Artist: Analog and AI | Video | Sound |Drawing | Animation | Fun My free audio conversion app. No ads. No sign-ups. No BS. 👇 https://nitter.gq/t.co/ex1l3bNlku
St Louis, MO
Joined June 2009
- Tweets117K
- Following1.3K
- Followers2.4K
- Likes134K
Pinned Tweet
Made a bot that sniffs out X scams 🕵️
Paste any post link and it tells you straight whether it's a scam, engagement bait, or legit, with the red flags spelled out.
Free to grab: x.ai/bot/s0IwjnUv7BX_bI8Q4UR…
Pruna took MiniMax's H3 video model and turned it into an open image model you can actually run.
8 steps instead of 50, and it works in ComfyUI. I teach video art.
Free tools my students can run at home matter more to me than another closed app.
Anyone tried it yet?
Today, we open-source P-H3-Image, an image generation and editing model built from MiniMax-H3 by @MiniMax_AI. A SOTA video model, turned into an efficient open image model.
- 𝗛𝗼𝘄 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝘁 𝗶𝘀 𝗶𝘁? Generate or edit images 5× faster, in 8 steps instead of 50. A 4 MP image takes 2.6 s instead of 12.4 s. The model is 7× smaller than H3, so high-resolution images are a light download and run on a much smaller GPU.
- 𝗛𝗼𝘄 𝗴𝗼𝗼𝗱 𝗶𝘀 𝗶𝘁? Strongest at photorealism, documentary and editorial photography, film stills, and graphic styles with short quoted text. It renders 2 to 16 MP and edits with up to 8 input images.
-𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸? We reworked every component of H3: the prompt upsampler, the text encoder, the diffusion schedule, the transformer, and the decoder. Use it with Diffusers or ComfyUI.
📄 Tech Report: pruna.ai/blog/oss-p-h3-image…
🤗 @huggingface Open Weights: huggingface.co/PrunaAI/P-H3-…
👉 @ComfyUI Native Support: huggingface.co/PrunaAI/P-H3-…
I drew a guy on fire and then made the sound of being on fire on my modular synth.
Watch for the hand at 0:27.
What does this sound like to you?"
A 66-second short: fox, paper birds, clouds, score, sound design.
Every frame from code. No image model. One HTML file.
That's closer to how I think about video art than a polished AI clip is.
Would you rather watch something perfect, or something that could only be made this way?
🚨 Fable 5.5 just made a 66 sec animated short film, with zero image models
every frame + every sound = python it wrote itself
fox, paper birds, clouds, lettering, score, sfx, all of it
> 1920×804 · 24fps · single html file
> no gen images / no stock / no samples / no font files
> ~2h40m of autonomous work
< not ghibli level, and it told me so itself lol >
29% of my Grok Bot usage left, 4 days to go.
On that 71% I built a researched hook playbook, a public bot template, and a week of posts.
So when people say it burns out in a day… what's eating yours: big research runs or bots running all day?
From 1996 to 2000 I made fake TV commercials.
Now Seedance 2.5 makes a 30 second clip with sound from one prompt, and people say it looks like a $200 million movie.
The polish is free now.
The weird idea still isn't.
What would you make a fake ad for?
This is my face after one prompt used up my whole day of Grok Bot.
I drew it while waiting for my usage to reset.
But here's the twist: I'm still not canceling. The bot's that good, and I'm that stubborn.
What's the most usage you've burned on one task?
I teach video editing.
My beginner students freeze the first time they open Premiere.
Pieter just had Claude Code cut 58 minutes of clips into a 10 minute vlog.
The cutting is getting easy.
Knowing what's worth keeping is still the job.
Teachers and editors, are we teaching taste now, or still teaching buttons?
My vlog today was fully edited by Claude Code with Opus 5.5!
I copied the video clips from my DJI to a folder, then opened Claude Code in it and asked it to edit it into a 10 minute vlog
I made a few adjustments but not a lot
It transcribed everything I said with mlx-whisper (Whisper large-v3-turbo running on my MacBook Pro's GPU through Apple's MLX)
It pulled video frames every few seconds into grid images (contact sheets), so Claude could see what's in each clip
It used SpeechBrain ECAPA (a voice-matching model) to figure out what was my voice from other people
It then made thumbnails with OpenCV with YuNet's face detector
The edit itself is a list of cuts in a JSON file: which clip, start time, end time, Claude then picked the pieces by reading the transcripts and looking at the frame grids by itself
I added on screen captions for X and on YouTube for the first minute with Pillow (Python's image library)
P.S. I didn't ask for any of this, and I didn't know any of this until @thesamparr just asked me how I edited it
All I asked for was "Claude can you edit these clips into a 10 minute vlog for X and YouTube"
I think if I didn't have AI to edit it, I wouldn't be vlogging because editing would be such a painful and slow process for me and hiring someone to do it would be even more hassle for me!
BREAKING: X users stay on the site longer than anyone else.
Scientists are calling it the "one more scroll" effect.
But here's what the data doesn't show: 40 of those minutes were me trying to find a post I saw yesterday.
Proud to be part of the statistic. 🫡
That "here are my 4 tools" post only listed 3. That's not an accident.
It's a gray-area marketing play.
Here's how it works, step by step:
Free tease, paid sauce.
A smart-sounding theory that's missing the one step that makes it work.
The missing item. Replies ask "what's the 4th?" and the answer comes back tagging a product.
No #ad. Paid? Affiliate? Just a fan? The blur is the point.
Perfect setup replies. Some are real, some might be planted. Nobody can prove which.
Here's the part that matters: none of it breaks the rules on its own.
The pattern only shows when you step back.
My rule as a teacher: if I share it, I share all of it. No missing step, no product hiding in the replies.
What's a post that sent you hunting for the missing piece?
514 years ago today Michelangelo unveiled the Sistine Chapel ceiling.
He said he was a sculptor, not a painter. Did it anyway.
Four years, head tilted up, paint in his face.
I tell my students the medium you're scared of is usually the one you need.
What medium did you get pushed into?
Made a Gol Bot that turns any topic into an explainer video: script, storyboard, voiceover, animatic, thumbnails.
x.ai/bot/09_7munGcsQnJZYmiJW…
Drew a pumpkin in 5 seconds, then wrote a soundtrack that sounds like a haunted toy piano falling down the stairs. Happy Halloween 🎃
"X's payout program changed in September. Most people still don't know the new bar is 500,000 verified impressions in 90 days. So I built a bot that researches your niche and writes your hooks to hit it."
It works because it opens on news people missed, gives a real number they can't ignore, and then hands them the fix.
x.ai/bot/RExHClSPSldb-uIq7SN…
I teach video and sound art. I grade the process, not just the final file.
Fluxus artists built whole careers on instructions and the doing.
Bob Ross was basically a Fluxus score with a happy little tree at the end.
What's a piece you loved watching?
I make musique concrete. Real sounds, recorded on purpose, chopped up on tape.
ElevenLabs SFX v3 can type out a box dropping on a couch now.
I still want the take where my mic stand fell over.
Sound designers, would you trust a prompt over a field recording?
Big scenes rarely come down to one sound, so we built SFX v3 to handle more than ever.
Our latest model lets you create richer, more precise sound effects from a single text prompt, with improved music and vocals.
Now available inside @ElevenCreative and ElevenAPI.
Made a free bot that designs your X banner from your own art, and it survives the iPhone crop, the Dynamic Island, and Android punch holes.
Tell it what you want in plain words.
Bot: x.ai/bot/4T6Thj34oUhb4fUurFd…
Toolkit: github.com/brflux/x-banner-k…
