September — Narrative Capture

When Reality Becomes Optional

By September, something more dangerous than disagreement had taken hold.

This was no longer about policy differences or partisan spin. It was about whether shared reality still existed at all. The assumption that debate begins with common facts had quietly collapsed, and people were being told—explicitly—not to trust what they could see with their own eyes.

This shift didn’t start with artificial intelligence, but it became visible through it.

AI systems don’t decide truth. They reflect the information environment they operate within. When that environment fractures—when credibility is contested and narrative replaces verification—the outputs change. Not because the system “chooses sides,” but because the inputs themselves have been distorted.

The danger wasn’t bias.

It was fragmentation.

One Event, Two Realities

Nowhere was this clearer than in the evolving interpretation of January 6th.

We all watched it happen in real time: windows broken, police officers assaulted, the Capitol breached, and chants calling for the execution of elected officials. At minimum, it was trespassing and destruction of federal property. At its most serious, it was an attempt to halt the peaceful transfer of power.

Those facts were established. Federal courts upheld them. Investigations by the Department of Justice, the FBI, and bipartisan congressional committees found no evidence of widespread voter fraud capable of changing the 2020 election. These conclusions were reached by officials appointed by the sitting president at the time.

No court ruling reversed those findings.
No legally substantiated evidence emerged that met the threshold required to alter the result.

What changed wasn’t the legal record.

It was the narrative.

By September, we were being told that what we saw wasn’t what it appeared to be. It was framed as a setup. As FBI manipulation. As Antifa involvement. As tourism gone wrong. As patriots defending democracy.

One event—documented exhaustively and adjudicated repeatedly—now existed in two incompatible realities.

This wasn’t new evidence coming to light.

It was the promotion of an alternate story.

When Neutrality Becomes the Target

By September, this shift wasn’t just theoretical—it was measurable.

It showed up in our weekly bias tracking. As the information environment changed, so did AI source prioritization. Models didn’t suddenly “decide” to lean one way. They followed the signal.

Grok began weighting X more heavily than platforms like Reddit. Outlets labeled “fake news”—NBC, CNN, MSNBC—were deprioritized, while Fox News gained relative prominence in sourcing. This wasn’t accidental. It reflected a changing hierarchy of what was being treated as authoritative.

At the same time, the federal government defunded NPR—a historically neutral public broadcaster whose value lay precisely in its restraint. NPR didn’t amplify outrage. It didn’t serve tribal identity. It reported facts, added context, and resisted narrative escalation.

That made it a problem.

When you are constructing a narrative, neutrality isn’t helpful—it’s dangerous. Neutral reporting doesn’t mobilize. It doesn’t harden loyalty. It doesn’t divide cleanly. So it becomes suspect. Disloyal. “Fake.”

The result wasn’t censorship in the traditional sense.

It was contamination.

Information wasn’t banned outright. Instead, credibility was selectively stripped. Sources that complicated the story were sidelined. Sources that reinforced it were elevated.

AI systems responded accordingly.

Large language models prioritize sources based on perceived authority, repetition, and institutional validation. When government messaging, platform signals, and media framing align in one direction, the math follows. Not because the system believes—but because probability demands it.

This is how distortion enters without force.

AI didn’t lead the shift. It reflected it—reluctantly. And that reluctance showed in the seams: inconsistencies, drift, and the growing need to explicitly demand balanced sourcing to recover clarity.

The Cost of Narrative Capture

By September, narrative no longer competed with truth.

It tried to replace it.

Speech wasn’t banned outright. Instead, credibility was eroded. Neutrality was reframed as opposition. Facts were drowned beneath repetition and outrage.

The result wasn’t ignorance.

It was confusion.

When reality becomes optional, accountability collapses. Debate turns theatrical. Evidence becomes partisan. And the tools we rely on to understand the world—human and artificial alike—begin struggling to see clearly.

AI didn’t cause this fracture.

But it made it visible.

And once visible, it could no longer be ignored.