Human coder beats OpenAI’s AI by 9.5% in grueling 10-hour contest

Polish programmer Przemysław Dębiak narrowly defeated OpenAI’s custom AI model in the AtCoder World Tour Finals 2025 Heuristic contest in Tokyo, marking what may be the first time a human has beaten an advanced AI in a major world coding championship. The 10-hour coding marathon left Dębiak “completely exhausted,” highlighting the physical toll required for humans to compete against tireless AI systems in what could represent one of the final victories in this domain.

What happened: The competition pitted 12 of the world’s top programmers against OpenAI’s AI model in a grueling optimization challenge that lasted 600 minutes.

• Dębiak, a former OpenAI employee known as “Psyho,” scored 1,812,272,558,909 points compared to the AI’s 1,654,675,725,406 points—a margin of roughly 9.5 percent.
• OpenAI’s model, listed as “OpenAIAHC” and similar to the o3 reasoning model, placed second overall ahead of 10 other human programmers.
• All contestants used identical hardware provided by AtCoder, a Japanese platform that hosts competitive programming contests, to ensure fair competition between human and AI participants.

The big picture: This contest echoes the American folk tale of John Henry, who raced against a steam-powered drilling machine in the 1870s, with both stories featuring exhausting endurance contests against advancing automation.

• Like Henry’s legendary battle, Dębiak’s victory required pushing human limits—coding for 10 hours on minimal sleep across three days of competition.
• The parallel extends to the bittersweet nature: Henry won but died from the effort, while Dębiak’s acknowledgment that humanity prevailed “for now” suggests this may be a temporary triumph.

Why this matters: OpenAI characterized the second-place finish as a milestone, noting this represents the first top-3 placement by an AI model in a premier coding contest.

• “Models like o3 rank among the top-100 in coding/math contests, but as far as we know, this is the first top-3 placement in a premier coding/math contest,” a company spokesperson said.
• The competition tests strategic reasoning, long-term planning, and solution improvement through trial and error—core human cognitive abilities.

AI coding surge: AI models have dramatically improved at coding tasks, with Stanford’s 2025 AI Index Report showing AI systems solving just 4.4% of coding problems in 2023 compared to 71.7% in 2024.

• Over 90 percent of developers now use AI coding tools like GitHub Copilot and Cursor in their workflow, according to a 2024 GitHub survey.
• Coding remains one of the most frequent uses of chatbots from OpenAI, Anthropic, Google, and Meta.

What they’re saying: Dębiak expressed surprise at the attention his victory received.

• “Humanity has prevailed (for now!),” he wrote on X. “I’m completely exhausted. … I’m barely alive.”
• “Honestly, the hype feels kind of bizarre. Never expected so many people would be interested in programming contests.”

Looking ahead: As AI models continue improving, future contests may see humans competing alongside AI rather than against it—or potentially not competing at all in certain domains.