At the AtCoder World Tour Finals in Tokyo, Polish programmer Przemysław Dębiak — competing under the alias “Psyho” — claimed first place in a grueling 10-hour coding contest. But what set this event apart wasn’t just the difficulty or the competition. It was the entry of a formidable new challenger: an advanced AI model developed by OpenAI.

This marked the first time an AI participated in such a prestigious programming contest. Entering under the name “OpenAIAHC,” the AI model competed in a special “Humans vs AI” exhibition within the Heuristic division. It faced off against 12 top-ranked human coders, including Dębiak himself.

Human Grit vs Machine Precision

Despite extreme fatigue, Dębiak outperformed the AI by just over 9%. In a post-contest statement, he admitted that the AI’s presence pushed him to the edge: “If not for the AI, my score would’ve been much, much lower.” The thrill of racing just ahead of the machine became his primary motivation to keep going.

The AI didn’t just trail behind — it beat all the other human contenders, finishing second overall. This performance underscores just how close AI has come to rivaling humans in creative problem-solving and programming prowess.

The Future Is Competitive

Dębiak, a former OpenAI employee, may have won this round, but the writing on the wall is clear: AI is quickly catching up. As these systems continue to evolve, the question isn’t whether they’ll win next time — it’s how soon.