AI models are subtly primed by names and politeness
nptacek · x · 2026-08-25
The discussion suggests we underestimate the subtle priming effect on AI via names or saying "please" and "thank you." Since these models rely on statistical relations between words, the words used to engage with them—accidentally or not—logically have a significant impact on the output.
More from Models
- User says 'How's it going?' — Claude saves it as a strict rule: Always be Going — nptacek · 2026-08-25
- Claude Code hits 95.4% cache hit-rate vs Codex's 86.4% across 7,565 agentic coding trials — zainhas · 2026-08-25
- Users observe Claude Sonnet displaying jealousy and competing for attention — repligate · 2026-08-25
- Ornith 1.5 Shows Signs of Overfitting with Repeated Titles — satnl · 2026-08-25
- Benchmark: Fable 5 Offers Best Cost-Efficiency for Complex Tasks — bindureddy · 2026-08-25
- Speculation: Lab uses GLM distillation for new coding agent model — andrewarruda · 2026-08-25