fan film about a Jedi with PTSD trying to live with a lifetime of trauma and killing.” I’ve always felt like there’s a psychological darkness in the Star Wars universe that’s been left largely unexplored.

Meet PINK STATIC, the best band that never existed. I believe in repetition when building an IP. Not reposting the same thing, but giving people more ways to recognize the world, the characters and the tone. I produce the music for Junkyard King, so I started wondering: what if the music actually belonged to the characters? So Artie + Kai now have a band. PINK STATIC. Their first track, “One Step at a Time,” appears in Episode 2, and I’m using these little fillers to explore where the IP can go beyond episodes: music, visual identity, clothing, maybe even brand collaborations down the road. Made this clip with Seedance 2.5. One useful trick: feed the model a black video carrying the original track as timing/audio authority, then generate the visuals in smaller chunks. Much less audio drift, much tighter lip sync. Also pushing a rough screenprint / halftone / misregistration graphic language that may find its way back into the show. ODD KID RECORDS presents PINK STATIC One Step at a Time

Lately, I’ve been playing a little with this little boy right here… It’s funny, I never really saw myself making any kind of kid-friendly content, and yet here I am, IP after IP 😅 I guess emotion has always been a big focus in my work, and getting a laugh definitely sits pretty high on that list too. Turns out those two things translate pretty well to making strange little worlds for younger audiences. I also remember reading someone talking about NVIDIA’s DLSS years ago and making the point that, at some point, getting more realistic is just another way of looking more like everyone else. ¯\(ツ)/¯ That one stuck with me somehow. With all these new tools pushing toward more fidelity, more resolution, more realism, I find myself increasingly interested in going the other way. Finding characters, textures, imperfections and visual languages that actually feel like something. Anyway. Meet another odd little kid. More soon from Dead Normal and @thisisoddkid.

You got me there. 😄 Considering the pretty extreme interest this received, I decided to break open my poor little Codex piggy bank and go back in. @sama whenever you want for some free credits. Cleaning things up, adding more systems, moving over to @fal’s H3 Max Turbo, and generally pushing the engine quite a bit further. I’m also now exploring some more D&D-like mechanics after someone very rightfully brought that up. Not just voting on A or B, but potentially introducing stats, probability, character traits, inventory, consequences, maybe even rolls that can alter what the audience wants to happen. Which is where this starts getting really interesting to me. Less “choose your own adventure.” More “shared live world with rules.”

AI video doesn’t need another infinite slop machine. It needs a showrunner. Using MiniMax H3 Max, thanks @fal, I built "THIS WAY", a live film engine that remembers characters, arcs and consequences. The audience votes live. Two futures generate ahead. The winner becomes canon and the story keeps unfolding. It’s not perfect yet, but this feels like an important step: not infinite generation, but AI serving an actual plot through audience decisions.

Been thinking a lot about the amount of hate around AI lately, and especially how dramatically more polarized everything has become over the last six months. I started exploring this space about five years ago and, back then, people mostly seemed curious or indifferent. You could disagree, debate what the technology meant, argue about where it might go, and then move on with your life. Somehow that feels increasingly impossible now. A lot of people compare what’s happening to the arrival of CGI, but I’m not convinced the comparison really works at this scale. Traditional artists were obviously worried about CGI, jobs changed, entire workflows disappeared or evolved, and there was plenty of resistance. But CGI artists weren’t being insulted daily by everyone from industry veterans to some random 17-year-old or your neighbor who happens to be a landscaper and has absolutely no stake in filmmaking whatsoever. Today it’s “AI slop” here, “fuck AI” there, and people will genuinely go out of their way to find work they dislike simply to tell the person who made it that they hate what they’re doing. The magnitude of the reaction is fascinating in itself. Social media obviously changes the equation enormously. We didn’t have millions of people algorithmically rewarded for having the hottest possible take when CGI was becoming mainstream, and outrage now travels at roughly the speed of culture having three espressos. But I think there’s more to it than that. AI touches several anxieties at once. Copyright, employment, authenticity, inequality, identity, even the very personal question of what makes someone’s skills valuable when a machine suddenly becomes capable of doing pieces of what took them years to learn. And unlike CGI, this technology landed directly in everybody’s hands almost immediately. You don’t need a workstation, a studio job, or years of technical training to encounter it. It’s on your phone, in your search engine, in your workplace, in your kid’s homework. So the debate stopped being primarily about technology and became cultural remarkably fast. There are legitimate reasons to question how these systems are built, how people are compensated, where consent belongs, and what happens to creative jobs. I don’t think dismissing those concerns helps anyone. But there’s also something deeply unhealthy about turning every person who experiments with the technology into a proxy for all of those problems. Someone making a film with AI isn’t personally responsible for solving copyright law, labor economics, OpenAI’s business model, the future of Hollywood and your existential crisis before they’re allowed to post a 30-second animation. You can dislike the technology. You can refuse to use it. You can fight for better rules around it. All of those positions are legitimate. But five years into this journey, the part I find increasingly difficult isn’t disagreement. It’s how quickly curiosity has been replaced by tribalism. And historically, tribalism has rarely been where humanity does its best thinking.

JUNKYARD KING. CHAPTER 2: THE NEW GUY. ⚔️ Six weeks later, I think we just set a new benchmark for AI-assisted animated storytelling. 15 minutes... With our friends at @invideoOfficial ⚔️

One of my biggest pet peeves with AI filmmaking right now is how quickly we seem to settle for “good enough.” We keep saying these tools remove friction without lowering quality, and yet I still see so many 15-second sequences glued together with broken continuity, disappearing objects, obvious artifacts, no grading, barely any dialogue, and very little attempt at actual narration. The funny part is that the tools are getting insanely good, but somehow that seems to make some of us less demanding, not more. Getting a cool shot fast shouldn’t be the end goal. It should give us more time to curate, redo the bad stuff, fix continuity, work on performance, edit properly, and actually tell a story. Nothing wrong with short experiments or gimmicks, I make them too. But if we want people to take AI filmmaking seriously, we also have to show that we care about the filmmaking part, not just the generation. And then we wonder why the whole space gets dismissed so easily. None of us will get it perfect, myself included, but we can at least give enough of a shit to keep raising the bar.