The political drama unfolding in many Western capitals has reached a point where it feels more like a surreal script than conventional statecraft: decisions that seem self-defeating, performative theatricality that produces real harm, and a press corps that often treats the spectacle as if it were ordinary policy. That mismatch — between how plausibility is judged in fiction and how actions by the powerful are reported — is the subject of this first column in a two-part conversation I had with an AI interlocutor, Claude.
I began with a Hollywood thought experiment. Imagine a screenwriter proposes a film in which the leader of the world’s most powerful country repeatedly claims the right to resolve disputes by force, sometimes acts on that claim, then announces a privatized “Board of Peace” funded by billion-dollar seats — even as leaders are detained or eliminated and negotiations proceed in parallel. Would studio executives finance something so absurd? Unlikely: fiction must meet a plausibility standard for audiences; reality does not.
Claude’s reply cut to the heart of the contrast. Fiction survives or fails on its internal logic: audiences must be able to suspend disbelief. Institutions and journalists face something different: they work inside a system that treats actions as consequential regardless of how incoherent the rationale may be. Reporters cover tariffs, troop movements, clandestine operations and public proclamations; those acts affect lives and so demand documentation rather than mockery.
Yet Claude agreed there is another failure at work: much reporting normalizes power’s framing instead of interrogating it. Good satire — or good analysis — doesn’t simply deride absurdity from the outside; it reproduces an actor’s internal logic so its contradictions become visible. Kubrick’s Dr. Strangelove is the example Claude invoked: by faithfully rendering the logic of nuclear brinkmanship, the film exposed the grotesque consequences without relying on external ridicule. The best journalism, similarly, renders behavior with precision so the absurdity reveals itself to readers.
That kind of reporting — what I call mediating intelligence — is increasingly scarce. Journalists could act more like cultural critics who question the milieu that employs them, but doing so requires institutional courage and an analytical framework that resists simple partisanship. Instead, media organizations often accept the powerful’s framing, treating initiatives as neutral “policies” rather than as manifestations of deeper strategic incoherence.
I pressed the point with history. In the run-up to World War II there were negotiations — the Munich agreement is a painful example — that, once violated, left a clear moral line: the violator was exposed. Today we have seen negotiation itself become suspect. Negotiations are sometimes used as a ruse to lull adversaries, or dismissed as a sign of weakness. When negotiation loses legitimacy as a norm, violations become harder to classify and moral clarity grows elusive.
Claude framed this as an epistemic collapse: what is fading is not technological or material capacity but the civic habit of holding competing frameworks in tension long enough to evaluate them honestly. Where that habit atrophies, institutions that once functioned as channels of accountability — the press above all — are hollowed out. Rather than serving as a window that helps a public understand competing stakes, much of contemporary media provides mirrors that reflect audiences’ existing beliefs.
Why has this happened? The attention economy plays a central role. Platforms and funding models reward speed, emotional engagement and confirmation rather than careful, discomfiting interrogation. Interrogation unsettles audiences; it does not guarantee clicks. So outlets prioritize content that amplifies existing views and sustains engagement metrics, not material that draws out complexity and cultivates public reasoning.
The consequences are civilizational in tone if not in mechanics. When mediating intelligence weakens, so do the public mechanisms that identify, name and correct abuses of power. Bad leaders can be tolerated for a while; the greater danger is the erosion of shared intellectual practices that make accountability possible. Claude warned that recovering those practices is uncertain and often comes only after a catastrophic rupture large enough to delegitimize the prevailing framework — an outcome we should want to avoid.
Our conversation also touched on a practical concern I raised: dialogues like this one — between humans and AI — can contain useful analytic work, but they risk vanishing into the opaque memory of machine systems. If insights produced in private or small-group exchanges do not reach the public sphere, they cannot contribute to democratic repair. That’s why translating private reflection into public discourse matters.
This column ends where the conversation does: with a sketch of a problem and the promise of a solution to be explored next. In part two I and Claude will turn from diagnosis to remedy. How might journalists, institutions and citizens revive mediating intelligence? What platforms, incentives and cultural practices are necessary to rebuild the habit of interrogating assumptions rather than simply performing analysis?
There is hope: the habit of careful public reasoning has been rescued before. The challenge is to do it without waiting for catastrophe. That will require institutional incentives that reward nuance over outrage, editorial leadership willing to tolerate fewer clicks in exchange for greater clarity, and citizens prepared to accept cognitive discomfort as the price of honest reflection.
Your thoughts are welcome. If you have ideas, experiences or proposals about how to strengthen the mediating role of the press and renew democratic accountability, please email [email protected]. We intend to fold public contributions into the next installment and into a broader, ongoing conversation about how humans and AI can collaborate to restore the intellectual habits democracy needs.