Joshua, [Jan 1, 2024 at 5:13:09 PM]:

I find recording with the GoPro to be much more fun than watching the footage. Once you start recording a lot of video, it becomes painfully obvious how uninteresting most of it is.

Still, there can be value in capturing some things.

For example, if you’re on a nature walk, and you want to be able to reference plants you saw.

We’re getting to a point where you can dump the video into a GPT and have it automatically index and describe your video.

I’ll test this because it would be awesome to have a searchable index of both the speech and visuals.

Shouldn’t be too hard with ffmpeg and some simple API integrations.

Reminds me of PNNL’s IN-SPIRE tool. They did that so many years ago…

AWS does it as a service too

What’s the service called? It looks like Rekognition is the main one.

I guess that would help. But, with ChatGPT, I think it could describe scenes in a lot more detail.

And then you could summarize the transitions between scenes very succinctly.