video

Eric Schmidt argues that long reasoning — AI models that can stay on a single problem for hours or days — is the one AI theme he would invest in, and that the market has not yet priced in how different this is from models that simply answer and stop.

作者分享了一个使用 AI 编程代理的工作流，并制作了视频辅助说明，建议读者将自己的文章内容喂给 Codex 或 Claude Code 让代理理解整个流程；他全程使用 Astra medium，消耗了大约 2-3 管 Pro 每周额度，认为用 Opus 5.5 效果可能更好。

SpaceX shares slow-motion videos of the Starship Flight 14 liftoff from Starbase, highlighting progress in aerospace engineering.

HN.watch is a website that provides the Hacker News front page with a short video explainer for every story, allowing users to watch video summaries of tech news posts.

A user describes how a video inspired them to test the limits of the Opus 5.5 AI model, following Donald's workflow and achieving impressive results.

A Twitter user announces plans to create a video explaining how they improved SEO for SuperX, seeking input on topics to cover.

This article announces an upcoming video that provides a comprehensive breakdown of the AI subscription top-up business, covering the business process to order acquisition, with detailed tutorials.

Remi Fabre shared a live-action video featuring the Microduck robot in episode 3, highlighting the robot's 'Oscar-winning performance' during filming.

OneStream Live 2.0 is a cloud-based platform that enables creators and businesses to create, schedule, and multistream live and pre-recorded videos to over 45 social media platforms and websites from a single dashboard.

Shared a video on content monetization, which is brimming with valuable tips.

An astronaut details the preparation for a spacewalk, emphasizing the role of NASA engineers and crew members in the process.

Glasp for Firefox is a tool that enables users to highlight and summarize any webpage, PDF, or video directly within the Firefox browser.

A tweet recommending a video on writing effective cold emails, with tips on subject lines, concise content, and clear calls to action.

A YouTube playlist featuring talks and slides from the 2026 EuroLLVM conference, a meeting for LLVM developers and enthusiasts.

A tweet by @pickover points to a video by user 'aila' featuring visual recipes for mathematical wonders, emphasizing curiosity and enlightenment in mathematics.

Google Gemini now supports sign language transcription from video, a notable accessibility advancement for AI models.

A Japanese court overturned a patent held by Red Digital Cinema related to RAW video compression, potentially affecting video technology licensing.

Introducing VGI-Bench, a multimodal benchmark that probes 12 distinct visual and audio-visual skills with 550 human-curated questions to expose failures in today's video benchmarks.

This paper introduces KVAE, a family of tokenizers for audio, image, and video designed for text-conditioned generative models, claiming competitive or superior reconstruction and generation quality compared to existing open-source tokenizers. The code and training details are publicly released.

ByteDance announced SeedRealtime, a native audio-visual model that can process continuous video, audio, and text while speaking in real time.