三家 AI 工作室同题做视频与图文资产对比

Branko Trcek @Brankotrcek原推发布于 2026.10.04

在 X 上打开原推

DOTS CONTENT STUDIO turned one presentation into 12 assets across two releases. The second package passed 6/6. Three studios got the same assignment: Dots, Grok Bot, and a custom Sonnet 5.5 setup. Same sources, same deliverables, same review rules. Research → Write → Produce → Review → Approve My Dot acts as Chief of Staff: defines the tasks, tracks dependencies, and collects the current files. The source packet supplies the recording, transcript, article, brand rules, and editorial feedback. I keep control of publishing, external messages, and spending. Each release contains four posts, one video, and one cover. The manifest records which versions are ready and what still needs fixing. Grok Bot took the opening round: 5/6 assets accepted without edits, against 4/6 for Dots and 3/6 for Sonnet. Then I changed the fact sheet, revised the opening style, and gave all three studios the same corrections. The second release flipped the result: Dots returned 6/6 ready for approval. Sonnet delivered 5/6. Grok Bot delivered 4/6. Dots completed the correction round in 14 minutes. Sonnet took 17. Grok Bot took 23. Across both releases, Dots needed 18 minutes of my attention. Grok Bot needed 43. Sonnet needed 31. That includes reviewing, editing, and directing the work. Every studio still needed my approval. The handoffs decided which one I kept: clear inputs, specific corrections, and one current package instead of scattered versions. The complete Dots setup, reusable prompts, and three-studio comparison are in my article below ↓

这条是引用着另一条写的：x.com/solomonrojie/status/2105713134174605327