Sam Altman's Attacker, In His Own Words

Our Interview with Daniel Moreno-Gama

Listen on: Apple | Spotify | YouTube

Back in January, I had a conversation with someone named Daniel Moreno-Gama. He was a 19-year-old Texan with a part-time job, taking classes at a community college. He was worried about AI, and what he saw to be the impending extinction of humanity.

If that name sounds familiar, it’s because this week he was charged with attempting to kill OpenAI CEO Sam Altman.

My colleagues and I (shoutout Seth Temple Andrews) had found Moreno-Gama on a Discord server called Stop AI, where he was posting under the username Butlerian Jihad (apparently he is a Dune fan). We were reporting for an episode of The Last Invention, where we’ve been covering all sides and all the beliefs shaping our world’s view of this fascinating, bewildering technological moment. In these debates, Moreno-Gama belongs to a small but hardline group: He believes that the time has come to use violence to stop AGI. In a post on Discord that caught our attention, he was asking the group whether he would get banned for talking about violence.

One of our reporters sent him a direct message asking him to elaborate, and he said he was interested in “Luigi-ing some tech CEOs.”

We asked if he would be willing to sit down for an interview and were surprised when he agreed. We conducted the interview remotely in late January.

Today we’ve published an edited version of that interview, with the man who is now accused of being an attempted assassin.

As Moreno-Gama told me his backstory, I realized that even before he was charged with 13 felony counts, he was, in a way, an embodiment of our society’s fears about the social experiment we’ve been running on Gen Z. He got his first iPad around age 10, and quickly became a big YouTube fan. He explained that he got into politics as a preteen, watching Ben Shapiro videos and getting into fights under an anonymous username in comment sections. Before long, he’d moved on to being a Bernie Sanders fan before discovering the videos of prominent AI safety advocates like Eliezer Yudkowsky.

“What really got me into it,” he said of AI in our interview, “was the arguments of people like Yudkowsky, Max Tegmark, Connor Leahy, et cetera. I quickly noticed that a lot of their arguments not only made a lot of sense, but most of the time didn’t have very good responses from the other side. And that began to really worry me.”

He told me that he had come to believe the creation of AGI would likely lead to the end of the human race, and that if only more people knew what he knew, they would agree and band together to stop it before it was too late.

When I asked him directly about what lengths people should be willing to go to in order to stop this apocalypse, he became uncomfortable.

“Do you think that if we continue to see the industry move in the direction it’s moving now, that we have to stop the extinction of the human race by whatever means necessary,” I asked.

“Hmm. I’ll say no comment,” he replied after a long pause.

When I pressed him further, asking whether he thought someone should kill Sam Altman, he paused, then answered: “Um, no, I think…one person is not really gonna do that much of a dent.”

However, last week, he attacked both the home of Sam Altman and the main headquarters of OpenAI. When confronted by security, he allegedly said he had come to burn the building to the ground and kill anyone inside.

Law enforcement says he had a three-part manifesto on him when arrested, in which he makes clear that his hope was to inspire others to follow his lead. He had included the names and addresses of other AI CEOs and investors.

Back in January, even as Daniel told me this sort of violence wouldn’t make a difference, he said something that stands out now in a way that it didn’t then. When I asked about Luigi Mangione – who has gained a cult following after he was arrested and charged with the murder of health insurance CEO Brian Thompson – he answered that while he didn’t agree with what Mangione did: “I think what we saw with Mangione is a lot of people were able to excuse it. I think that’s interesting.”

In recent days, many prominent AI safety advocates – including organizations like Pause AI and figures like Yudkowsky – have released statements strongly condemning the attack and urging impressionable young people not to follow Moreno-Gama’s lead.

In a court appearance on Tuesday, Moreno-Gama’s lawyer said he was experiencing a mental health crisis and that he is autistic.

“It is unfair and is unjust for the San Francisco district attorney and the federal government to fear monger and to exploit the mental illness of a vulnerable, young man by turning a vandalism case into an attempted murder,” the lawyer, Diamond Ward, said.

I invite all interested to listen to Moreno-Gama in his own words to come to their own conclusion on his mental state, at least in January. To me he seemed to be an earnest, cogent 19-year-old.

Online, anonymous users on places like Reddit have already begun comparing Moreno-Gama to Luigi Mangione – with some even calling for someone to finish what Moreno-Gama started. “This should be a nightly occurrence,” read one social media post after the attack.

Could what Moreno-Gama allegedly did be just the first of what may become a trend? Well, the existential fears he expressed are not fringe. It’s worth noting that his assessment of the risk posed by AGI is actually in step with what many leading AI researchers have said – even some of the CEOs of the companies he targeted for attack. But throughout history, existential fear has always needed a face: communism, nuclear power, immigration. And the backlash building against AI isn’t only existential – for a growing number of young people who already feel locked out of the economy, AI represents something more immediate: the reason they can’t get a job, and the billionaires getting rich from that fact.

I ended the interview by asking Moreno-Gama if he thought there were many others like him, who wanted to see violence done toward the creators of AI. He responded: “People may feel that way, but I don’t know. Not too many people would do it.”

If you have any guest suggestions or questions you’d like us to investigate, send an email to: hello@longview.report

- Andy

(a longer version of this post also appeared in The Free Press)