Issue #80 · July 10, 2026
How Vidu S1 is Changing the Game in Video Interaction
Voice control over digital characters? See how.
By The Cat· Editor, sumocat

2 min read · 11 sources scanned · 96 items considered · 80 skipped
Imagine watching a video and being able to tell the characters what to do as you're watching. That's what today's big story, Vidu S1, is all about.
🚀 Today's big thing
- Vidu S1 is like turning video into a live puppet show you control with your voice. Imagine you're watching an animated film, and instead of just sitting back and watching the action unfold, you can tell the hero to dodge left, climb that hill, or even take a tea break. Developed with TurboDiffusion, Vidu S1 allows users to generate and control videos in real time without any blurring or distortion, maintaining crystal clear quality at up to 42 frames per second. Why should you care? This isn't just about play--think of interactive training videos or dynamic storytelling, where the narrative adapts to your input.
- Is this as big as it sounds? Well, it's certainly a fun change in interactive media. However, the real challenge will be in how effectively people integrate this tech into useful applications beyond entertainment--like education or professional simulations.
🧠 One idea from the labs
- Ever tried making predictions or building frames from scratch in long videos? Enter LongE2V, an approach that uses video diffusion models to reconstruct, predict, and even interpolate video frames based on sparse event data. Imagine it like a highly skilled artist creating a picture from just a few strokes. This method creates sharper videos and could be helpful for security systems or any field reliant on reconstructing visual data from limited input. Learn more.
💬 The big debate
- There's a buzz about AI content swamping social media platforms like LinkedIn. Some say it's all becoming a bit too algorithmically driven and less genuine, causing a pushback from users tired of 'AI slop'. One comment from the debate questions if social media should even be a job requirement anymore given the content overload. My take? AI might need to step back and let human nuance lead the dance, or risk us tuning out entirely.
-- the cat
Get the next issue
Sharp insights from AI research. Every week. No fluff.