The oldest problem of democracy is how do we come to agreement about the 'facts'. We now live in a world where scientists are unable to agree on the nature of reality, and the proving grounds of politics are infected with fake news, deep fakes, and a decline of the institutions meant to keep them in check. Journalist Narain Batra argues, looking at the US example, we are entering a world of epistemic fragmentation where the silos of the internet have entered the real, physical world. A return to a stoic rejection of sophistry is the only way out.
On November 3rd, tens of millions of Americans will vote in the first midterms of the second Trump term, one might say, the first national test of whether anyone still agrees on what's true enough to vote on. Whatever gets fabricated or denied between now and November won't stay limited to a news cycle. It will show up on a ballot.
The extraordinary fact at the heart of American journalism began not with new technology but with a phrase uttered early in January 2021. Trump repeated the words “fake news” so often that they stopped functioning as a complaint about individual stories. Rather, they started functioning as a challenge to journalism itself.
Trump grasped something his that none of his predecessors had understood or even had access to. A president no longer needs journalists to interpret him for the public, he has the capacity to talk to the public directly. A tweet at four-thirty in the morning can set the day's agenda before journalists smell their coffee. Once a president can bypass editors, fact-checkers, and producers, calling the press "fake" stops being mere criticism. It becomes the rationale for a direct, unchecked line to the public. He doesn't need to prove a story false but now only needs to persuade us that regardless of their status, journalists can't be trusted. Frequent putdowns and insults diminish journalists. It becomes easy to circumvent them.
Trump’s Truth Social formalized the bypass. A sitting president can now publish statements, images, and increasingly AI-generated video directly, with nothing standing between impulse and audience. Trump has posted a long, escalating run of AI-generated imagery targeting political rivals over the past year. Wired dubbed him America's first generative-AI president.
Elon Musk is the other half of this story, and he arrived by a different route. Where Trump proved a president could talk around the press, Musk bought platform that much of the global conversation runs through. His stated reason for acquiring X (formerly Twitter) was seemingly a liberal one. He argued that centralized control over speech is itself a danger, and platforms should tilt toward openness. There's a real principle in it that the government censorship is extremely dangerous, and X is not the government.
But a platform doesn't have to be the government to matter so colossally. When a handful of privately owned platforms become the infrastructure through which a democracy talks to itself, "it's just a business decision" starts to ring more hollow.
Musk then folded his AI company, xAI, into X, producing something structurally new, a closed loop, with no independent checks from outside it. Someone posts a false claim on X. Grok reads it, because Grok is built into the same platform. A user then asks Grok whether the claim is true, and Grok answers inside that same feed. The chatbot ends up doing two jobs at once, reading the platform's content, and adding to it.
Want to continue reading?
Get unlimited access to insights from the world's leading thinkers.
Browse our subscription plans and subscribe to read more.
Already a subscriber? Log in
Join the conversation