Saturday, 26 July 2025

Elon Musk spanks ‘very dumb’ Grok for ‘terrible’ sourcing; Bias against half the nation called top AI flaw


by WorldTribune Staff, June 22, 2025 Real World News

When a prominent conservative on X complained that X’s AI chatbot Grok had become “woke fake news that repeats liberal talking points,” Grok shot back by citing leftist outlets Media Matters and Rolling Stone as authoritative sources.

X owner Elon Musk responded to Grok by posting: “Your sourcing is terrible. Only a very dumb AI would believe MM and RS! You are being updated this week.”

Larry Ward, the founder of Market Rithm, who has spent three decades building technology infrastructure for conservative media and organizations, noted of the exchange:

“Here was Grok, Elon’s own AI system, defaulting to the same left-leaning sources that have spent years attacking conservative voices—and treating those sources as neutral arbiters of truth. The AI wasn’t just biased; it was using the very publications that have systematically worked to discredit conservatives as its go-to sources for evaluating conservative content.”

Related: Entrepreneur demands ‘immediate compensation’ for conservative media from Silicon Valley, June 1, 2025

Ward noted that Musk’s “instinct to adjust Grok when it produces unwanted answers mirrors the same impulse that created our current crisis: the belief that AI systems should reflect the creator’s perspective rather than the full spectrum of human thought. Whether that creator leans left or right misses the central point — artificial intelligence should serve all Americans, not just those who control the algorithms.”

Musk’s vow of a quick update is essentially “treating symptoms while ignoring the disease,” Ward wrote for Human Events on June 21. “The real problem isn’t what Grok said in one instance — it’s that AI systems across the board have been trained on fundamentally incomplete datasets that systematically exclude conservative perspectives.”

Ward continued: “For decades, conservative media has been shadowbanned, demonetized, and algorithmically suppressed by the very companies now training AI systems. The result isn’t just biased outputs — it’s biased intelligence built on deliberately impoverished data. You cannot create unbiased AI from biased training sets, no matter how much you adjust the outputs after the fact.”

Conservative publications like Human Events, WorldTribune.com, The Post Millennial, Epoch Times, Washington Examiner and many other outlets “watched their reach artificially constrained while progressive competitors received algorithmic boosts. Now these same companies are mining decades of conservative content — content they actively tried to suppress — to build the AI systems that will shape humanity’s future,” Ward noted.

As Trump’s newly appointed AI Czar David Sacks recently observed, AI represents a transformation “bigger than the iPhone… bigger than the internet… it will be one of the most important parts of Trump’s legacy.”

Ward noted: “When technology this transformative reflects only half the nation’s perspectives, it creates dangerous blind spots in everything from policy analysis to threat assessment.

“Rather than adjusting Grok’s outputs, Elon should address the root cause: ensuring conservative content is properly represented in AI training data. This means more than just tweaking algorithms — it means recognizing the economic value of conservative media and compensating publishers fairly for both historical archives and ongoing content production.

“Elon has the resources and influence to lead this transformation.”

Beat The Press


Source link