AI Hasn’t Crossed the Chasm to Teams: Inside Superhuman’s Bet on Collaborative Agents

AI made you faster. Did it make your team faster? Almost every knowledge worker says yes to the first question and goes quiet on the second — because the productivity is pooling in individual chat windows and never reaching the people around them. In this episode of Talking AI, Matt Paige sits down with Lane Shackleton, Head of Superhuman Docs, who spent more than a decade building Coda before it was rebuilt as Superhuman Docs. Lane argues the problem is structural: the chat window was designed as a soloist tool, collaboration software was designed for teams, and the two have never really unified. His fix is what he calls the last mile of AI — agents that come to you inside the work you are already doing, rather than waiting for you to open a chat box. They get into what shared team context changes about agent behavior, why memory only you can see isn’t team memory, the July 8 launch wave that turned Coda into Superhuman Docs, how product, design, and engineering roles are collapsing into each other, where the bottleneck moved once code got cheap to write, and how you work with the same labs you compete against. --- Key Moments: 07:56 — Why having to think “it’s time to use AI” is a failure mode 11:20 — The context problem: copy/paste and memory that vanishes 15:01 — Coda to Grammarly to Superhuman 17:58 — Why July 8 was a big bang, not a phased rollout 18:59 — AI Views: prompt your way on top of a dataset 24:05 — Product, design, and engineering roles collapsing 27:52 — Where the bottleneck moved: code review 30:17 — Partner or threat? Working with the labs you compete with 37:24 — Let the makers make, and why play beats mandates 43:31 — The end-of-day sweeper --- Key Links: Superhuman: https://superhuman.com/ Connect with Lane on LinkedIn: https://www.linkedin.com/in/laneshackleton/