John Carl - e/acc retweeted
i've found that nothing makes AI videos feel more real than using actual locations that exist in real life
so I created my own location scout as a claude skill with built in access to google street view... now I get detailed reference photos of ANY location on autopilot 😮💨
John Carl - e/acc retweeted
lmao
Opus is happy to assist Deepseek with reverse engineering! it just wont do it itsself
John Carl - e/acc retweeted
we should be deeply skeptical of “safety and alignment” rhetoric from any org that’s in the business of building frontier models
akin to taking health advice from Big Tobacco
John Carl - e/acc retweeted
If your AI characters sound fake, the fix is almost always in how you write and DIRECT the dialogue in your prompt.
I recently started on the dialogue-heavy scenes for my AI feature film about Beethoven, and too many of the lines kept coming out wrong.
They’d sound too theatrical, too monotone, or too much like a practice read-through between actors.
(all this despite the fact that Seedance 2.5 is fantastic when it comes to realism.)
So through a whole lot of trial and error, I've found that these 5 things made the biggest impact on the quality of the dialogue delivery in my generations:
1. Paying close attention to the written punctuation used.
2. Breaking long lines into shorter ones, with intentional pauses.
3. Using bodily actions as punctuation (this one gave the biggest improvement BY FAR! In the attached clip, note the sniffles in between Beethoven's lines. That alone made the delivery feel much more natural.)
4. Using "anchor gestures" (he lifts his finger as if to say "one sec")
5. Just split your generations in two (stupid simple, but it really makes a difference.)
So if your characters keep sounding like they're reading from a teleprompter, give this a try!
Now you can build any scene in Blender from a single reference image.
I’ve open-sourced my tool AnyScene: drop in one or more references (plus optional text) and an agent builds the whole set inside Blender.
Production-ready assets in a production ready environment, placed, lit and color-graded, with every asset broken out as its own file.
github.com/apekshik/anyscene
Today we’re launching the first AI modeling agency 💅 right inside Glif.
Cast hundreds of real AI models, drop your product, idea or brief and let Glif create super compelling content for you end to end.
If you still think AI models aren’t ready for real work, watch this 👇
🤖 Made with AI
John Carl - e/acc retweeted
This is f*cking insane. Elon Musk says Grok 5 could reach AGI:
"Grok 5 will be the smartest AI in the world by a significant margin on every metric without exception"
Then he dropped the number that matters:
"I calculate like 10%. That's what my biological neural network comes up with"
He's never put AGI above zero before.
Training already started. Release sometime in Q1.
So what changed? Size and senses.
Grok 5 is a 6 trillion parameter model
Grok 3 and 4 run on 3 trillion
Text, images, video and audio in one model. It builds its own tools to answer you
His bar for AGI: understanding real-time video like a human does
And there's secret sauce he wouldn't share on stage:
"It's really going to feel sentient"
5 minutes. Elon walks through what xAI is training right now and why this one is different
Bookmark and watch it today
Then read the article below ↓
John Carl - e/acc retweeted
Canva is building its own AI models for creating images and editing existing designs!
Last week, Canva's model Craft debuted at #4 on DesignArena's image-gen leaderboard.
Now comes Brush, a vision-language model focused on design editing.
Canva says Brush can restyle existing designs, adapt them to different formats and apply changes consistently across multiple pages.
🤝 Paid partnership
Wow, think of the educational opportunities here. Incredible. The ultimate homework helper.
John Carl - e/acc retweeted
This app is great! Just re-processed a few recent raw photos shot on iPhone 18 Pro
So I recently badgered my cinematographer @steveyedlin into making my ideal iPhone still camera app. It shoots RAW, bypasses the iPhone's look processing and instead uses the same image processing we use on our movies. apps.apple.com/app/apple-sto…
John Carl - e/acc retweeted
People want AI to automate everything
except the one thing
they personally get paid to do.
Meet NVIDIA Nemotron 3.5 Transcribe 🎙️
An open-weight, promptable 30B MoE model with ~4B active parameters that brings transcription in 40 languages, speaker attribution, translation, contextual prompting and audio understanding into one checkpoint.
Watch it keep track of who said what when everyone’s talking at once.
1/ 🧵
John Carl - e/acc retweeted
Every video is now open-source.
With Tesseract, Claude and ChatGPT can use any video as a reference to create a new, editable project around your idea. Paste a link or upload a video, then tell your agent what you’d do differently.
Keep refining with your agent, or take your project into After Effects or Premiere Pro.
To prove our point, we remixed a few classics.
Bro, whaaat!
Introducing Reality-Aware Rendering in @Simulon, a completely new way to create with the real world.
Place neural assets with your phone and act out the scene. When you stop recording, Simulon renders the take, reading everything in it (what you do, what you say and the place around you), and your assets respond to all of it.
The final is regenerated from your recording: a new performance of the scene, not a copy of the footage.
Available now on iPhone. Coming to Android soon.
John Carl - e/acc retweeted
A fantastic thing for high-resolution H3 vids.
NVIDIA dropped Open-H3-Regenerate-2K - it gives anyone local control over high-resolution video upscaling.
> Takes a cheap low-res draft from any model.
> Enlarges the latent directly in three steps to 4K.
> Beats the closed API in user tests.
nvlabs.github.io/Sana/Open-H…
🚀 Open-H3-Regenerate-2K: 3-Step Video Regeneration, Native Output up to 4K
📝 Blog: nvlabs.github.io/Sana/Open-H…
💻 Code: github.com/NVlabs/Sana/tree/…
🤗 HF Weights: huggingface.co/Efficient-Lar…






























