❓questions i'm sitting with

not everything needs a conclusion. here are a few questions i've been chewing on. i'll update this when i get closer to answers.

• what does a "good" ai coach look like — one that makes you more capable, not more dependent?
• why do so few fitness apps feel like they were made by someone who actually trains?
• how do you write on the internet in a way that the algorithm doesn't warp into something you didn't mean?
• is there a version of "working in public" that doesn't slowly turn into performance?
• what's the smallest possible piece of software that changes someone's habit for a whole year?
• can an nlp model ever really understand a joke? (i don't think so. but i keep checking.)