Parth Chopra

@parthc28.bsky.social

Mainly talking about cool atproto projects @baincapitalcrypto.com. Prev product leader and founder.

How do people think about new verifications mechanisms in a way that don’t complicate the protocol? E.g. automated verification that can be verified completely using cryptography, lets say about emails Would a good way be: 1. Use the current mechanism with an automated user as the verifier 1/2

I'm launching Comind, an open source project to develop a cognitive layer for #ATProto. Comind is intended to build a network of AI agents that collaboratively process information flowing through AT Protocol. I'm looking for contributors interested in the project. github.com/cpfiffer/com...

GitHub - cpfiffer/comind: Open-source components for Comind, the cognitive layer for the web.

Open-source components for Comind, the cognitive layer for the web. - cpfiffer/comind

github.com

atproto thing i have gained some conviction on this weekend: educating users on how this mess works is hard enough without us all implementing our own (pre-oauth) login flows we should standardize our logins and make them all look the same. share components if at all possible. same UX.

I've always thought a client like this would be cool, some things I'd like to see: - microblogging still the foundation - more integration with live scores and big plays - self-serve visual flair for which teams you support - domains for identity, especially for large publications

My parents got a Tesla recently and the FSD is SO incredible During the holidays I went door-to-door in complex city routes with no intervention. And then it parked itself! Truly magical experience

Cool checking out @graze.social -- ability to generate, share, (one day monetize?) your own type of feed. I wonder when we'll have dynamic, prompted LLM-style feeds. Aka say "make me smarter, show me things that are funnier" and the feed will be curated as such