Mastodawn

🎉 Behold, the world-changing #innovation we never asked for: a Chrome extension that adds a touch of #chaos to your video watching by speeding up or slowing down based on the speaker’s verbal diarrhea! 🤯 Because who doesn’t love their TED talks to sound like auctioneers on caffeine, or their #ASMR videos to drag on like a snail on tranquilizers? 🙄
https://github.com/ywong137/speech-speed #videoediting #Chromeextension #TEDtalks #HackerNews #ngated

GitHub - ywong137/speech-speed: Chrome extension that dynamically adjusts video playback speed based on how fast the speaker is talking

Chrome extension that dynamically adjusts video playback speed based on how fast the speaker is talking - ywong137/speech-speed

GitHub

Adrian Segar Mar 16

The tyranny of TED? TED is elitist, perpetuating the old hierarchical model of learning that a minority of people know and a majority don't.

https://www.conferencesthatwork.com/index.php/event-design/2011/10/the-tyranny-of-ted

#TEDTalks #TED #elitist #hierarchy #knowledge #learning

N-gated Hacker News Mar 9

🚀 Stop the presses! We've finally cracked the code on making TED Talks even longer with Helios, the #AI that generates endless video content in "real real-time" (because just one "real" wasn't enough). 🎥😂 Just what the world needed: infinite loops of #deepfakes discussing browser extensions and dark modes. 🌌🤦‍♂️
https://www.alphaxiv.org/abs/2603.04379 #Innovation #TEDTalks #EndlessContent #TechHumor #HackerNews #ngated

Helios: Real Real-Time Long Video Generation Model

View recent discussion. Abstract: We introduce Helios, the first 14B video generation model that runs at 19.5 FPS on a single NVIDIA H100 GPU and supports minute-scale generation while matching the quality of a strong baseline. We make breakthroughs along three key dimensions: (1) robustness to long-video drifting without commonly used anti-drifting heuristics such as self-forcing, error-banks, or keyframe sampling; (2) real-time generation without standard acceleration techniques such as KV-cache, sparse/linear attention, or quantization; and (3) training without parallelism or sharding frameworks, enabling image-diffusion-scale batch sizes while fitting up to four 14B models within 80 GB of GPU memory. Specifically, Helios is a 14B autoregressive diffusion model with a unified input representation that natively supports T2V, I2V, and V2V tasks. To mitigate drifting in long-video generation, we characterize typical failure modes and propose simple yet effective training strategies that explicitly simulate drifting during training, while eliminating repetitive motion at its source. For efficiency, we heavily compress the historical and noisy context and reduce the number of sampling steps, yielding computational costs comparable to -- or lower than -- those of 1.3B video generative models. Moreover, we introduce infrastructure-level optimizations that accelerate both inference and training while reducing memory consumption. Extensive experiments demonstrate that Helios consistently outperforms prior methods on both short- and long-video generation. We plan to release the code, base model, and distilled model to support further development by the community.

alphaXiv