Psyho’s final score of 1,812,272,558,909 points was enough to put him ahead of OpenAI’s model at 1,654,675,725,406 points by a margin of about 9.5 percent. The head-to-head competition lasted 10 hours and left Psyho “completely exhausted.” He said he was “barely alive” by the end.
Can humans prevail against AI in coding competitions?
Answer: Yes!
“Humanity has prevailed (for now!),” was the sentiment of programmer Przemysław Dębiak after narrowly defeating an artificial intelligence model on Wednesday. Dębiak, also known as Psyho, competed against a custom AI model from OpenAI in the final round of the AtCoder World Tour Finals 2025 Heuristic contest in Tokyo.
Psyho’s final score of 1,812,272,558,909 points was enough to put him ahead of OpenAI’s model at 1,654,675,725,406 points by a margin of about 9.5 percent. The head-to-head competition lasted 10 hours and left Psyho “completely exhausted.” He said he was “barely alive” by the end.
“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,” an OpenAI spokesperson said of the event. “Events like AtCoder give us a way to test how well our models can reason strategically, plan over long time horizons, and improve solutions through trial and error — just like a human would.”
Psyho’s final score of 1,812,272,558,909 points was enough to put him ahead of OpenAI’s model at 1,654,675,725,406 points by a margin of about 9.5 percent. The head-to-head competition lasted 10 hours and left Psyho “completely exhausted.” He said he was “barely alive” by the end.
