I intend on building a TikTok clone, and the way it will replicate the TikTok algorithm is to have a lightweight image recognition model and a speech-to-text model run locally when you upload a video, and generate kind 1985 (?) labels. Then it runs a local algorithm that chooses notes based on your interests. You don’t need to know about all the notes that exist, only all the kind 22 notes with associated kind 1985 labels

I’m going to finish my degree and prove you wrong.

I think you're missing the point. Nostr works through countless asynchronous websockets. There is not such thing as a true "global feed" at an architectural level.

That's okay, I think few are used to "global" algo ranked posts anyway. Maybe on Reddit or YouTube, but for example on yt I either view the subscription time based feed or the home "for me" feed. The global trending feed is never interesting but that may be just me. Whatever feed should probably be personalized which means only running algo on one's wot. Problem, I think is that's either self-host, subscription, or always running resource intensive app for a couple hops. There's wot relays, algo relays, personal haven relays. Be nice to get it wrapped up in a bun. X has probably proved though people are willing to pay for the experience if it's worth it.

Sounds similar to what nostr filter relay does. Sounds nice!