Stripe ships 65KB llms.txt, but founder sites lack it, reveals asymmetry
Dry_Steak30 · reddit · 2026-09-01
The author tested /llms.txt deployment across 12 major tech companies. Eight, including PostHog (375KB) and Stripe (65KB), serve the file to help models understand their products. However, Anthropic and LangChain return 404, while OpenAI and Perplexity block access (403). Notably, while companies optimize for models, the founders' personal sites (Patrick Collison, Paul Graham) do not offer it. This highlights an asymmetry: social platforms (LinkedIn, Instagram) block AI crawlers, forcing models to rely on stale bios or old pages for personal data. The author questions if the link-list structure of llms.txt fits people (who need provenance) and whether adding new pages actually displaces outdated info in model outputs.
More from Companies & People
- Why Microduck is Winning: Cuteness Overload and Sim2real Tech — RachelVT42 · 2026-09-01
- Lee Cronin: AI Coding Dramatically Accelerates Scientific Idea Validation — PaulYacoubian · 2026-09-01
- Top AI trends for developers in 2026 and strategic thinking — rseroter · 2026-09-01
- User shares ChatGPT billing trick to reset weekly limits and avoid double tax — andybarriateth · 2026-09-01
- Polymarket: Sam Altman has 9% chance of leaving OpenAI CEO role before 2027 — Polymarket · 2026-09-01
- Deep Dive: The Rise and Fall of OpenClaw and the Evolution of the Agent Industry — 新智元 · 2026-09-01