Why Lisp Thrives In The Age Of AI (with Cameron Barré)

Earlier this year it looked like AI might kill programming language innovation. Coding agents were at their best in the mainstream languages with decades of training data, which made the barrier to adopting anything new higher than ever. That's changed. With the right skills files, an agent will write Haskell or Gleam as happily as Java, and that flips the situation on its head: it's never been cheaper to take a new language, or a new architecture, out for a test drive. All we need is ideas worth trying. So this week I'm joined by an enthusiastic proponent of a language, an architecture and a library you might want to try. Cameron Barré is a long-time Clojure programmer and the author of Grain, an open-source library for building event-sourced information systems. We start with what makes a system simple or complex, and the difference between essential and accidental complexity. Then we look at why Cameron thinks Lisp is an excellent choice for expressing both code and constraints, and why a live REPL might matter even more to an agent than it does to a human. From there we get into event modeling, an architecture so simple to understand that even a clanker can get it right, and how Grain uses Clojure macros to turn a handful of primitives (commands, events, views and todo processors) into a DSL that stops agents drifting from the architecture. If you've been looking for an excuse to revisit Lisp or event sourcing, or you're wondering what a language should give us now that agents write most of the code, this is a tour of a different way of thinking: power where you need it, constraints where you don't, and cheap enough to try for yourself. --- Support Developer Voices on Patreon: https://patreon.com/DeveloperVoices Support Developer Voices on YouTube: https://www.youtube.com/@DeveloperVoices/join Grain: https://github.com/ObneyAI/grain Cameron's Clojure/Conj 2026 Talk - Building Simple Systems in the World of Natural Language Programming: https://www.youtube.com/watch?v=YKFxDDtfJVI Obney.ai: https://obney.ai/ Event Modeling: https://eventmodeling.org/ Allium: https://juxt.github.io/allium/ What's Worth Knowing In AI Right Now? (with Henry Garner): https://www.youtube.com/watch?v=43lsrgkhJ8c Clojure: https://clojure.org/ Out of the Tar Pit (paper): https://curtclifton.net/papers/MoseleyMarks06a.pdf Rich Hickey Fanclub: https://github.com/tallesl/Rich-Hickey-fanclub Design, Composition and Performance (Rich Hickey): https://www.infoq.com/presentations/Design-Composition-Performance/ Calva: https://calva.io/ Datomic: https://www.datomic.com/ Nubank: https://nubank.com.br/ ClojureScript: https://clojurescript.org/ ClojureDart: https://github.com/Tensegritics/ClojureDart Babashka: https://babashka.org/ Jank: https://jank-lang.org/ Jolt: https://jolt-lang.net/ Glojure: https://github.com/glojurelang/glojure Clojure Test Suite: https://github.com/jank-lang/clojure-test-suite Polylith: https://github.com/polyfy/polylith Datastar: https://data-star.dev/ htmx: https://htmx.org/ Playwright: https://playwright.dev/ Code as Agent Harness (paper): https://arxiv.org/abs/2605.18747 Recursive Language Models (paper): https://arxiv.org/abs/2512.24601 EuroClojure 2027: https://2027.euroclojure.org/ Kris on Bluesky: https://bsky.app/profile/krisajenkins.bsky.social Kris on Mastodon: https://mastodon.social/@krisajenkins Kris on LinkedIn: https://www.linkedin.com/in/krisjenkins/