The problem with approximating an interest graph with a social graph is that social graphs have negative network effects that kick in at scale. Take a social network like Twitter: the one-way follow graph structure is well-suited to interest graph construction, but the problem is that you’re rarely interested in everything from any single person you follow. You may enjoy Gruber’s thoughts on Apple but not his Yankees tweets. Or my tweets on tech but not on film. And so on. You can try to use Twitter Lists, or mute or block certain people or topics, but it’s all a big hassle that few have the energy or will to tackle.
Eugene goes on to explain that TikTok's brilliance is being 100% about your interests and curating your feed for you. No need to manage human relationships and ensuing conflict that causes, or struggle to categorize yourself and admit to embarrassing interests. The algorithm will take care of everything, simply launch the app.