then the footage is edited in a way to suggest a kind of narrative about the person and what's going on with them.

for all we know, this guy is thinking "wtf is this!?" but the comments are full of people projecting all kinds of stuff onto him like he's a character in a movie.

here was the prompt:

this time Sonnet 3.7 didn't really do what i asked, but it turned out to be a better way to illustrate things:

Claude was able to generate a list of examples, put them into a static HTML page, and i just made a screen recording with Cleanshot and inserted it into the edit

I'm creating a video about the constructionist view of language that says that language is made up of lexical chunks (not traditional grammar rules)