OpenAI’s latest model, o3, has claimed victory in a high-stakes AI chess tournament, defeating Elon Musk’s xAI model, Grok 4, in a decisive final match. The tournament, organized by Kaggle, featured eight leading large language models from companies including Anthropic, Google, and DeepSeek, showcasing the rapid advancements in artificial intelligence.
The competition took place over three days and drew attention from the tech community as AI models battled for supremacy. Going into the final, Grok 4 was already a winner to be reckoned with due to a series of decisive wins. Pedro Pinhata, a writer for Chess.com, noted that “up until the semi-finals, it seemed like nothing would be able to stop Grok 4 on its way to winning the event.” Yet, the decisive third bout would go on to become a watershed moment.
In an unexpected outcome, Grok 4’s performance was described as “unrecognizable” and “blundering,” paving the way for OpenAI’s o3 model to secure a series of convincing victories. Chess grandmaster Hikaru Nakamura commented on the match, stating, “Grok made so many mistakes in these games, but OpenAI did not.”
Musk had earlier dismissed xAI’s role in chess. He charged the company with having made an “incredible little effort on chess.” This comment followed Grok 4’s victory as the strongest player in the tournament. Though he was very confident at the outset that his model would win, the result in the end was an example of the competitive advantage that o3 has.
The historical context of AI in chess recalls the famous matches between Russian grandmaster Garry Kasparov and IBM’s Deep Blue in the late 1990s. As Kasparov once infamously stated, if you want to know how smart Deep Blue is, don’t ask an alarm clock. He expressed his frustration about losing to such technology: “Losing to a $10m (£7.6m) alarm clock did not make me feel any better.” His experience is a perfect example of the shifting paradigm between human intelligence and artificial intelligence.
We know that technology is advancing at an incredibly fast pace. It’s not unlike the real-life story of South Korean Go master Lee Se-dol, who retired in 2019 following his 4-1 defeat to AlphaGo. Lee stated, “There is an entity that cannot be defeated,” signaling a recognition of AI’s growing prowess in strategic games.
OpenAI co-founder and AI entrepreneurial superhero, Sam Altman, and Musk would both argue that their latest models are the smartest in the world. This tournament will be the litmus test that proves Altman’s claims about o3’s capabilities. If o3 beats the competition that has historically done well, the scorecard will tell the story.