Human programmer beats OpenAI's custom AI in 10-hour marathon, wins World Coding Championship — Polish programmer might be the last human winner
Przemysław 'Psyho' Dębiak, a 42-year-old programmer from Gdynia, Poland, made history by defeating OpenAI's custom AI model at the AtCoder World Tour Finals (AWTF) 2025 "Humans vs AI" contest in Tokyo. Considered one of the most prestigious coding tournaments in the world, the AWTF invites just 12 of the top-ranked human programmers—and, for the first time, an AI competitor—to tackle its grueling challenges. After a 10-hour coding marathon, Dębiak edged out the AI by roughly 9.5%, clinching first place while the OpenAI-built model settled for second.
'Humanity has prevailed (for now)!' Dębiak wrote on X, confessing he'd slept only around 10 hours over three days while pushing himself to the limit. OpenAI CEO Sam Altman aptly responded with, 'Good job, Psyho.' The AI model, branded OpenAIAHC, was widely expected to dominate the market. Still, Dębiak's innovative, heuristic-driven approach—using problem-solving shortcuts and educated guesses instead of brute-force calculation—secured the win. Contest administrator Yoichi Iwata praised his distinctive method, noting that while the AI excelled at raw optimization, it 'fell short of human creativity.'
The AtCoder World Tour Finals (AWTF) is regarded as the ultimate stage for heuristic programming contests that focus on 'good-enough' solutions to complex, unsolvable problems rather than perfect ones. This year's challenge tasked competitors with plotting a robot's path across a 30×30 grid using the fewest possible moves, an NP-hard optimization problem with countless possible outcomes. With no access to external libraries or documentation, success relied on intuition, creativity, and adaptability—qualities where human ingenuity can still outthink the raw speed and precision of AI models.
Dębiak, a former OpenAI engineer who helped develop OpenAI Five (the Dota 2 AI), competed using only Visual Studio Code with basic autocomplete and admitted the AI pushed him to his limits: 'I was close to the model's score, and that pushed me to give everything.' Near the end of the 10-hour marathon, he overtook OpenAIAHC to claim victory and a 500,000 yen prize. A veteran algorithmic competitor, Mensa member, and four-time TopCoder Open Marathon champion, Dębiak has never held a full-time job and once even joked about considering careers from DJing to professional poker. His win carries symbolic weight, a human triumphing over AI in a field where machines typically excel.
All that being said, AI's steady progress is undeniable—Stanford's 2025 AI Index found coding benchmarks jumped from 4.4% AI success in 2023 to 71.7% in 2024. With tools like GitHub Copilot used daily by over 90% of developers, AI is reshaping workflows. However, this win highlights that creativity, endurance, and intuition —especially in long-form heuristic challenges —remain human strengths.
Dębiak isn't naive about the challenge ahead, saying, 'It's easy to imagine a different problem where AI would win and humans would be far behind.'. Still, this feels like a human John Henry moment—a testament not just to sheer human will, but to the spark of creativity machines haven't replicated and (hopefully) will never be able to.
Follow Tom's Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
15 minutes ago
- Yahoo
SoundHound AI Stock Jumps 25% on Strong Results, Raised Guidance
Key Takeaways SoundHound AI set a sales record as all of its businesses grew. The artificial intelligence voice software maker also reported a lower-than-expected loss. SoundHound AI raised its full-year revenue AI (SOUN) shares took off, soaring 25% in morning trading, when the maker of artificial intelligence voice software reported record sales and boosted its outlook as it added more customers, especially in China. The company reported second-quarter revenue that skyrocketed 217% to $42.68 million, about $10 million more than analysts from Visible Alpha anticipated. Its per share loss of $0.03 was less than a year ago, and also beat estimates. Co-founder and CEO Keyvan Mohajer said it was SoundHound AI's 'strongest ever quarter,' and that several of its previous strategic moves 'delivered high-impact results with growth across every business unit.' Among the gains, the Automotive business scored what the company called a 'significant new win in China for our voice assistant to be integrated in a major OEM with vehicles spanning China and worldwide, including the development of multiple new languages.' SoundHound AI raised its full-year revenue guidance to a range of $160 million to $178 million, compared to its earlier forecast of $157 million to $177 million. Despite today's advance, shares of SoundHound AI have lost about a third of their value in 2025. Read the original article on Investopedia
New York Post
17 minutes ago
- New York Post
Anti-DEI activist Robby Starbuck settles lawsuit over Meta on AI chatbot defamation
Conservative firebrand Robby Starbuck and Meta have agreed to a settlement after he sued the company alleging its AI chatbot defamed him. As part of the settlement, Starbuck will serve as a consultant to Meta, working with its Product Policy team to bolster their existing efforts to combat political bias in its AI models and reduce the risk of 'hallucinations' – made-up information – generated by the chatbot. Advertisement 'Both parties have resolved this matter to our mutual satisfaction. Since engaging on these important issues with Robby, Meta has made tremendous strides to improve the accuracy of Meta AI and mitigate ideological and political bias. Building on that work, Meta and Robby Starbuck will work collaboratively in the coming months to continue to find ways to address issues of ideological and political bias and minimize the risk that the model returns hallucinations in response to user queries,' Meta and Starbuck said in a joint statement. 'I'm extraordinarily pleased with how Meta and I resolved this issue. Resolving this is going to result in big wins that I believe will set an example for ethical AI across the industry. I look forward to continuing our engagement as a voice for conservatives to ensure that we're always treated fairly by AI,' Starbuck told Fox News Digital. Starbuck filed suit in Delaware Superior Court against Meta in April after its chatbot allegedly falsely accused him of being a 'White nationalist' who was arrested on Jan. 6 and claimed he had been sued for defamation. The chatbot allegedly recommended that the right-wing influencer lose custody of his children, saying he was a danger to them. Advertisement 3 Anti-DEI activist Robby Starbuck and Meta have settled in court after he sued the company, William DeShazer for NY Post The anti-DEI activist was first made aware of the AI chatbot's smears against him in August 2024, when an X user posted several instances of Meta AI spreading misinformation about him online. The chatbot falsely claimed Starbuck was present at the Jan. 6 riot, was linked to the Q-Anon conspiracy, and was anti-vaccine. Starbuck proceeded to investigate for himself whether Meta AI was spreading lies about him and uncovered additional falsehoods from the chatbot. Advertisement He decided to file a lawsuit but not before he tried to resolve the issue with Meta. Woke bias has been a concern across multiple AI platforms produced by different companies. 3 Starbuck filed suit against Meta in April after its chatbot allegedly falsely accused him of being a 'White nationalist' who was arrested on Jan. 6. REUTERS Google Gemini faced backlash after its platform claimed that Memorial Day is a controversial holiday, and generated racially inaccurate images of historical figures. Advertisement ChatGPT, the chatbot pioneered by OpenAI, had previously been found to refuse requests to praise Donald Trump, but would accept requests to praise Kamala Harris or Joe Biden. OpenAI subsequently implemented programs designed to combat bias in its models. Meta claimed that addressing political bias in its AI models has been a longstanding goal for the company. 'It's well-known that all leading LLMs have had issues with bias—specifically, they historically have leaned left when it comes to debated political and social topics. This is due to the types of training data available on the internet, our goal is to remove bias from our AI models and to make sure that Llama can understand and articulate both sides of a contentious issue.' Meta wrote on their blog in April. The conservative activist told Fox News Digital he intends to use his new role with Meta to ensure that no American of any political stripe is subject to ideological bias. He's optimistic that his work with Meta will resonate across the AI industry as it continues to become an increasing presence in Americans' daily lives. 3 As part of the settlement, Starbuck will serve as a consultant, working with Meta's Product Policy team to bolster their existing efforts to combat political bias in its AI models. Ana Gioia/NY POST DESIGN 'I think a tech leader like Meta working with me is a critically important step to producing a product that's fair to everyone. I think what we do to improve AI training could become an industry standard and I also think we'll set an example for the entire industry when it comes to ensuring fairness. I'm laser focused on that and ensuring that the highest quality training materials are used,' Starbuck told Fox News Digital. Meta has attempted to address areas of perceived anti-conservative bias within its company since President Donald Trump returned to office. In January, the social media behemoth announced it was ending its Diversity, Equity and Inclusion policies. Advertisement The company also brought on former Republican political consultant Joel Kaplan, who worked as deputy chief of staff in the George W. Bush administration, to serve as its chief global affairs officer. Kaplan told Fox News Digital in January that eliminating DEI would ensure the company builds teams with 'the most talented people.'
Engadget
17 minutes ago
- Engadget
Apple will bring GPT-5 to Apple Intelligence in iOS, iPad OS and macOS 26
OpenAI finally released its long-awaited GPT-5 model this week, unsurprisingly proclaiming it its best yet with regards to coding, accuracy, safety and more. CEO Sam Altman even compared the jump up in quality to when the iPhone first adopted a Retina display in a press briefing ahead of the announcement. Big talk indeed. Given ChatGPT's integration with Apple Intelligence , you might be wondering when the latest model will arrive on the devices that support it. The answer is sooner rather than later, according to 9To5Mac . Apple told the publication in a statement that Apple Intelligence will leverage GPT-5 in iOS 26 , iPadOS 26 and macOS Tahoe 26. Those software updates are set to arrive in their final form, as they traditionally do, at the tail end of the summer, so you won't have to wait long. Currently, ChatGPT is an optional add-on to Apple's own AI model. It can be accessed by Siri when you're making a request that Apple Intelligence on its own might not be able to handle, with Apple using the example of questions about photos and documents. ChatGPT also integrates with Camera Control in your iPhone's Camera app to learn more about your environment, and if you connect your ChatGPT account you can save your requests in your personal history. ChatGPT has to be activated as an extension to Apple Intelligence in settings, and if you turn it off you won't have to worry about Siri turning to an external model if you're concerned about it doing so. OpenAI says GPT-5 is its safest model to date, thanks in part to a new feature called Safe Completions, which means it will attempt to give the most helpful answer it can within safety parameters imposed by its maker. GPT-5 is free for everyone, but OpenAI offers tiered paid plans for heavier users, from its $20 per month ChatGPT Plus subscription, right up to the $200 per month Pro plan that offers unlimited use. If you have one of these plans, you can log in through Apple Intelligence and gain all the same benefits. Engadget has reached out to Apple for comment as we'll update this article if we get a response.
