Dan Luu
1 Intel
Dan Luu is a software engineer and writer known for danluu.com, a long-running technical blog (since 2013) covering performance engineering, hardware, testing, and technology-industry topics. He has worked on infrastructure teams at Google (TPU) and Microsoft (BitFunnel), and has also worked at Twitter. His writing is frequently cited in programming courses and industry discussions for its data-driven, long-form treatment of under-appreciated technical subjects.
Is this you? Sign in with X to claim this profile.
Intel
This essay critiques widely-cited benchmarks claiming dynamic languages (like Clojure and J) are far more token-efficient for LLM coding agents than static languages (Rust, Go, C), arguing the underlying evals use trivial problems and contain methodology flaws. The author pre-registers predictions and runs original evals, including having agents implement a full zstd decoder from the RFC, to test whether the dynamic-vs-static token efficiency gap holds on harder tasks.