Anthropic engineer: prompting now means goal, effort level, and a verification check
pvncher · x · 2026-10-07
Anthropic's bcherny argues prompting Claude no longer needs heavy scaffolding: state the goal, how much effort to spend, and how to verify the result — a big shift from the Sonnet 3.5 era when prompt wording mattered most. pvncher pushes back on "talk to Claude like a coworker": talking to coworkers well is itself a skill, and your new coworker is a Fields Medalist managing swarms of Fields Medalists.
More from coding & agent
- Agents love tidying up files nobody asked them to touch — JFPuget · 2026-10-08
- Open-source bridge turns a browser chat tab into an OpenAI-compatible API endpoint — harshanacz · 2026-10-08
- GraphRAG Bug: Deleted Documents Stay Indexed and Retrievable After Updates — JeremyCMorgan · 2026-10-08
- Paper: Vibe Coding Kills Open Source as AI-Recommended Repos Lose Stars — soumitrashukla9 · 2026-10-08
- Every publishes definitive guide to Compound Engineering, the philosophy behind 7k-star plugin — every · 2026-10-08
- Claude Haiku 5.5 matches GPT-6 Luna pricing but a stingier tokenizer hides a 1.25x cost hike — Simon Willison · 2026-10-08