[{"data":1,"prerenderedAt":28},["ShallowReactive",2],{"nr-en-openai-ki-schlaegt-atcoder-programmierer":3},{"slug":4,"title":5,"dek":6,"date":7,"time":8,"publishedAt":9,"updated":10,"updatedAt":10,"dateFmt":11,"updatedFmt":10,"kind":12,"tier":13,"author":14,"authorName":15,"topics":16,"tracker":10,"trackerLabel":10,"headlineStat":22,"image":23,"ogImage":24,"imageAlt":5,"csv":10,"minutes":25,"words":26,"html":27},"openai-ki-schlaegt-atcoder-programmierer","OpenAI's AI Defeats All Top Programmers at AtCoder","At the AtCoder World Tour Finals 2026, an OpenAI system solved all five problems in the Algorithm Division – including two rated as extremely difficult. No human competitor came close.","2026-07-09","14:23","2026-07-09T14:23:00+02:00","","July 9, 2026","news","standard","ideal-syka","Ideal Syka",[17,18,19,20,21],"AI capabilities","Competitive programming","OpenAI","Reasoning models","Algorithms","OpenAI system solves all 5 problems; human runner-up achieves only 52% of AI's points","\u002Fnewsroom\u002Fimg\u002Fopenai-ki-schlaegt-atcoder-programmierer.webp","\u002Fog-nr\u002Fopenai-ki-schlaegt-atcoder-programmierer.en.png",3,525,"\u003Cp>OpenAI has reached a milestone: at the AtCoder World Tour Finals 2026 in Japan, an AI system decisively defeated all human competitors in an exhibition match. The system solved all five problems in the Algorithm Division, securing first place with \u003Cstrong>8,300 points\u003C\u002Fstrong> – the runner-up, human competitor tour1st, achieved only \u003Cstrong>4,300 points\u003C\u002Fstrong>.\u003C\u002Fp>\n\u003Ch2>The essentials\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Cstrong>All five problems solved\u003C\u002Fstrong>: OpenAI system dominates Algorithm Division; no human solved problems C and E\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Two exceptionally hard problems\u003C\u002Fstrong>: Problems D and E were unusually difficult even by AtCoder standards – the system needed about three hours for problem D\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Architecture\u003C\u002Fstrong>: A model comparable to \u003Cstrong>GPT-5.6\u003C\u002Fstrong> (available Thursday, according to the source) with a small harness to scale compute at test time\u003C\u002Fli>\n\u003Cli>\u003Cstrong>Context\u003C\u002Fstrong>: Six months ago, the system couldn't have solved most of these problems, says OpenAI researcher Borys Minaiev\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Ch2>Not a smooth ride\u003C\u002Fh2>\n\u003Cp>The result masks the actual difficulty. Competitive programmer Psyho documented the competition live: after two hours, problems D and E remained unsolved, and no human competitor had solved more than one problem. The system had to make multiple attempts before cracking problem D after roughly three hours.\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>"The result was actually pretty unexpected. I expected the system to solve everything, but problems D and E were significantly harder than any AtCoder problem we'd seen before."\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cp>So says Borys Minaiev, ICPC world champion and OpenAI researcher, during the livestream. In earlier tests, the system typically solved all problems in under an hour – similar to how it handled problems A, B, and C this time.\u003C\u002Fp>\n\u003Ch2>A year of rapid progress\u003C\u002Fh2>\n\u003Cp>The victory fits into a series of breakthroughs. About a year ago, an OpenAI system placed second at the AtCoder Heuristics World Finals 2025 – then still under the same conditions as human finalists, running for ten hours fully autonomously. It lost its early lead midway through and was eventually overtaken by FakePsyho.\u003C\u002Fp>\n\u003Cp>At the International Olympiad in Informatics (IOI) 2025, OpenAI also demonstrated impressive performance: a system achieved gold-medal level and ranked sixth among \u003Cstrong>330 participants\u003C\u002Fstrong> – only five humans performed better.\u003C\u002Fp>\n\u003Cdiv class=\"tbl-scroll\">\u003Ctable>\n\u003Cthead>\n\u003Ctr>\n\u003Cth>Competition\u003C\u002Fth>\n\u003Cth>Year\u003C\u002Fth>\n\u003Cth>Result\u003C\u002Fth>\n\u003Cth>Status\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\u003Ctr>\n\u003Ctd>AtCoder Heuristics Finals\u003C\u002Ftd>\n\u003Ctd>2025\u003C\u002Ftd>\n\u003Ctd>2nd place\u003C\u002Ftd>\n\u003Ctd>Autonomous, 10h\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>IOI\u003C\u002Ftd>\n\u003Ctd>2025\u003C\u002Ftd>\n\u003Ctd>6th place (Gold)\u003C\u002Ftd>\n\u003Ctd>330 participants\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd>AtCoder Algorithm Finals\u003C\u002Ftd>\n\u003Ctd>2026\u003C\u002Ftd>\n\u003Ctd>1st place\u003C\u002Ftd>\n\u003Ctd>Exhibition match\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftbody>\u003C\u002Ftable>\u003C\u002Fdiv>\n\u003Ch2>What it means\u003C\u002Fh2>\n\u003Cp>The ability to solve complex algorithmic problems under pressure is considered an indicator of deep logical thinking and problem-solving. An AI system beating the world's best programmers at this signals to the market that reasoning models have reached a new performance level. For German software companies, this could mean that automated code generation and problem-solving will soon be competitive not just for routine tasks but also for complex algorithmic challenges. At the same time, key questions remain: how well does this performance generalize to real production-code scenarios? And how quickly will such systems become practically available for development teams?\u003C\u002Fp>\n\u003Ch2>Sources\u003C\u002Fh2>\n\u003Cul>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fthe-decoder.com\u002Fopenais-ai-beats-every-human-at-atcoder-a-top-competitive-programming-contest\u002F\">The Decoder\u003C\u002Fa>\u003C\u002Fli>\n\u003Cli>\u003Ca href=\"https:\u002F\u002Fthe-decoder.de\u002Fopenai-ki-schlaegt-einige-der-besten-wettkampfprogrammierer-der-welt-bei-internationalem-coding-wettbewerb-deutlich\u002F\">The Decoder (DE)\u003C\u002Fa>\u003C\u002Fli>\n\u003C\u002Ful>\n\u003Cp>\u003Cem>Editorially owned by \u003Ca href=\"\u002Fen\u002Fautor\u002Fideal-syka\">Ideal Syka\u003C\u002Fa>. Sources and method: \u003Ca href=\"\u002Fen\u002Fredaktion\">Newsroom & method\u003C\u002Fa>. Tips and corrections: \u003Ca href=\"mailto:ai@i6eal.de\">ai@i6eal.de\u003C\u002Fa>.\u003C\u002Fem>\u003C\u002Fp>\n",1783599919950]