The lights dimmed as 5 actors took their locations round a desk on a makeshift stage in a New York Metropolis artwork gallery turned theater for the night time. Wine and water flowed by the intimate area as the home — filled with media — sat to witness the premiere of “Doomers,” Matthew Gasda’s newest play that’s loosely based mostly on Sam Altman’s ousting as CEO of OpenAI in November 2023.
The play fictionalizes occasions that occurred after OpenAI’s co-founder and former chief scientist Ilya Sutskever knowledgeable Altman he was fired — a call the board remodeled issues that the CEO was mishandling AI security and interesting in abusive, poisonous conduct. Regardless of the apparent meticulous analysis that went into Gasda’s depiction of that night time, the playwright instructed TechCrunch his aim wasn’t to create a documentary, however quite to make use of that setting as a microcosm for the larger philosophical questions of AI security and alignment.
People have for millennia created fable and lore round humanity’s subsequent nice innovations and the dangers of pursuing them. Like Prometheus stealing fireplace and Oppenheimer splitting the atom, humanity can’t resist the lure of its personal innovations. With Gasda’s play, the humanities are actually weighing in on the philosophical debate round speedy technological innovation — cementing know-how and its barons into the zeitgeist.
“The humanities, the arts, we can say something about this,” Gasda instructed TechCrunch. “We’re maybe toothless financially and toothless technologically, but we’re not toothless in the way that we have the right to represent this world as much as anyone else.”
In Gasda’s play, the corporate known as MindMesh, and the egotistical, infantile, spurned CEO is called Seth.
The primary act happens in Seth’s “war room” as he and people closest to him debate the deserves of the board’s ousting, what their subsequent steps ought to be, and whether or not the CEO is correct to doggedly pursue such society-altering know-how. The second act takes place in MindMesh’s board room and descriptions the varied fears amongst its members, together with that the newly-ousted Seth would possibly take retribution on those that betrayed him, and that “we’re gonna get wiped out by a competitor species.”
The central pressure of the play is one which’s taking part in out on the world stage in the present day – the existential risk of AI versus the existential promise of it.
‘I was fired for creating miracles’
Gasda says he wrote 35 drafts of this play, which he previewed for early audiences again in August. After extra journeys to San Francisco — and plenty of Celsius-fueled writing periods later — he arrived with the “Doomers” model that’s premiering in New York this weekend by February and shall be proven in San Francisco in March.
Gasda, who is thought for writing and directing “Dimes Square” and “Zoomers,” instructed us he needed to grasp character archetypes and the psychology of a gaggle of those who don’t essentially “engage in self-reflection.”
The result’s a solid of 10, half of that are based mostly on actual individuals akin to OpenAI CEO Sam Altman, former chief technologist Mira Murati, and co-founder and president Greg Brockman. Murati served as interim CEO in the course of the govt shakeup. She left the corporate in September 2024 to begin her personal firm.
Different characters are based mostly on Helen Toner and Adam D’Angelo, two former board members who voted to oust Altman; and even Eliezer Yudkowsky, a researcher who has known as for OpenAI to be shut down earlier than it ends the world.
Gasda mentioned that Seth, the character based mostly on Altman, is probably essentially the most true-to-form in his depiction, however he additionally left room for fictional portrayals of characters these conversant in the Bay space will acknowledge — a callous VC who thinks porn is the way forward for AI, a newly-minted Gen Z millionaire founder, and a know-it-all lawyer from Stanford.
“I wanted to extract enough sense of fidelity and sense of realness to make the play challenging and to make the characters effectively real enough that it won’t turn off people who actually know what happened or know what a board room at an AI company is like,” Gasda mentioned, noting that Altman was despatched a duplicate of the play earlier than it premiered.
The questions the solid debate are pertinent: Ought to AI growth be sped up so “we” can win; ought to its growth be slowed down to permit for higher security and alignment; ought to it’s shut down altogether to guard the human race?
By means of these debates, we see the archetypes of every character fulfilled: If there’s a selection between profitable and being ethical, Seth, the character based mostly on Atlman, chooses to win.
He declares loudly that the board fired him “for creating miracles,” and argues that alignment can be a “poor use of a sacred resource.” It’s human, he says, to pursue excellence and provides that MindMesh is the world’s “immune system,” a benevolent American-made AGI that may shield us when the “bad” AGI goes rogue.
“The only thing to do is outcompete and out-engineer,” Seth says. Characters based mostly on Mira Murati and Greg Brockman largely again Seth, at the same time as he insults them, arguing for a imaginative and prescient of an AI utopia the place know-how cures illness and opens up interplanetary area journey. To which the security ethicist character, Alina, says, “You make it sound like a genie in a bottle.”
Gasda sprinkles dry humor all through “Doomers” — lightening the temper of an in any other case tense material. He additionally introduces humor by capturing nuances of Silicon Valley tradition. Polycules and ketamine have been talked about greater than as soon as, and at one level, the characters casually take mushrooms. There are references to Waymo robotaxis, and at one level a personality remarks, “I know drinking is low status, but I really need a drink,” in reference to the Cali-sober development overtaking the Bay.
Security questions stay
The occasions depicted on this play occurred virtually 15 months in the past, and already the dialog round AI has shifted because the race for domination outpaces issues of safety.
Altman ended up instantly returning to energy after OpenAI engineers threatened to give up en masse if he wasn’t reinstated. A brand new board that’s comfy with OpenAI shifting right into a for-profit construction has since consolidated underneath the CEO. Sutskever and Jan Leike, the co-lead on OpenAI’s now-defunct superalignment workforce, have defected. Different safety-focused researchers who raised issues about AI labs have additionally departed.
That hasn’t damage OpenAI.
The corporate is reportedly elevating a $40 billion spherical that will worth it at $300 billion, whereas President Donald Trump guarantees to guard AI from regulation as a new arms race towards China heats up and new rivals, like DeepSeek, enter the ring. In brief, AI innovation is rushing up, not slowing down, simply as Seth’s character needed. The query everybody awaits the reply for is whether or not or not it is a good factor.
“It’s ugly to build God,” Alina, the ethicist within the play, says. “Because we’re so ugly, and it’s based on us.”
