Burkov: Models Still Sycophantic; Without Harness, Intelligence on Par with o1

Andriy Burkov argues that today's models still exhibit sycophancy, manufacturing flaws in flawless inputs, and that stripped of evaluation harnesses their true intelligence is roughly at the level of o1.

2026-09-13 ~ 2026-09-13 · 2 related posts

• Burkov: Models Inventing Fake Issues Is the Same Sycophancy We Saw in 2024 — burkov · 2026-09-13
• Burkov: strip away the harness and models are only o1-level smart — burkov · 2026-09-13