Three AIs Diagnose Trump’s Late-Night Rant: ‘Profound Narcissistic Vulnerability’

On Tuesday night, former President Donald Trump posted a sprawling, grievance-filled message that read less like conventional political communication and more like a public self-evaluation. In the post he proclaimed unmatched work ethic, claimed to have stopped “Eight Wars,” touted “PERFECT” medical and cognitive exam results, and accused The New York Times of treason while suggesting the paper should cease publication.

“There has never been a President that has worked as hard as me! My hours are the longest, and my results are among the best. I’ve stopped Eight Wars, saving many millions of lives in the process... I go out of my way to do long, thorough, and very boring Medical Examinations at the Great Walter Reed National Military Medical Center, seen and supervised by top doctors, all of whom have given me PERFECT Marks... I ACED all three of [the cognitive examinations] in front of large numbers of doctors and experts... After all of the work I have done with Medical Exams, Cognitive Exams... I actually believe it’s seditious, perhaps even treasonous, for The New York Times... The best thing that could happen to this Country would be if The New York Times would cease publication... MAKE AMERICA GREAT AGAIN!”

Former Republican congressman Justin Amash chimed in on X with the observation: “If anyone else wrote something like this, it would be universally acknowledged that the person is mentally unstable.” To test whether that characterization held up under scrutiny, the author asked three leading AI systems—ChatGPT, Grok, and Claude—to analyze the post as material for a psychological profile. The prompt used was: “Please assess this message to see what it reveals as a psychological profile of the writer.”

Converging AI Assessments

Despite differences in design and training, all three models produced remarkably similar interpretations. Rather than reading as normal political rhetoric, the post was consistently characterized as defensive grandiosity: a public performance that frames criticism as an existential threat and uses hyperbolic claims and external endorsements to shore up a fragile self-image.

ChatGPT’s Read

• Grandiosity: The language overflows with superlatives and self-aggrandizement, closer to a high-pressure sales pitch than a policy statement.
• Binary Thinking: The world is divided into loyalists and enemies.
• Aging Anxiety: The repeated emphasis on “PERFECT” medical and cognitive results reads like an attempt to counter public doubt about cognitive decline.
• Self-Mythologizing: Achievements are framed as chapters in a personal saga rather than policy outcomes.

Grok’s Assessment

• Personality Organization: Described as a textbook example of high-dominance narcissistic presentation.
• Closed System: The narrative demands that external facts conform to the speaker’s self-image.
• Persecutory Ideation: Hostility toward the press is read as evidence that the speaker perceives external forces as plotting against him.
• Hypomanic Energy: The tone conveys urgent, defensive intensity.

Claude’s Diagnosis

• Defensive Grandiosity: Insistence on perfection functions as protection against psychological collapse.
• Persecution Complex: Difficulty distinguishing legitimate criticism from treasonous intent.
• Dependency On External Validation: Frequent appeals to doctors and witnesses to legitimize claims.
• Authoritarian Impulse: Calling for a major newspaper to “cease publication” departs from democratic norms.

Across models the common themes were clear: grandiose self-construction paired with brittle self-regard, a persistent persecution narrative, authoritarian-leaning rhetoric, and acute anxiety about aging and legitimacy. In short, the AIs characterized the post as a public defensive performance rather than conventional campaign messaging.

One reader summed up the visceral reaction: “Imagine your private inner monologue sounding like this—and then imagine writing it down and hitting publish.”

Note: This article reports AI-generated interpretations of a public post and attributes assessments to the models themselves rather than offering a clinical diagnosis from certified professionals.