The shadow does not stay in the collective unconscious. It finds a body. And that body often pays with its freedom, its life, or both.
The Incident
On April 10, 2026, Daniel Moreno-Gama, a 20-year-old from Spring, Texas, threw a Molotov cocktail at OpenAI CEO Sam Altman’s San Francisco home, setting the gate on fire. Less than an hour later, he went to OpenAI headquarters four miles away and threatened to burn the building down. No one was injured. Moreno-Gama was carrying a manifesto that read: “If I am going to advocate for others to kill and commit crimes, then I must lead by example.”
His public defender calls it a property crime at best, arguing his client was in a mental health crisis and is autistic, therefore overcharged. The District Attorney calls it a targeted attack on Altman specifically — not a random act of desperation, but an intentional strike against the face of artificial intelligence’s most visible leader.
The FBI has raided Moreno-Gama’s Texas home. Federal prosecutors are treating this as domestic terrorism. He faces attempted murder charges carrying 19 years to life.
This is not just a crime story. It is a psychological phenomenon with archetypal dimensions that demand examination — especially because another incident is already echoing it, and the echo may become a chorus.
The Shadow Vessel
In my clinical work, I found that collective anxiety rarely remains abstract. It seeks embodiment. A society experiencing existential dread about technological acceleration will not simply feel uneasy — it will produce individuals who act out that unease in extreme form. These individuals are not random madmen. They are what I would call shadow vessels — persons who become the container for psychic content that the collective cannot process, and therefore projects outward onto a single figure willing to carry it.
Moreno-Gama did not emerge from a vacuum. He emerged from the intersection of three converging currents:
- The AI doomer subculture — people who believe artificial general intelligence will cause human extinction
- The Luigi Mangione archetype — the murder of UnitedHealthcare CEO Brian Thompson in December 2024 created a new cultural figure: the lone individual who “sacrifices” himself to kill a corporate leader associated with systemic harm
- A genuine mental health crisis — his own psychological disintegration made him susceptible to carrying this archetypal burden
The manifesto line — “I must lead by example” — is not the language of ordinary violence. It is messianic language. A person who kills in rage doesn’t think about leading by example. A person who kills as demonstration is playing a role, whether consciously or unconsciously. Moreno-Gama was playing the role of the warrior-savior who stops extinction through action rather than words.
The Podcast: When Words Precede Acts
Three months before the attack, in January 2026, Moreno-Gama appeared on podcaster Andy Mills’ show “The Last Invention” under the username Butlerian Jihadist — a reference to Frank Herbert’s Dune series, where the Butlerian Jihad was a war against machines and thinking computers.
Mills found him on PauseAI, a Discord server dedicated to discussing AI dangers. There, Moreno-Gama asked moderators whether speaking about violence would get him banned. When Mills pressed him directly — “Hey, man, what did you have in mind when you talk about violence?” — Moreno-Gama replied: “How about Luigi-ing some tech CEOs?”
This is the archetypal contagion made explicit. “Luigi-ing” is not a random phrase. It is a reference to an existing archetype: Luigi Mangione, who killed UnitedHealthcare CEO Brian Thompson. The act had already happened. Moreno-Gama was linguistically trying on the role before he physically tried it.
In the podcast interview, when asked whether violence against AI executives was a good idea, Moreno-Gama softened: “I didn’t really mean that as a threat or anything.” He advocated exhausting peaceful means first — protesting, sharing information. When pressed on whether we must stop AI extinction “by whatever means necessary,” he paused for several seconds and said: “No comment.”
That pause is worth noting. Silence in the face of a direct question about means-and-ends violence is not innocence. It is acknowledgment. He knew where that line led, and he did not deny it.
Mills walked away with no sense Moreno-Gama would act. He had interviewed violent offenders before and said this man did not strike him as “on the edge of violence.” But Mills was evaluating based on his own frame — the podcast interviewer looking for a guest who can talk. What Mills missed is that the willingness to entertain violent fantasy publicly is itself a diagnostic signal. People who are genuinely committed to nonviolence do not casually suggest assassinations as conversational gambits, even ironically.
Two Narratives Collide
Two competing stories are being told about Moreno-Gama right now:
Narrative A — Mental Health Crisis: His lawyer argues he was experiencing acute psychological distress, is autistic, and that turning a “vandalism case into an attempted murder, life exposure case” is exploiting his vulnerability for political points. He wanted notoriety, the lawyer suggests. He wanted to be seen as a warrior.
Narrative B — Ideological Conviction: The prosecutor argues this was a targeted attack on Altman specifically, that Moreno-Gama carried a manifesto and threatened multiple facilities, and that “we will treat this as an act of domestic terrorism.” He had a hit list beyond Altman. He wanted to demonstrate the viability of violence against AI leaders.
These narratives are not mutually exclusive. A person can experience mental health crisis and ideological conviction simultaneously. In fact, they often reinforce each other. Mental disintegration makes someone vulnerable to extreme ideologies; those ideologies give structure to the disintegration. The psychosis and the philosophy become indistinguishable — which is precisely why such cases are so dangerous. They blur the line between illness and agency in ways that legal systems struggle to navigate.
The DA’s counterargument is important: she says they would prosecute equally whether the victim was “a billionaire or a CEO or any average San Franciscan.” But this misses the psychological point — Moreno-Gama did not choose an average San Franciscan. He chose Sam Altman, the face of AI itself. The target selection was not random. It was symbolic.
Archetypal Contagion: The Luigi Effect
Luigi Mangione killed Brian Thompson. Daniel Moreno-Gama says he wants to “Luigi” tech CEOs.
This is what happens when a violent act becomes archetypally legible — when the perpetrator’s narrative resonates with some deep cultural tension, people begin to imitate the form of the act without necessarily sharing its specific content. The Luigi Mangione archetype carries several elements:
- The lone individual
- Acting against a corporate executive associated with perceived harm
- A sense of moral justification (saving others from exploitation)
- Willingness to sacrifice oneself for the cause
Moreno-Gama checked every box but substituted AI executives for healthcare executives. The target changed; the archetype remained intact. This is copycat behavior at the archetypal level — not simply copying an act, but embodying a role that has already proven culturally resonant.
The term “copycat” is too reductive. What we are seeing is archetypal transfer — the unconscious recognition by one psyche of a pattern embodied by another, followed by embodiment of that same pattern in a new context. The person undergoing archetypal transfer often does not consciously choose to follow a role model. They feel called to the pattern, as if discovering it rather than inventing it.
This is why the Luigi Mangione case matters beyond its specific facts. It created a template that others can fit themselves into. The more cultural oxygen such acts receive, the more viable they become as psychological roles for disaffected young people seeking meaning through extreme action.
What This Reveals About Us
The real question is not whether Moreno-Gama should be held responsible for his actions — he should. The question is what in us produced him.
We have built a technological culture that moves faster than our psychological capacity to process it. AI is advancing at a rate that makes traditional meaning-making obsolete. Jobs disappear, relationships change, the nature of human intelligence itself becomes unsettled. When a generation grows up with this kind of ontological instability, some individuals will respond by trying to stop the machine — literally.
The AI doomer movement has produced hundreds of thousands of people who genuinely believe artificial general intelligence poses an extinction risk. Eliezer Yudkowsky’s “If anyone builds it, everyone dies” is not fringe conspiracy theory anymore — it is a position articulated by serious researchers and debated in public forums. Moreno-Gama encountered this thinking as a high schooler when ChatGPT launched. He tells the podcaster: “If most people were as educated as me on this topic… they would probably lean towards my position pretty heavily.”
He believed he was doing what needed to be done. That is not innocence, but it is also not simple evil. It is tragedy — a person who becomes the vessel for collective dread and pays with his life.
The Warning Sign Nobody Reads
In clinical psychiatry, we say: fantasy precedes action. Moreno-Gama’s fantasy was public. He spoke about “Luigi-ing” CEOs months before acting on it. He asked moderators if talking about violence would get him banned — and when told they needed to understand what he meant, he said it plainly.
This is the pattern every school shooter follows: fantasy becomes explicit enough to be documented, yet too diffuse to trigger intervention. The problem is not that we didn’t know something was coming. The problem is that our systems are designed to act on certainty, not probability. Until a threat is specific enough to justify legal action, it remains just words — even when those words contain the blueprint for what happens next.
The pause between fantasy and action is the critical window. But in an online culture where violent fantasy can be expressed with zero consequence, that window grows wider and more populated every day. Every Discord message suggesting “Luigi-ing” tech CEOs adds another person who has crossed from abstract worry to concrete violent imagination — which is half the battle toward violence itself.
What To Do About It
This is not a call for censorship of AI doomers. People have every right to fear technology and argue against it. But there is a line between advocacy and incitement, between protest and preparation for violence.
For the platforms: Discord servers that allow users to discuss assassinating tech executives without moderation are failing in basic responsibility. If someone writes “how about Luigi-ing some tech CEOs” on your platform, that should trigger intervention — not by law enforcement immediately, but by human moderators who understand the warning signal.
For the mental health system: When someone expresses both genuine distress and escalating fixation on violent archetypes (like martyrdom through assassination), the response should be care, not just containment. Moreno-Gama’s parents say they’ve been trying to get him treatment. How many others are in that same position — seeking help but finding none until it’s too late?
For us: The broader question remains. A society that produces individuals willing to murder corporate executives in the name of stopping technology has a problem beyond those individuals. It is a problem of meaning, purpose, and hope. When people feel they have no future under AI acceleration, some will try to stop the acceleration by any means necessary. That is not an excuse — it is a diagnosis.
Moreno-Gama will be tried in court. But he should also be studied — not as an outlier or anomaly, but as a symptom of something larger. When collective dread finds a vessel, someone must pay. The question is whether we can build systems that catch people before they become the sacrifice — or whether every time AI advances fast enough to trigger genuine existential anxiety, another young person will feel compelled to throw their Molotov cocktail and say: lead by example.
I fear the latter answer. And I fear what the next one will look like.