@adamconover.net is here though, and also worth checking out — not as focused on tech and policy, more on fact-finding journalism on everyday topics with a humouristic bent (e.g., "Adam ruins everything").

I don't think Cory is on Bluesky, but his website is pluralistic.net. Highly recommend giving his writings a read (though be warned — he writes a *lot*).

My sense from talking to a few folks is that Claude Code is currently the leader of the pack, so was planning to start there

This will be my first foray into agentic AI-based coding, so you'll get my unfiltered first-and-second impressions 😅 And the opportunity to point out how wrong I've been all along if it turns out to be a humbling experience 👀 Ultimately, I'm aiming for a reality-check for my skepticism!

Excellent! Now we just need the same for rustc lints 👀

That helps, for sure. But still requires that a) they all come with automated fixes, b) those fixes are generally correct, and c) teams prioritize reviewing those. It's also generally inefficient to be forced to review reactively, than decide when to bulk-address.

Oh, for individual projects this isn't *that* bad. The issue arises when you have this across all of a company's projects. They all have other things to prioritize, and end up simply running with older Rust versions instead. Which is then a security risk. And then their dependencies fall behind...

Challenge with this is that it means I'm blocked from upgrading my Rust version until I've addressed the lints. Which also may mean I'm blocked from updating dependencies that make use of newer Rust. The current system soft-forces that the Rust upgrade and lint fixes must happen jointly and first

I want this so badly. I remember having a chat with Niko and Felix about it ages ago, and I think their main concern was that it'll end up being approximate at best (e.g., if the semantics of a liny changes). But I think that'd still be a huge net positive.

Or is that need perhaps sufficiently covered by ChatGPT and friends?

Sadly not! I wish 😅