An Unexpected Endorsement From Crypto's Philosopher

Ethereum (ETH) co-founder Vitalik Buterin dropped an interesting take Thursday: Grok, Elon Musk's AI chatbot on X, is actually making the platform better at sorting truth from nonsense. In fact, he says it's the second-most important feature for "truth-friendliness" on X, trailing only community notes.

That's a bold statement for a platform many critics say has become a haven for conspiracy theories and algorithmic manipulation. But Buterin's reasoning is worth unpacking.

Why Buterin Thinks Grok Works

In his X post, Buterin highlighted the "easy ability" to summon Grok for instant fact-checks, calling it the "biggest thing" after community notes that's been genuinely positive for the platform.

He's noticed something entertaining: people call on Grok expecting it to validate their wild political beliefs, and instead the AI essentially tells them they're wrong. "I've seen many situations where someone calls on Grok expecting their crazy political belief to be confirmed and Grok comes along and rugs them," Buterin wrote.

The Musk Problem (And Why It Might Not Matter)

Buterin isn't naive about the obvious issue here. Grok's training could easily skew toward Musk's worldview since the Tesla (TSLA) and xAI CEO literally owns the platform. That's a legitimate concern about bias baked into the system.

But here's where Buterin gets philosophical. Despite "negative" assumptions about Musk, he believes X with Grok beats the alternative: unchecked "third-party slop" flooding the platform.

He invoked "stationary bandit theory," an economics concept suggesting that long-term players like Musk have incentives to maintain some level of transparency and fairness. Short-term opportunists, by contrast, just exploit systems and disappear. A bandit who plans to stick around has reasons not to burn everything down.

Buterin's Complicated Relationship With X

This isn't exactly a full-throated endorsement of Musk's platform. Earlier this month, Buterin criticized X for abandoning its supposed mission of promoting free speech, instead becoming a venue for "coordinated hate sessions."

He's also accused Musk of "actively tweaking" algorithms to boost certain content while suppressing other voices.

What's Coming Next for Grok

Last month, Musk teased Grok 5, scheduled for early 2026. He claimed the new model will be "extremely" intelligent and fast, with roughly a 10% chance of reaching human-level intelligence.

Whether that happens or not, Buterin's take suggests that even imperfect AI fact-checking might be better than no fact-checking at all when you're dealing with the chaos of social media. It's a pragmatic, if somewhat resigned, perspective on where we are with online discourse.