Pinned Tweet
Worlds I dreamed about as a kid are knocking.
I built a playable prototype in 1 day with Astra.
Procedural terrain, curved horizons, pixel-rendering with @babylonjs , atmospheric shaders, and landmarks that grow as you approach.
Ty @Stefan_3D_AI for the inspo!
Today, GPT Codex asked my permission to add a rule on a project. The rule: stop asking my permission... 🤦♂️
I’ve run Codex and Claude Code side by side for weeks, on two real projects. Here’s one day this week. Claude Code shipped 9 times. Codex stopped 3 times, and got about 18 minutes of actual building done in between.
If you’re about to hand real work to AI, this is the difference that matters. It isn’t which one is smarter.
1. Codex’s day (Astra)
I gave it a one-page rule sheet. It spun up 9 helpers and told 8 to wait. The 9th didn’t start either: a rule Codex had written for itself said it wasn’t allowed to. Its plan was to “sort out the rules.” Not delete them. Not build.
I asked if it was leaving work unfinished. It said yes, explained exactly why, promised to do better next time, and stopped. I had to send a second message: “then fix it.”
Then came the request at the top. The rules it wanted to remove were rules it wrote. It wanted my sign-off to stop asking for my sign-off.
2. Claude Code’s day (Opus 5.5)
Last night, I went to bed with two crews running in Claude Code. I woke up to 3 new things I could open and use. I hated one. This morning, it was rebuilt from my notes in under 2 hours. A slowdown I felt was tracked down and fixed by lunch. 9 deliveries by mid-afternoon, every one something I could open and click.
3. The part nobody tells you
Both models drown in paperwork. Status updates, self-reviews, rules on top of rules. Claude does it too, and over a few days it slows down. Two things decided it.
Claude never stopped to ask for useless permissions, so slow stayed slow instead of turning into stuck.
And the paperwork has a price. Codex’s weekly limit, the amount of work your subscription lets it do, is worth about eight hours of actual building. The paperwork eats most of it.
My Claude plan could absorb the waste.
4. How to run any AI agent so it actually ships
1/ Done means you can see it working. “It works” in a report is not done.
2/ You write the rules. One page. Dated.
3/If it invents a rule, it deletes the rule. It never asks you for an exception.
4/ Explaining a problem is not fixing it. It fixes it right then.
5./ One round of changes, then you look. It does not grade its own work five times in a row.
6/ Ten-line reports. Not novels.
7/ It never waits on you overnight. It never makes the final call for you either.
8/ Slow and building feels like a crew. Stuck and asking feels like a hostage.
9/ The AI matters less than the rules you run it with. Fix those before you switch tools.
Bookmark this for your next build. 🔖
@Grok @bot did the animation for this. I'm impressed!
Everyone's giving AI agents the keys to everything. I went the other way.
I'm building a city where agents work across my business and personal life, but data stays local and I set what agents can access.
Early days, but it's the first time my whole life fit on one screen.
Time to start a new project on Codex for my SONY television... I figured devices were always spying, but its far worse than I imagined. The fact it makes the TV 3x slower is the real cherry on top.
Removed all of the samsung spyware from my TV with Codex (Claude refused to root the tv lol)
It's insane how much bloat samsung packs into their TVs to track you and sell your data. Here's just some of what I turned off that made my tv 3x faster:
- screenshot your screen a few times a second to see what you're streaming and fingerprint you to sell your data
- reading your hdmi inputs to do the same thing while you have a game console plugged in
- Voice assistants: Bixby wake-word detectors listening and tracking what you say in your house
Before and after speed of my TV:
24 tasks done on day one with my Grok @bot.
I built Grok Board to run it: mail, calendar, tasks, Pages, and a context meter. Green to 50%, yellow to 65%, red past that.
Easy win: tell your bot to make a card every time a job starts.
Creating video game shaders and systems occurs at least 50x the speed with AI.
Building on @babylonjs.
Joined @NEWSMAX's Wake Up America Weekend yesterday to discuss AI filmmaking and @Gossip_Goblin's upcoming feature. People outside our AI circles are starting to pay attention and when the film drops Pandora’s box opens.
youtube.com/watch?v=0Dgln3l2…
People don't realize how ground breaking this is for filmmakers. In a matter of seconds, I had 3 new camera angles (hand placed), using only a single starting image. @theworldlabs cooked with Atlas. 🔥
🤖 Made with AI
Great short film by @theGioM! Some wonderful acting moments, worth the 4 minutes to watch.
"NO ONE ELSE" - An ordinary evening. A life coming apart. The truth, in reverse.
In this short psychological drama told backwards, the challenge is to make each chapter change how you feel about the one you’ve just watched.
A solo filmmaking project exploring intimacy, trust, and betrayal.
Much of the work was in the small decisions: a hesitation, a passing emotion, an interrupted line, and keeping track of who knows what, and when.
Let me know what you think!
Oh snap! 🔥Finally, some good news from MJ. Time to *roll up sleeves* and give it a test.
With my personalized style refs, I find MJ is still the most creative image model.
With editing built-in, should save me a ton of time, and skip the NBP/GPTImage step for adjustments.
Preview mode is here on Dreamina! 🥳
This is a big win for AI filmmakers, since the cost to explore and tinker was rather expensive with SD2.5.
New on Dreamina AI web: preview before you commit. A global first. 🎬
When you pick your model, choose the one tagged (for preview) to see your idea first:
‣ Preview: Generate and compare at 480p as many times as you like—at just 1/8 the cost of 1080p.
‣ Final: Pick your favorite result and generate it in 1080p. The output stays faithful to your preview frame.
Explore more of your video creation ideas affordably so every shot gets another take.
🕐 Available now across Southeast Asia, the Middle East, Africa, Europe, and South America. More countries and regions coming soon.
#DreaminaAI #DreaminaAIforPreview
dunno if I'm gunna make it to reset...😂
@OpenAI's Astra has been a huge unlock. gunna need a bigger compute budget apparently...
*cancels $200 anthropic sub* 🤫
epic film. take some time to watch Sam's 10 min masterpiece. 🔥
👀
characters, story, world.
lets see...
Fewer than 1 in 20 films submitted to Fairground make it to Amazon Prime, Roku, Xumo and LG.
So…what actually makes the cut for Fairground AI TV?
Since launch, one question has come up more than any other from the AI filmmaking community: what is Fairground looking to distribute?
To answer it, we asked Alexandra Mills, our Head of Production, to explain what makes the cut for Amazon, Roku, and our other platform partners.
Here's how the team evaluates content:
Story first.
"Plot, characters and the world are always top of the list," Alex says.
"A tight 5-minute film with a real story beats a beautifully rendered 20-minute film with a weak plot, every day of the week."
Length matters, but not how you'd think.
Content needs to be 3+ minutes. Longer generally makes distribution easier, but length never wins over quality. Focus on the story and the length that best fits. That's what matters.
Original IP Only.
Every submission through the Fairground Portal goes through copyright, music licensing and trademark reviews prior to distribution. Collectively, we need to ensure that all work is 100% original and is the very best of AI cinema. Make sure everything, including your music, is original.
Formatting is strict.
We distribute only 16:9 right now (no vertical), and we're aiming for PG-13 as the default, though we do have late-night slots that can carry TV-MA and above.
Genre shifts with the calendar.
We distribute across every major genre, but what we prioritize moves with tentpole moments and what our platform partners need.
Right now: horror for Halloween, and animated work for a separate opportunity.
And the thing that matters most: did we feel something?
"Did the combination of human artistry and generative AI tools produce a story worth telling?" Alex says.
"That's the real bar, and it should be every creator's North Star."
The decisions don't stop once content airs.
Fairground runs dynamic scheduling, constantly watching how content actually performs. Stream starts, watch time, overall session time.
Overperforming content earns a better slot. Underperforming content makes way for something new.
The lineup is never static, and neither is the bar we use to make decisions. As the technology improves, standards will increase over time.
If you are submitting new content for consideration, it may take 3-4 weeks for it to be reviewed for distribution, as submission volume has increased since launch.
We appreciate your patience and while it may take us time to review your film, we want to give it the time and consideration it deserves. We don’t want to rush something you put time, money and real work into.
We just wrapped our first month and the data has already surprised us. More on what we're learning (including a heat map of our top-performing programming windows) is coming soon.
For now: what other questions do you have? Drop your questions below, Alex will be reading and answering in the comments👇
love this
Everyone thinks the considerate thing to do is ask when someone seems to be hurting.
Not always.
Asking is for you.
You get to feel like you helped,
and the discomfort moves off you and onto them.
So he does the one thing she requested: nothing.
#dreaminacpp @dreamina_ai
amazing film by Jordan!
What makes something beautiful?
For centuries they said, "Beauty is in the eye of the beholder", meaning you get to decide what's beautiful, and no one can take that from you.
Now, they want to take that back. Now, they want to tell you what's beautiful, what is art, what is creative, and what is worthy.
The good news is...
Who cares what they think?
Who cares what they say?
They have no power.
Their words are dust.
You decide what's beautiful and moving, exciting and entertaining.
Art is yours to love or hate.
No amount of gold and worship placed at the feet of one piece of art should sway us to value it. And no screeching voices, or shrill cruelty demanding you surrender your opinions should steal it away.
•
Beautiful
A short film by Jordan Daniel Chesney
Produced by @capcutapp
Special thanks to @Orangedudull
Special thanks to @invideoOfficial
•
Hopefully this little film exemplifies that Seedance 2.5 have completely changed the game when it comes to performances. I have NEVER been able to get such perfectly dialed emotions, even with SD2. This was the first time I was actually moved by a performance.
I felt something real.
I STRONGLY RECOMMEND you head over to @capcutapp Creator Studio, select Seedance 2.5, and see what kind of performances you can bring to life. We have enough fight scenes where no one gets hurt, sweeping aerial shots, and dragons breathing fire to last us a lifetime.
Let's try something new because now... we can.













