I get the impression, especially these last few months, that Klabnik is undertaking a heavily couched form of political speech. He appears to be deeply uncomfortable with anyone who would question the cloud-based, de facto centralized and commercially enclosed development and usage patterns that have built his career, and that informed ATProto's heavyweight participation requirements as well as the entire AI-startup market. But rather than opening himself to that broader conversation, he instead attempts to manipulate these discussions by pathologizing individuals with whom he's having a disagreement, on top of throwing in questionable arguments from authority.

Consider, for instance, lecturing quips like "Because you don't always get to pick when you ship" or "They did extensive research on what was out there..." etc. from the lobsters thread. No room to discuss the confused product/protocol approach that Bluesky is taking, or that the "move fast break things" logic that one might apply to a web app does not really fit developing and documenting a broader protocol, if that's what Bluesky is really trying to create. No room to question that this approach may have harmfully skewed the Bluesky team's thinking about what they really need. It's an echo of A16Z's intentionally braindead "just build things" rhetoric, where the "builders" (formerly "founders") are above reproach.

That is an incredibly weaselly way of approaching public dialogue, and has strongly impacted my trust in him for the worse. I think he probably feels a strong need to be perceived as level-headed and civil ("Quiet, Calm, and Sincere" as Something Awful's community-complains subforum used to be called), as long experience leads a lot of internet denizens to fear discussion forums dying in a flamewar spiral, but uhhh... he's not meeting that bar.

It certainly feels like there's been a concerted push to concern-troll about anti-AI sentiment on lobste.rs. Every day I recalibrate if it's worth keeping my account there.