Arsaze is an AI-native video editor built around a real multitrack timeline that people can edit manually or control through connected AI assistants such as Claude, ChatGPT, Gemini, Cursor, and others. Its central idea is to make video edits visible, reviewable, and reversible: an agent’s changes play back live, are automatically versioned, and can be undone or explored on a branch. The editor also offers Tab, an autocomplete feature that previews a likely next edit - such as removing similar pauses or matching a color grade - and applies it only when accepted. Tab adapts to editing choices, with an option to disable that learning.
Editing can also happen through a transcript: deleting words removes them from the video, while cuts appear as tracked changes that can be restored. Clients can review cuts through a link, draw on specific frames, and leave time-stamped notes that an AI can act on. Other tools include transcription and captions, silence and filler-word removal, color correction, audio cleanup, motion graphics, background removal, and AI-generated media. Users can export MP4s or transfer projects to DaVinci Resolve or Premiere Pro via FCPXML. Arsaze has a free plan and runs in browsers or desktop apps for Windows, Mac, and Linux; desktop local projects keep source media on the user’s computer, though editing requires an internet connection. The company says uploads are not used to train models without consent.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
Elon Musk and Mark Zuckerberg used President Trump’s preferred term “super intelligence” on their companies’ earnings calls, following his order to replace “artificial intelligence” in federal agencies. Microsoft, Amazon, and Alphabet leaders continued to use “AI.” (cnbc.com)
Democratic lawmakers led by Greg Casar and Maxine Waters introduced a resolution opposing taxpayer bailouts for companies and executives if an AI-driven market bubble collapses. It calls for federal support to protect workers, retirees, and low-income communities from an economic downturn. (progressives.house.gov)
Companies that cut jobs citing AI are beginning to recruit some former employees again. Returning workers should weigh the role, pay, and whether the reasons for the earlier cuts have changed. (apnews.com)
![Fiber lasts decades, GPUs last three years: the AI boom isn't the dot-com boom [video]](https://hntoday.pixelshot.io/screenshots/20261030/6123fca6c1e94075_tn.jpeg)
The title argues that AI infrastructure differs from dot-com-era investments because fiber lasts for decades while GPUs become outdated in about three years. The provided content contains no video transcript or further details. (youtube.com)
Anthropic added a Claude Code setting that disables the last approval prompts for risky commands like deleting files. The company warns through the setting’s name, but removing the safeguard raises the risk of agents causing data loss. (theregister.com)
Spawn’s ten-person team says it shipped 77,000 substantial changes in a month by relying on permanent AI agents, spending $3.1 million on tokens. The company uses agents to own areas such as engineering, finance, and bug triage while humans guide larger projects. (twitter.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.