Dan Luu · 2026-10-24 · notable
Dan Luu — coding agents mean we should build more ambitious software
Dan Luu argues that coding agents make once-impractical projects, like a new regex engine or a Rust interpreter, take hours instead of a year. He says verification, not code writing, is now the bottleneck.
Projects that once took a year of expert work now take a few hours, so the old cost model for 'too big to build' is out of date.
What is it?
Dan Luu's essay 'We should be more ambitious with the software we build' takes its title from Michael Malis, the builder of pgrust, a Postgres rewrite. Luu argues that many developers still treat projects like a new compiler or a faster ripgrep fork as too costly, and that this assumption no longer holds when coding agents do most of the implementation.
How does it work?
The essay uses Luu's own side projects as evidence: a regex engine, a faster ripgrep fork and a Rust interpreter, tools he calls 'works for me' software that took a low single-digit number of hours of his time. Trust comes from testing, not from who wrote the code. pgrust is credible because its builder tests it heavily, and that testing has found real Postgres bugs; Malis has since moved from Kani to Lean to verify the harder parts.
Why does it matter?
Luu says the gap between what a model can implement and what it can verify keeps growing, so the useful skill is building self-checking setups such as fuzzing and property-based testing. An appendix gives Yossi Kreinin's counterpoint: most people should be less ambitious, because agent loops need clear pass/fail criteria and giant black boxes are hard to maintain.
Who is it for?
engineers deciding what to build with coding agents