Why Some AI Workflows Stick—And Others Don’t

This issue closes out Every’s Think Week, which culminated yesterday in our first-ever conference, Thesis: 2027, at Pioneer Works in Brooklyn. Nearly 400 founders, builders, and experts joined us in person, and over 3,000 watched online. If you missed it, watch the recording, and look for Grant Harvey’s recap on Sunday.

Today, we’re republishing Katie Parrott’s piece from July on how to decide which AI workflows to retire or revisit. “Change is the only constant in AI,” Katie says, looking back on it. “Just because a solution doesn’t work for you today doesn’t mean it never will. A new one might come around that abstracts away the technical difficulty, or just fits into your workflow in a way the previous solution didn’t. The example in the piece is Tend: I thought I didn’t need it because I wasn’t busy enough, but now my Dot is doing much of that same stuff for me.” —Kate Lee

Was this newsletter forwarded to you? Sign up to get it in your inbox.

I was on a video call with Every CEO Dan Shipper, getting a tour of the increasingly elaborate ways he uses OpenAI’s Codex app to run his work life, when I saw the next AI workflow that was going to change mine. At least, that’s what I told myself.

Dan was sharing an early version of Tend, the Codex-native system he created to pull email, Slack, meeting notes, and company updates into one place and help you act on them.