Yeah, it’s really important to avoid having it look like a lecture. You could also try learning animation yourself and seeing if you’re decent at it (or even really good at it), that would make it easier to collaborate with rationalanimations/@Writer since they’re bottlenecked (there’s also tons of great Yud and Scott Alexander stuff, and rationalanimations has been growing/sped up lately but even 5 videos per quarter won’t get nearly enough of them animated in time).
What I recommend is reading the CFAR handbook or tuning cognitive strategies and having this be one of the real-world challenges you test yourself against as you read and grow and test yourself, just like with the Sequences except you’ve already read them. Sequences-based optimization ideas are probably well worth pursuing on the margin, so don’t get intimidated, but the final winning idea makes you and many others win big so it’s worth it to iterate and try things and fail many times and learn many lessons until something incredible happens.
It’s a lot of work to learn to create animations and then do them for hours of content. Creating AI images with Dall-E 3, Midjourney v6, or SDXL and then animating them with RunwayML (which in my testing worked better than Pika or Stable Video Diffusion) could be an intermediate step. The quality is already high enough for AI images, but not for video without multiple tries (it should get a lot better in 2024).
Yeah, it’s really important to avoid having it look like a lecture. You could also try learning animation yourself and seeing if you’re decent at it (or even really good at it), that would make it easier to collaborate with rationalanimations/@Writer since they’re bottlenecked (there’s also tons of great Yud and Scott Alexander stuff, and rationalanimations has been growing/sped up lately but even 5 videos per quarter won’t get nearly enough of them animated in time).
There’s also galaxy-brained high-risk high-reward ideas like @dkirmani’s attention-based “zoomer-friendly” format of the sequences. The Sequences are great material to work with in general because all you have to do is something that Yud didn’t think of.
What I recommend is reading the CFAR handbook or tuning cognitive strategies and having this be one of the real-world challenges you test yourself against as you read and grow and test yourself, just like with the Sequences except you’ve already read them. Sequences-based optimization ideas are probably well worth pursuing on the margin, so don’t get intimidated, but the final winning idea makes you and many others win big so it’s worth it to iterate and try things and fail many times and learn many lessons until something incredible happens.
It’s a lot of work to learn to create animations and then do them for hours of content. Creating AI images with Dall-E 3, Midjourney v6, or SDXL and then animating them with RunwayML (which in my testing worked better than Pika or Stable Video Diffusion) could be an intermediate step. The quality is already high enough for AI images, but not for video without multiple tries (it should get a lot better in 2024).
For ayone trying to keep up with AI for film making, I recommend the youtube channel curious refuge https://www.youtube.com/channel/UClnFtyUEaxQOCd1s5NKYGFA