In this episode of the Slow Hunch, I spoke with Jay Graber, CEO of Bluesky. 

Originally conceived as an initiative within Twitter under Jack Dorsey, Bluesky was designed to transform Twitter from a closed platform to an open, protocol-based network. For those of you who caught the episode with Brad & Fred, this is a subject we got into in some detail: years ago, USV had an intuition that Twitter might be better designed as an open protocol, but at the time, the tools didn't exist to make it a practical reality. In 2019, Twitter launched Bluesky (and the underlying AT Protocol) as a research initiative, and that's when Jay joined to lead the effort. She subsequently lead the spin-out of Bluesky into an independent effort, developed the app and the protocol, and launched the network. Since launching publicly in early 2024, Bluesky has since grown to roughly 30 million users.

I've always liked the core design of ATproto and Bluesky, which lets each user configure core elements of their social experience, pulling from an open library of feeds, filters and client apps; all tied together with a single, user-owned identity (my identity on Bluesky is actually my DNS domain name, nickgrossman.xyz). In our conversation, Jay talks about why this architecture matters so much to her: primarily by giving users the ability to vote with their feet if the platform makes changes they don't like.

Jay is one of the clearest thinkers I know on the relationship between protocol design, system architecture and user experience, and it's been pretty incredible to watch her build Bluesky from a high level idea into a scaled consumer product over the last few years.