GPT Beats Human Top Score, Raising Doubts Over Competitive Programming Benchmarks

GPT reportedly scored 600 in an IOI-style contest versus a human best of 500, prompting claims that competitive programming is losing value as a human intelligence benchmark. A former contestant dismissed proposals to raise difficulty, arguing models will outsmart all human competitors and problem setters by 2027.

2026-08-26 ~ 2026-08-27 · 2 related posts

• GPT outperforms human max score, questioning the value of IOI as a benchmark — rpeng233 · 2026-08-26
• Former CP contestant rejects "raising difficulty to counter AI": pointless — anpaure · 2026-08-27