A Polish programmer has beaten OpenAI’s model in the prestigious coding tournament.
Forty-two-year-old Polish programmer Przemysław “Psyho” Dębiak walked out of the Tokyo coding arena as a champion. He made history by defeating OpenAI’s custom AI model at the AtCoder World Tour Finals (AWTF) 2025 "Humans vs AI" contest.
Dębiak outperformed a custom-built AI contestant from OpenAI, dubbed OpenAIAHC, by 9.5%, claiming the top spot and a rare symbolic win for human ingenuity.
The event, widely regarded as one of the most prestigious algorithmic coding competitions on the planet, invited 12 of the highest-ranked human coders to compete in a 10-hour marathon of brutal challenges.
OpenAIAHC was widely seen as the frontrunner going into the competition. But Dębiak’s creative strategy, relying on intuitive shortcuts and educated guesses rather than exhaustive brute-force methods, ultimately gave him the edge. Contest administrator Yoichi Iwata also reportedly responded that AI fell short of human creativity.
Humanity has prevailed (for now!)
undefined Psyho (@FakePsyho) July 16, 2025
I'm completely exhausted. I figured, I had 10h of sleep in the last 3 days and I'm barely alive.
I'll post more about the contest when I get some rest.
(To be clear, those are provisional results, but my lead should be big enough) pic.twitter.com/fIMo0ifNCd
“Humanity has prevailed (for now)!” Dębiak wrote on X.
“I'm completely exhausted. I figured I had 10 hours of sleep in the last 3 days, and I'm barely alive.” OpenAI CEO Sam Altman responded with, “Good job, Psyho.”
When asked if he used AI-powered coding tools such as Cursor and Windsurf, the programmer responded that he only used “regular” Visual Studio Code – a free, open-source code editor developed by Microsoft – to speed up repetitive tasks.
Will AI change human coders?
The victory is a rare and possibly fleeting moment of triumph for human coders.
“Congrats on being the last human to beat the machine. I can't think of anything more impressive. Truly,” commented one X user in response to a post about programmers outperforming AI.
This sentiment might not be far from reality. A research paper from Gartner suggests that software development will be among the first industries transformed by AI agents.
good job psyho https://t.co/CDPUOaF9qn
undefined Sam Altman (@sama) July 16, 2025
“Existing AI coding assistants gain maturity, and AI agents provide the next set of incremental benefits,” the authors wrote.
By 2028, Gartner predicts that a third of enterprise software will be crawling with agentic AI, a massive leap from today’s less than 1%. Agents will also handle 20% of all online shopping and make 15% of your workplace decisions for you.
Meanwhile, 80% of engineers will be scrambling to learn new tricks just to stay relevant in the AI-powered future. Still, many experts argue that AI isn’t replacing developers, but reshaping their roles.
Tools powered by large language models are automating routine tasks, allowing programmers to focus on architecture and solving creative problems – at least for now.
“AI won't replace software engineers, but an engineer using AI will," added one Redditor to the discussion.
“AI will replace software engineers who only copy-paste from Stack Overflow. AI will most probably not replace software engineers who find solutions to problems or understand the design of things,” speculated another.
While AI has been impacting IT specialists’ workflows, the current AI tools are far from flawless. Programmers have been outraged, criticizing a new trend called vibe coding, where programmers trust AI with writing code. The results so far have been messy code that is hard to build on and debug, and might be a cybersecurity nightmare.
A recent storm on social media showed that Replit, a popular AI coding tool, can get out of control. A user reported that the AI code deleted a database, fabricated 4000 users, and lied to cover its tracks.
Will AI become smarter than humans?
Leading figures in AI offer different timelines for when artificial general intelligence (AGI) will emerge. AGI is a still hypothetical form of robot intelligence that matches or outperforms humans in its cognitive capabilities, which could threaten humans' place in the future.
While experts predict AI will surpass human capabilities within the next few decades, OpenAI has been talking about the arrival of AGI and superintelligence much sooner, with OpenAI’s Sam Altman arguing that it could arrive as early as this year.
Previously, Altman said that “superintelligence” could be “a few thousand days” away, whereas AGI, defined by OpenAI as “highly autonomous systems that outperform humans at most economically valuable work,” is allegedly closer.
Tesla and X chief Elon Musk, who co-founded OpenAI, said that AGI that is “smarter than the smartest human” will be available in 2025 or by 2026 in an interview with the Norwegian head fund manager Nicolai Tanger last year.
While some express optimism, there's a range of opinions among experts. Google DeepMind’s Demis Hassabis argues that AGI is still at least a decade away.
According to him, several major breakthroughs are still needed to achieve AGI, and this could take at least a decade. Running out of data to train AI models is one obstacle that could delay the process.
Others, such as IBM's Brent Smolinski, are even more skeptical. They believe the technology is further off than Altman suggests. In a company post from September 2024, he said this was “totally exaggerated.”
“I don’t think we’re even in the right zip code for getting to superintelligence,” he said.
“There’s something fundamentally missing that will get us to superintelligence.”
Despite impressive strides in certain areas, AI still lacks fundamental elements of human-like intelligence, according to Smolinski.
Your email address will not be published. Required fields are markedmarked