Berkeleychat

here are my unstructed notes since i'm running out of time:

social media was created because there’s something that people instinctually crave but takes effort to get, and that’s learning about other people. social media made this much easier

but still we can do better. you only get a superficial view into people’s lives from their profiles, and people don’t really use social media to strike up conversations with people they’ve never talked to

what makes this work, is that the personal nonsuperficial data is analyzed and presented to other users in an unintrusive manner it's something that aggregates and presents meaningful data to optimize for group outcomes while preserving privacy

here's the ultimate vision: someone’s going to build the personal universal assistant, but i’m not sure if anyone will build the communal universal assistant if i don’t. relationships, group dynamics, societal relations are the most important context. a social central planner—something that maximizes personal serendipity and social clarity, optimizes group outcomes, and improves societal relations—would improve culture

the reason why current algorithms seem not to provide long-term value is because they're not really complex enough to do so. the only algorithms that can hold a user's attention are those optimized by AI, but the only thing current algorithms can really optimize for is short-term engagement.

in the future, i think we'll be able to optimize for more complex forms for benefit for the user, beyond just short-term engagement.

this might sound idealistic, but i do think you can extract more value when you provide more value.

Built With

• grpc
• keydb
• protobuf
• redis
• swift
• swiftui