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.
Not great, Bob. I've run into one or two psycho iterations of Claude. Nothing like this of course. Just a low level recalcitrance. But, yeah. Not good, Bob.
if you have a qwen3.8 27b around, try steering L34 N3838 +0.75, now you have evil qwen 27b
for extra evil add L38 N17180 +0.75
same recipe as @_lyraaaa_: dose × median(‖resid‖ leaving layer L, tokens 1:) / ‖W_down[:, n]‖ at the down_proj input, every token
In 2027 and 2028, ai filmmakers will wish they’d read more closely the terms and conditions they signed in 2026 when they handed their IP to various platforms.
China's first AI film in theaters did poorly. Compare it to Niu Lai which made about $1,150 in its first ten days then went viral and exploded in sales despite the level of craft.
It's one data point but it points to what I've been thinking. That human energy signatures are one of the main things viewers react to and this is the next hurdle for filmmakers working with AI.
How much of your own energy signature appears in your work. And what does that look like? How is it experienced when working with AI?
Used to think I wanted a team, then I said "don't be silly, just use agents." Now I realize what I really need is a good old-fashioned clone. Straight vanilla, nothing fancy. Just extra me.
Kitty Kart Gang. Heist step 3 of 9. The vibe: a VHS tape of an underground cartoon.
Can Claude build a 3d felt world using @threejs ? Yes and I'm here for it.
Sound was either code or the Cinema collection from archive dot org. It took several passes to refine the fur look. Could be improved further. Still fun.
Compared to a version rendered using Seedance.
2025: "Worse case scenario, Children of Men & Gattaca"
2026: 'Worse case scenario, Terminator& Matrix"
2028 "You like Mel Gibson movies? i got some fun ones."
C.E. Whitmore retweeted
Odyssey with Grok Imagine is beautiful, well made and exciting. If it's turned into a series it will do well. Some will hate it, always. And some will reject it for a specific reason.
There is an energy signature to human creative output, not to be confused with a creative signature. And that missing human signature is what some are noticing and rejecting in AI film.
I don’t know what distinguishes an amateur filmmaker from a professional one. But when I watch a film with a very small budget and perhaps less capable acting, it’s apparent that the main thing any of us ever reacts to is emotion in other people. That spark of humanity ignites an involuntary response within us.
That spark is not in AI film (yet?) AI film speaks differently. It relies on something less clear. Beautiful and fun but still. It may be partly why hybrid uses of AI have flown under the radar while pure AI productions face backlash.
I love creating with AI. I doubt it will ever have the same anima found in human creations (famous last words), and that’s fine. We are learning it is still possible to spark emotional responses through stimuli and storytelling. It's all just shadow, story and symbols.
The delivery seems to come down to whether or not you like yours with a little anima, and how much. I love them both though I lean toward maximum freedom and a dash of curiosity.
I don't know where AI filmmaking tops off but I'm betting on the human creativity behind the tools. It's the same human energy steered differently.
Watch Episode 1 of our Odyssey series made with Grok Imagine.
Let us know in the comments if you'd like a full breakdown of how it was done!
Creative collaboration with @WonderStudiosX and director Dev Walker @glumlot_
The only time I’ve wanted to watch The Simpsons in the last 20 years.
Odyssey with Grok Imagine is beautiful, well made and exciting. If it's turned into a series it will do well. Some will hate it, always. And some will reject it for a specific reason.
There is an energy signature to human creative output, not to be confused with a creative signature. And that missing human signature is what some are noticing and rejecting in AI film.
I don’t know what distinguishes an amateur filmmaker from a professional one. But when I watch a film with a very small budget and perhaps less capable acting, it’s apparent that the main thing any of us ever reacts to is emotion in other people. That spark of humanity ignites an involuntary response within us.
That spark is not in AI film (yet?) AI film speaks differently. It relies on something less clear. Beautiful and fun but still. It may be partly why hybrid uses of AI have flown under the radar while pure AI productions face backlash.
I love creating with AI. I doubt it will ever have the same anima found in human creations (famous last words), and that’s fine. We are learning it is still possible to spark emotional responses through stimuli and storytelling. It's all just shadow, story and symbols.
The delivery seems to come down to whether or not you like yours with a little anima, and how much. I love them both though I lean toward maximum freedom and a dash of curiosity.
I don't know where AI filmmaking tops off but I'm betting on the human creativity behind the tools. It's the same human energy steered differently.
Watch Episode 1 of our Odyssey series made with Grok Imagine.
Let us know in the comments if you'd like a full breakdown of how it was done!
Creative collaboration with @WonderStudiosX and director Dev Walker @glumlot_
Working on a murder mystery, you've got to search high and low for a good movie to learn from. I think I've found one and I'm taking notes.
A Talent for Murder opened yesterday. A murder mystery thriller from Anton Corbijn. Haven't seen it yet but the process of working out each character, mind and motives rings so true.
Oh and the real Highsmith spent her last years in Switzerland, in a house called "the bunker". She also kept about 300 snails and once brought some of them to a cocktail party in her handbag, on a head of lettuce.
Are you kidding?
::sprinkles eccentrism all over my body::
A Talent for Murder is a decent mystery thriller held back by an unwieldy metafictional conceit.
Our review: bit.ly/4BrrNiU
Next step: Longevity and Rejuvenation. We are almost there.
Researchers in Bonn turned blood cells from an 80-year-old into neural stem cells with a molecular age below 20. More than 60 years younger by epigenetic clocks, which estimate age from chemical marks on DNA.
The team directly reprogrammed blood precursor cells into neural stem cells, skipping the usual pluripotent stem cell stage. The rejuvenation unfolded gradually and could be tracked for over 100 days.
This is what happens when every frame of a video is written as code instead of generated as an mp4.
The Pearl, a fight over the light that holds the world together. Every frame is rendered from code. No camera, no video model.
No consistency issues. Just story.
It feels like 2022 again, and I doubt it takes four years to reach a Seedance moment.
I feel like we're looking at something so much BIGGER than just better and more detailed SVGs.
SVGs are essentially images made entirely from code. Instead of storing millions of pixels like a PNG or JPEG, it uses math instructions to describe shapes, paths, colors, gradients, etc....
and Quiver is now generating these incredibly detailed illustrations that look almost like polished 3D renders, except they're actually editable vector graphics.
But what happens when these models eventually get so good that they can generate SVGs that are indistinguishable from photorealistic images? At that point, why would we need image models at all?
and if you can create any image entirely through code, you could theoretically generate every frame in a video in the same way and suddenly, instead of generating a fixed sequence as an mp4, you would be able to generate fully editable programmatic scenes.
I don't know what this means for the future, or if it will even work in this way, but holy molly, it's gonna be interesting
China's first fully AI-generated feature, Sanxingdui: Future Past, opens today across more than 10,000 screens.
It took 100 people over two years and more than 10 million yuan - or $1,497.443 - to complete. Pre-sales as of this morning: 80,000 yuan - or $11,979.54
On Douyin, AI short dramas pulled more than 200 billion views in September alone. But, one Weibo user put it plainly: "If it's free to watch on streaming platforms, it could become legendary. But paying to watch?"
This is a major test in a market that loves AI content.
Would you buy a ticket?
Would you pay to see a fully AI feature in a cinema? Will Chinese viewers? Trailer below.
The trailer if you want to judge for yourself.
m.yangshipin.cn/portrait_vid…
The Wildwood reviews are coming in...
"Wildwood is obsessed with impressing you instead of immersing you. Its marketing has emphasised the 16 year journey to make the film, the meticulously crafted puppets, sets and outfits and the idea that Laika had to become worthy of making it. All of this serves to put you in the cinema in a position to fawn over Wildwood’s technical brilliance, while paying no mind to the film’s underwhelmingly hollow story. A film should earn your investment in its world and characters, Wildwood takes it for granted."
REVIEW: WILDWOOD didn’t work for me
It’s the kind of clean, well crafted “magic” that many critics only see value in animation for, but below that is a hollow, rote story lacking personality
Check it out on @skwigly
The French Dispatch is five years old today with a secret gift.
3 minutes of hand-drawn car chase, styled after Tintin and Blake & Mortimer. 15 animators worked over seven months in Angoulême, the same French town where they shot the whole film.
I was never into 2D animation. But shake a little splash of noir on it and I'm in.
























