Award-winning filmmaker. The Last Waltz - Winner, Dreamina AI Animation Summit at Annecy 2026. Dark, strange, and occasionally funny films.
Joined April 2020
- Tweets2.7K
- Following260
- Followers526
- Likes30.7K
Pinned Tweet
The Last Waltz trailer is here.
A game of cat and murder.
It started with a single image: a ballroom at night, all shadow and silhouette. The image became a concept reel. The concept reel went to Annecy and came back with a win. Now the film has a trailer.
Three hundred guests. A fortune in roses. A bride dead by eleven.
Watch it below. And when the police arrive — remember that everyone was watching, and no one saw a thing.
The Last Waltz, an animated pink noir murder mystery, currently in production at WellerVision. Winner, Project Proposal Track, Dreamina AI International Animation Conference, Annecy 2026.
Every one has a distribution program. New market places are needed.
"Just a prompt/skill" is the most expensive lie in AI in SF right now.
A friend's going all in on UGC content analysis. I almost told him not to. It's just a prompt, right? Nothing to build.
Then he asked what I use for it, and I couldn't name one product. Not one I've seen. One I use. Every day. For money.
What I had used was Claude. My logic was the same as everyone's: anyone with a prompt gets 80% of the way there in an afternoon. So I'd ask it to go find the videos, break them down, and it would. Structure, hook, a score. Just like Exolyt, just like every "paste a link, get a 7/10" tool out there. And I'd read it, nod, and close the tab. Not a single piece of content we've made came from one of those breakdowns. They're fun to look at. They don't change what we ship.
Could I get there with a better prompt? Probably. But a prompt that actually works here, a skill that knows which details matter, isn't something you write in an afternoon. It's weeks of running it, reading the output, finding what it missed, fixing it, running it again. The "80% in an afternoon" was the lie. It's the 80% nobody needs.
That's when it flipped for me. Look at this category for ten seconds and it's AI's comfort zone, too shallow to build on. Look at it for twenty minutes and there's a huge, unserved need sitting in plain sight. And it's not one category. Content analysis. Viral post scraping. Competitor research. UGC insights. Every one of them has real demand, real budgets, and nothing that owns it.
So here's the strange picture.
Users have the need. I'm one of them. Founders are going after it. My friend is one of them. And somehow almost nothing in these categories gets used, gets bought, or makes money.
Meanwhile I can't find a product that solves my problem, so I do what everyone does: write my own skills on top of a general agent. Those skills are shallow. They skip the details that matter, so I'm back tweaking them every single day. Prompting is convenient. Getting the output I actually want is not. It eats more of my time than a product ever should.
Demand on one side. Builders on the other. Nothing in the middle that works.
Why?
Two reasons, and neither is the model.
First, thinking. Everyone's using AI for the same things. Same queries, same tools, wildly different results. My friend explained why. "Break down this script" sounds simple until you ask: by what? Storyline? Visual style? The stickers? The motion when the product appears? The model will do some of this on its own, but the clearer you've thought it through before you prompt, the better it executes. That's the whole gap.
Second, data, and I'm sure this one won't get fixed. TikTok cut the search function out of its own Creative Center in 2024 after researchers used it to study the platform. The one public trend tool, and they pulled it. I used to work at TikTok. I know what's on the priority list. The API platform has never been near the top, and for most Chinese companies it never will be. Every analysis product downstream of it is built on sand.
So the work is in the details. Hundreds of small decisions nobody sees. Which is what I told him: don't go big this time. Find one wedge so small it looks embarrassing, and go deeper on it than anyone has bothered to.
And if you actually do that work, here's the trap.
You spend weeks on details nobody sees.
When you finish, you have to prove you beat a user who just prompted a general agent. The same user who, like me, is quietly unhappy with their own output but will never admit it on a sales call.
Then a new model drops, and half your skills, structure and fine-tuning look pointless.
Then you launch. Into a feed where "built with AI in a day" products are everywhere. I don't even open them anymore. One glance, swipe. Nine out of ten are slop, so the reflex is right. But the reflex has a cost. The products built with real care get swiped past too. Not because they're bad. Because they're buried under more noise than the internet has ever had, and I'm one of the people scrolling.
I know founders who built genuinely good things in these categories this year. Almost none of them got found. They didn't fail at building. They failed at being seen.
That's the one thing I don't see getting fixed: distribution. For Chinese teams going global especially, it's not something we're born with. We have to earn it, and most of us haven't yet.
So this is the market right now.
Users with a real need who can't find a product, so they prompt, and aren't satisfied, and tune, and tune again.
Builders who did the hard work, invisible.
And a whole set of categories everyone wrote off as "just a prompt" that are, in fact, wide open.
This is the best era for people like us. It's also the worst.
The hard part was never the idea.
AI has completed ruined my expectations of productivity. If I haven't made to the moon and back in 6 hours - I feel like I'm doing something wrong. Which means, 6 more hours! [whips cracks]
Terrible thing.
I separate the work and conduct audits. I use Anthropic to create one set of outputs and use OpenAI to conduct audits of the same. All work is done based on a set of principles and project files kept in an Obsidian vault.
I've been able to increase the amount of work I'm able to complete and don't have to deal as much with the agony of Claude.
Quick reflection staring at my computer at 10pm: if AI is taking my software job, why have I never worked so hard in my life?
Two options:
1/ The narrative is wrong. There’s so much to build that there’s more work than ever.
2/ I mostly work with AI agents now, so I’m the bottleneck. My human speed just can’t keep up with them.
Thoughts?
OMG @theworldlabs cooked with Atlas. Just cooked.
I loved production design before. Now I’m giddy. The Last Waltz just got an interactive upgrade.
In a world of abundance it will be interesting to see 1) how little self control people have and 2) what they lean on to bring themselves back to equilibrium.
Seedance 2.5 has a 15,000 character limit. This, I have come to learn, is surprisingly easy to reach.
Why don't more people enjoy stop motion? Really happy to see Wildwood get the press it deserves but all of Laika's films are excellent; they've just been ignored.
I've seen several stellar stop motion films in recent years that haven't captured the imagination of the public. Critical acclaim? Yes.
If you don't like stop motion, what type of animation do you like? Why?
Anyone watching this space would have realized this. People are going to collapse their own markets. t's just supply and demand (and bad writing). Fewer productions, made well, with great writing and attention to world building is how we win.
BTS footage of AI generated film is a challenge. Am I really go to show you clips of me hunched over the keyboard with an open journal and Sharpie pen near a half eaten bowel of cherries?
How about my head in my hands in the dark at 11:45 at night trying to not wake my spouse while I troubleshoot a prompt or threaten to choke Claude by its zeros and ones?
Writing the story, setting creative direction, sourcing the music - those are never the BTS stuff. And you'd never share that anyway. That is why writer's room are sacrosanct and the look of the film is kept under wraps.
With BTS, people want to see how a story comes together. They want to understand the process of how something moves from an idea to something that seems tangible even if it's just lights and shadow. It's why people watch cooking shows and This Old House.
This process is not like that. It's mental. It's ideation --> troubleshooting, pivot? Pull hair. Epiphany! Ideation --> troubleshooting, pivot. Epiphany? Bowel of cereal.
I suppose I could craft a short film of me running through a clock, dodging gears and sharp edges trying to find my way to the other side. That's pretty close.
Stop motion is timeless. And in this case remarkable too.
Today's Oscar-qualified short is Bára Anna Stejskalová's '9 Million Colors,' a stop-motion musical about a mantis shrimp who befriends a blind fish that we've had our eye on for more than four years.
cartoonbrew.com/awards/9-mil…
When a studio film uses AI, it arrives with lawyers, union agreements, and a press tour, and nobody calls it slop. The independent creator gets the default label. The technology gets absorbed and the word gets quarantined around the poors.
More below.
Practically all VFX workflows are incorporating AI, just not necessarily in the way people here expect. It's sourcing textures, references, previz, modeling, cleaning, roto, syncing. Aside from brute force tasks they're using hybrid tools to generate backgrounds, extras, and sometimes shots that may get modified. It saves endless man hours and compute time. It's generational shift tech, it's here, and giant productions pretending it's not is a temporary smokescreen.
Not sure where it goes or how fast it gets there, but the irony is major studios and VFX companies are already sailing on that ship while low budget filmmakers are publicly whipped when they step their toes into a dinghy.
The Bayhemification of AI Video
There are two full-length AI features with theatrical releases. Alkan Avcıoğlu’s Post Truth opened in Turkish cinemas with critical acclaim already attached. Gossip Goblin’s Gods Don’t Give Gifts opens later this year. Ground is being gained. But the tug of war continues.
AI video is in danger of being Bayhemified. A whole medium around maximum spectacle and minimum story. Explosions, dragons, morphing faces, twelve impossible things before breakfast, compressed into eight seconds.
Spectacle is fine. There’s real craft in maximalism, and I like a little Bayhem myself. The problem is that spectacle is the main thing the medium is seen doing. “AI-generated” used to describe how something was made. Now it’s turning into a genre tag.
- The feed rewards Bayhem -
AI video lives on the social feed, and the feed rewards as much visual information per second as it can get. A character’s face held for three beats of silence is three seconds of dead air in which the viewer might leave. Story-driven work takes time, so you can’t post every day. And posting every day is the business model.
- The stigma spiral -
So the public sees spectacle, called “slop” made by “clankers.” People doing considered, story-driven work are busy working. “Made with AI” then puts a negative label on the work before anyone sees it.
- The ouroboros -
Studios are already using AI, quietly. De-aging, background replacement, previz, VFX work nobody sees. Budgets will depend on it and guilds will negotiate its terms. When a studio film uses AI, it arrives with lawyers, union agreements, and a press tour, and nobody calls it slop.
The independent creator, with no marketing department and no guild, gets the default label. Stigma fades when respectable people are seen using the disreputable thing openly or when stories begin to break through. If the respectable people keep hiding the tool and creatives are forced to swim through the muck of vitriol, the use of it can stay disreputable indefinitely. This could be permanent: a two-tier system. The technology gets absorbed and the word gets quarantined around the poors.
It’s happened before. Sampling was theft when hip-hop producers did it on cheap gear. Gatekeepers reject your use of a tool right up until their own use is too established to question. The snake eats its tail, and the tail is where the new voices are.
- The missing Finchers -
Every medium gets defined by its prestige figures. Television was assumed to be shallow, until it wasn’t. AI video needs its Finchers and its Hitchcocks: filmmakers who care about what a new tool lets you withhold. The held shot. The story that trusts its audience. With everyone chasing Bayhem, the quiet lane is wide open.
- The venue problem -
Because I live in the real world, I must then ask, “Even if they make that kind of film, where does it go?”
Venues for films made with AI are like hens’ teeth. A twenty-minute story-driven AI piece has no natural home. YouTube’s algorithm punishes it for its length and pacing, or demonetizes it for AI content! Much of the festival circuit bars AI work. The festivals that do take it often won’t accept a film that’s already been online, so the work has to stay out of circulation for the whole festival run.
Tech companies are starting to figure this out. Their products will be judged alongside the content, and they have begun to nudge creators toward longer form. But the algorithm is still the final arbiter, and so far creators have kept shaping their work around its needs.
The new platforms for independent filmmakers working with AI come with their own trap: the terms of service. Some words that come to mind: perpetual, irrevocable, transferable, sublicensable. Put a character or a world on those terms and you’re building equity in a house you don’t own.
- My stake -
I have a stake in this. I’m making a feature-length murder mystery with AI tools. Completion is targeted for later this year. The film exists because a project proposal I submitted won at the Dreamina AI International Animation Conference, held at the 2026 Annecy festival. When the film is complete, I’ll submit it back to Annecy for 2027, to be judged on the same terms as every other feature.
A murder mystery runs on planted clues, withheld information, and long conversations in quiet places, and the audience trusts that nothing is wasted. It’s the opposite of Bayhem.
Bayhem is the default right now. It changes when creators make things that are patient and undeniable, and a place is built for them to be seen, on terms that let creators keep what they made. That’s when AI film is born.
But who knows, any number of things could happen from here to eternity.
Sitting down to watch 8 hours of film as a judge of the Los Angeles Movie and Music Video Awards. 🍿 So excited. Understanding and evaluating the creative choices of film makers from around the world is an honor and incredibly enriching.
Adventures in the Screen Trade is one the best books I've read on making films.
In it, William Goldman talks about the role of a director as being the person everyone turns to, all day, with an endless stream of questions from every department about wardrobe, props, lighting, blocking and line readings. Each one needs an answer, and the answers have to be consistent with a single vision of the film.
AI tools collapse those departments but the questions and decisions still land with you, the director. Depending on the bounds of your project, those decisions are made under financial, labor or time constraints - with an added component, the frustration involved with a new, sometimes fickle technology.
You are not doing less. You are doing more with less.
The anti-AI crowd keeps confusing “that became easier” with “you’re doing less.”
We’re doing more because it became possible.
You can make a scene without waiting for actors to become available, finding a location you can afford or building an entire VFX pipeline. Those barriers getting smaller means you can attempt things you would have abandoned before.
But now you’re responsible for all of it.
Every character has to remain the same person across different shots. Every location needs a layout that survives a camera change. Performances, lighting, movement, editing and sound have to belong together.
Generating one impressive shot is easy compared with making it work alongside the next twenty.
And I’m not talking about photorealism. A completely imaginary world can feel convincing. The audience just needs to believe the characters inhabit it, rather than watch everything reinvent itself between cuts.
Making the footage became easier. Holding the whole film together is still your job.
Apparently that’s called doing nothing. 👻
Something something Roy Schneider We're going to need a bigger boat.
We are very excited to announce this next chapter! AEIP is a new banner dedicated to platforming emerging filmmakers in the indie space and we can’t wait to bring more films to life 🎥
variety.com/2026/film/news/b…
