Simon Willison on evals, going weird, and tools for Pulitzer-level journalism
TheTuringPost · x · 2026-11-22
Turing Post interviews Simon Willison on software engineering in the AI era. Key points:
- 1 hour now = 1 week before: the speedup has drastically changed his ambitions.
- Advice for builders: "go weird" — lean into unconventional approaches.
- Evals take center stage as a core engineering practice; Willison is among the first people to listen to on this.
- Next goal: build tools helping local reporters find stories in data they can't dig through themselves — "my goal is for someone to win a Pulitzer Prize using my software for perhaps 3% of what they did," ideally at a paper that can't afford developers.
Related event: Simon Willison: AI Turns a Week's Work Into an Hour(2 posts)→
More from AGI Musings
- Anthropic's opening bell marks the end of the age when only humans could think — r0ck3t23 · 2026-11-22
- 'Sending commands to Claude doesn't make you the author': pushback on AI lab math dumps — gerardsans · 2026-11-22
- Reddit asks: how do AGI believers hedge their finances and careers? — shameless_johnnyboy · 2026-11-22
- LangChain's "knowledge factory" essay sparks debate: better tools or training our replacements? — hwchase17 · 2026-11-22
- After two weeks in China, author argues Western AI safety playbook doesn't transfer — and catching up is getting costlier — JacquesThibs · 2026-11-22
- Researchers ask whether a huge Lean proof could hide bugs in the proof assistant itself — LucaAmb · 2026-11-22