3.4x their usual
46.4K viewsusually 13.3K
Comment “Codex” and I’ll send you my migration guide
Charlie Hills 🦩Transcript
217 words a minute
- 0:00So this is the same recording edited by two different AIs.
- 0:03Which one did better?
- 0:04Because both models opened the identical raw file and project,
- 0:07my palette, my typefaces, my written brief, and nothing swapped between them.
- 0:11All I asked for was a cut I could drop straight onto the feed without touching
Vs their usual views
3.4times
OutlierHigh confidence
46.4K views against a usual 13.3K for @charliehills.
- Views
- 46.4K
- 46% of followers
- Likes
- 605
- 1.3% of views
- Comments
- 30
- Length
- 0:40
- Outlier
- 3.4x
- their usual
- Posted
- Sep 10
- 21 days ago
How this is worked out
Caption
Comment “Codex” and I’ll send you my migration guide. I gave Fable 5.1 and GPT-6 Astra the same raw video, the same brief and my real YouTube editing process (the one I use for my own videos). Both had to cut the pauses, turn my words into visuals and time everything to my voice. Then I put their edits together (Fable above, Astra below). For this comparison, there’s one voice track. Astra’s panel has short holds totalling 2.2 seconds so the two edits stay aligned. And now I want your verdict. Which edit came out better? Comment “Codex” for the guide to bringing your Claude Code setup into Codex.