One-line prompt fix for sycophancy: make the model argue against you before it agrees
victor_explore · x · 2026-09-26
A practical prompt tip: put "argue against this" in the prompt before asking for anything.
- A model tuned to agree will sign off on a bad plan — and an agent reviewing its own code is just that same model agreeing with itself.
- The quoted context cites Nobel laureate Craig Mello's advice that the safe way to handle information is to use it to try to disprove what you believe.
More from coding & agent
- Grok now drives nearly 60% of agentic trading on Coinbase, dwarfing Claude and custom CLIs — XFreeze · 2026-09-26
- QuackIR: Jimmy Lin's EMNLP Paper Shows RDBMSes Match Vector DBs for RAG Retrieval — lintool · 2026-09-26
- Where to draw the determinism boundary in LLM pipelines — aronchick · 2026-09-26
- The only pipeline shape that survives production: LLM proposes, rules dispose — aronchick · 2026-09-26
- 100-slot Terminal-Bench 2.1 rerun shows Luna 6 far behind Luna 5.6 — s1lverkin · 2026-09-26
- Qwen3.8-Omni-Flash: natively multimodal agent model with 1M-token context — dair_ai · 2026-09-26