Sam Altman toppled Elon Musk in a final showdown of a tournament that sought to crown the best artificial (AI) chess player. It’s just that the ones pushing their chess skills into overdrive for the big prize on either side of the table weren’t the aforementioned tech giants in the flesh, but their very recent AI models – OpenAI’s o3 and xAI’s Grok4, respectively.
While the AI programs meant for everyday use fanned the fire of professional rivalry lit between Altman and Musk, it was ultimately Magnus Carlsen, who actually humbled the tech titans. Giving his verdict on each of the leading AI overlords’ (most importantly, Elon Musk, Sam Altman and Mark Zuckerberg) chess acumen, Carlsen put at least one of them to shame more than the others. Ultimately, neither of the three emerged as victors.
Carlsen not only stepped in to offer commentary for the AI chess tournament between several large language models (LLMs) on Google-owned platform Kaggle, but has also seen the tech titans pioneering the AI landscape in action on the chessboard. Emerging as arguably the greatest chess player of time, Magnus has instead made more headlines on the Internet’s viral scene. After his table-slamming stunt on losing to India’s D Gukesh, the Norwegian chess grandmaster has particularly been the focus of mainstream social media trends.
Magnus Carlsen’s verdict on Elon Musk, Sam Altman and Mark Zuckerberg’s chess skills
In light of that newfound popularity on the viral stream, the Chess star’s comments about having played in the same tournaments as Altman, Zuckerberg and Musk are the latest bit of trivia making headlines. While the first two influential men from the AI scene were still somewhat spared, Elon Musk got the rough end of the stick.
Unlike his in-person encounters with the Facebook and OpenAI bosses, Carlsen shared that while he had seen him up close, he hadn’t talked to him. “You know he famously doesn’t have the greatest respect for chess players or the game which, to some extent, I understand because it is not a very complicated game, right?” said Magnus. “It is very simple in many ways but I think that’s also the beauty of the game. Like it’s obviously simple since it took computers not that long to master it, right? Compared to some other games.”
He then went on to gush about the game at length, calling it “rich and very difficult to play.” And yet, he continued, “Like it’s simple enough to play that you can get joy from playing after practice. Like not maybe immediately but after practice practicing a bit but hard enough that you can never actually get particularly good at it as a human which we’ve found out by seeing engines play,” indirectly taking a dig at Musk.
In the same manner, the chess grandmaster ripped Musk’s Grok 4 after its 4-0 loss to OpenAI’s o3 just this week. He likened the game between the two AI models to “kids’ games,” adding “Hope everyone feels better about their games after watching this.” He eventually estimated Musk’s AI model’s skill at a low 800 against o3’s 1200.
“It was an alternate moves tournament in Silicon Valley. It was around the Champions Chess Tour final in 2022,” Carlsen said of having player in the same tournament as the OpenAI CEO. The Norwegian player recounted that Dutch chess grandmaster Anish Giri was on the team with Altman at the time, and the AI pioneer’s chess skills failed to meet his standards.
“So Giri was very unhappy about being teamed with him,” Carlsen added, while laughing as the past account was revived in his mind. He still managed to cut Sam Altman some slack, adding, “But he actually learned a lot through the experience. You could tell that he was very smart.”
Akin to his verdict on Altman’s skills, Carlsen said that even Zuckerberg didn’t know much of the game. “He was a little bit the same as Sam Altman. I thought he was even better,” he divulged.
After torching Mark’s efforts, the Norwegian player still made amends by noting that the Facebook chief was actually “learning very, very quickly” and “forming his own opinions very quickly.” Hailing his “impressive” and “useful” skill, Carlsen said that while his decisions weren’t always right, it was nothing out of the ordinary. “…but they were always well-reasoned,” he finally let Zuckerberg off the hook.
British AI researcher, DeepMind CEO – the only champion in Magnus Carlsen’s eyes
Demis Hassabis, the DeepMind Technologies CEO and recent Nobel Prize (Chemistry) laureate, also became the subject of Carlsen’s critical observations a few years ago at the London Chess Classic. Unlike others, he straight out went ahead and called him “fantastic.”
“I’ve met him many times, but in terms of chess, he’s kind of set in his ways: like he loves playing the Botvinnik System as white. And I asked him, ‘Do you know anything else?’ And he was like, ‘No, not really. That’s what I like to play.’ And I was like, ‘Yeah, fine. That’s great,'” Magnus detailed their shared experience.
Having already beaten some good players during the tournament, the two ended up with a tiebreaker in the finals. “And as a tiebreak the two amateur players had to play each other, right? And then we discussed the game before and he said, ‘My opponent’s going to play 1.c4. What should I do?’ And I said, ‘What do you normally do?” Well, I play 1.c5, then I play g6 and so on. So I told him ways you could get around (the opponent’s tricks in the opening),” Carlsen went on.
Contrary to what had been decided, Hassabis eventually forgot everything and the pair fell to a “horrible passive position,” eventually losing it all. Despite the loss, Carlsen praised Demis for his contribution to chess, while particularly celebrating his offerings to humanity with medicine.