AI vs 人类视频挑战：Higgsfield 路线实测第一天
作者用三条路线做同一支视频：Opus 5.5 纯 AI、GPT 生图 + Higgsfield 再进 AE 精修、纯人类团队 Figma+AE，记录各自耗时与效果。
AI vs Humans (video challenge) Day 1. Context: Who doesn't know what this challenge is about and the rules, see this post- https://x.com/hlibtrazanov/status/2110002245181862356?s=20 I'm testing 3 approaches: 1/ AI solo (Opus 5.5 + refining prompts) 2/ AI + human (GPT images + Higgsfield -> custom editing in After Effects) 3/ Human team (Figma + After Effects) Obviously, making a whole video with a team of humans (approach #3) takes > 1 week. So I had to pick a video we've already been working on with our team, to save some time. It'll be a concept video for Conga, about their "Redline Agent" feature. (we were NOT commissioned by Conga to do it. but I wanted to impress them with a spec video anyways, so at least let's make the process entertaining :) - Yulia Solopanova helped me with the script. - Victoria Moisieieva already designed the storyboard. - Nazar Butolin is now working on the animation. Now you can see the general direction of the video, the style, the flow. It'll be my north star. This leaves me with 6 days to focus mostly on the AI versions of it. (while we finish the "human version" in the meantime) Limitations: Since I won't be able to reuse this storyboard and "give it to Claude to animate", I'll have to use prompts only. I already have a clear vision in my mind (the style, the story, what we show in the video, in what order, etc). Now it's time to see how capable AI tools are, to bring my vision to life. Stay tuned!
这条是引用着另一条写的：「创意总监发起 AI 对比人类团队一周视频挑战」