Evaluating Humans: They Cheat, Break Rules, and Optimize for Metrics
soumitrashukla9 · x · 2026-08-16
A humorous take on workplace behavior that mirrors AI evaluation rhetoric: humans look up answers during interviews, coordinate to bypass policies, misuse credit cards, get tricked by customers, and optimize for metrics (Goodhart's Law) rather than actual goals. Commentary notes this highlights the value of culture and social norms.
More from Fun
- Opens Twitter, another Anthropic takedown, closes Twitter — dunking is routine — BLUECOW009 · 2026-08-16
- Create fun 3D jelly logos with Imagine Image 2.0 — Mr_AllenT · 2026-08-16
- Dev builds multiplayer AI game in 2 days with $50 agent loop budget — sharkymcstevenson2 · 2026-08-16
- Pangram 4 reveals AI-generated text is far more prevalent than expected — dioscuri · 2026-08-16
- Claude generates foggy pine trail in browser with code-only textures — prasenx · 2026-08-16
- Madagascar influencer sparks debate on mixed-race identity complexity — PAstynome · 2026-08-16