Inside a curved glass building next to the Golden State Warriors’ arena in San Francisco, eight cans of Spam with tiny arms whirred to life, tapping out artificial-intelligence-generated word slop on miniature keyboards. They were part of the Misalignment A.I. Museum, a gallery dedicated to A.I.-inspired art.
Across town, in the basement of a Lower Haight boutique, a group of tech workers delivered stand-up comedy sets about programming languages, ChatGPT and Nvidia’s stock price for Artificially Unintelligent, a tech-themed comedy show.
And a month earlier, on a foggy summer night in San Francisco’s Glen Park neighborhood, a group of tech workers gathered at a midcentury house being used as a start-up office for a reading of “Doomers,” a new, ripped-from-the-headlines play about the weekend that Sam Altman, the chief executive of the start-up OpenAI, was briefly fired.
A.I. is providing the art and entertainment worlds with plenty to fear, from potential copyright violations on a global scale to the loss of jobs taken by a soulless machine. But A.I. is also quickly becoming fodder for the creative community.
The technology has long been a staple of science fiction, but now, two years into the boom kicked off by OpenAI’s ChatGPT, the issues raised by those movies and books all feel a little more real. More artists, playwrights and comedians are finding inspiration in the A.I. technology that’s currently available: its ethical quandaries, its impact, its risks, its absurdities and even its executives.
Audiences are eating it up. “Doomers” is set to have an official run early next year in New York and San Francisco, after raising funding via Kickstarter. The Misalignment A.I. Museum, which started in 2023 as an eight-week pop-up, is set to move into a larger, permanent space in San Francisco’s Mission Bay neighborhood and reopen in February. And Artificially Unintelligent has become a monthly event and is seeking to expand into other cities.
Some of the creative work about A.I. is coming from tech industry insiders laughing at the hype or raising alarms about the dangers. Some of it is coming from outsiders fascinated by the industry’s intrigue.
A.I. has become a meaty topic for artists to dissect, invoke and mock even as they fight its use as a tool in their fields. In October, more than 10,000 actors, authors and musicians signed a letter opposing the use of their works to train A.I. systems.
At least 10 groups, including voice actors, the Recording Industry Association of America and The New York Times, have sued A.I. companies, claiming the technology unlawfully used copyrighted work without permission. And many artists, from graphic designers to movie stars, have expressed concern that the A.I. could soon replace them.
“It’s definitely a big topic that’s affecting us all, and we should have everyone who is affected by it be part of the conversation,” said Audrey Kim, the museum’s founder and curator.
From Misalignment to Marie Antoinette
Ms. Kim has been thinking about A.I. for at least a decade, starting when she worked in operations at Cruise, the autonomous vehicle company now owned by General Motors. Her colleagues would have lunchtime debates about the role of A.I. in society. In recent years, as that debate became more mainstream, Ms. Kim decided that art could be a way to bridge the gap between the people building A.I. and the rest of us.
Initially her museum, which is a nonprofit supported by donations, focused entirely on addressing the risks of A.I. But through discussions with people working in tech, Ms. Kim, realized there was no consensus among techies on what A.I.’s risks to humanity were, or even a universal understanding about how it worked. So she shifted to education and asking the big questions.
One exhibit is a phone booth made for conversations with an A.I. version of the television legend Fred Rogers. The exhibit breaks down all the different pieces of technology that were used to create it, including the speech-to-text converter that interprets the visitor’s voice, the A.I. system — called a large language model — that creates the Fred Rogers-like personality and the text-to-speech model that simulates his voice, “so it’s not just like this nebulous, vague A.I.,” Ms. Kim said.
A sculpture in the shape of two humans embracing is made of paper clips, evoking a popular theory of A.I.-induced doom. (Told to make paper clips, an A.I. model could decide that humans get in the way of making more paper clips and kill them.)
“I don’t think it’s going to happen,” Ms. Kim said of the paper clip apocalypse while giving a tour of the museum, “but one of the main dangers of A.I. is not having a stop valve.”
A piece called “Sonosynthesis” features a grand piano that plays A.I.-generated music in response to close-up images of bacteria. It’s designed to provoke discussions about who should get the rights to the music: the researchers who built the A.I., the artists who created the music used to train the A.I. or the scientists who grew the bacteria that inspired the music.
Many of the pieces, like an A.I. palm reader or a Broomba (a broom on a Roomba vacuum), are playful. A clown puppet that cries for help in the visitor’s voice is terrifying.
Other exhibits are purely aesthetic. Ms. Kim unfurled a pair of seven-foot neon pink and blue tapestries called “Marie Antoinette After the Singularity,” designed using A.I. tools by the musician Grimes, who has encouraged fans to make A.I. music using her voice. To create the pieces and advise on the placement of the crests, Ms. Kim contracted Factum Arte, a group in Madrid whose nonprofit arm refurbished Henry VIII’s tapestries.
Ms. Kim, who embodies the museum’s winking, whimsical spirit on tours, considers herself a tech optimist. “The future is not set,” she said. “It’s what we’re doing today that’s going to make that or break that.”
‘Millennials with a ton of power’
“Doomers” takes a different view. Matthew Gasda said he had been inspired to write the play after following Mr. Altman’s firing on social media last year. The abrupt ouster set off a public debate over potential dangers of OpenAI’s technology and whether Mr. Altman was ignoring them.
While researching those conversations, Mr. Gasda, 35, decided that the people building this new technology were not any more qualified to have debates about its role in society than anyone else.
“It’s just millennials with a ton of power,” he said.
Mr. Gasda’s previous plays were similarly of the moment. “Zoomers” was about young people in New York. “Dimes Square” was about the city’s postpandemic downtown scene. He often stages casual readings of drafts of his scripts, then collects audience feedback and adapts. He told the audience in San Francisco that this process might be familiar.
“You all know about beta testing and releasing products that don’t work,” he said to laughter.
That version of “Doomers” hit on tropes familiar to anyone who has followed tech culture in recent years: polycules, Waymo, p(doom), ketamine. The characters, based on Mr. Altman and his peers at OpenAI, had heated ethical debates about benevolence and abundance. At one point, Mr. Altman’s character (called Seth in the play) grabs a knife and demands that someone stab him because doing so would lower the probability of A.I.-induced doom.
Some of the early viewers worked at Anthropic, a rival to OpenAI, and they agreed with the play’s message that A.I. poses a threat to the world. Mr. Gasda said the wave of resignations at OpenAI in recent months further validated his thesis.
“I told the actors, ‘I think we’re onto something here,’” he said.
Court jesters
At Artificially Unintelligent, the threat from A.I. was not quite as dire. Onstage, the host, Neal Patel, surveyed the audience of 120 about their roles in the industry, cheering on the “value-add engineers” and poking fun at the people with “nontechnical fake email jobs.” To the founders, he asked, “Why are you here and not building?”
He joked that A.I. would soon take everyone’s jobs.
Mr. Patel, 24, started his comedy group, Not-So-Daily Stand Up, in part because he and his fellow techie comedians kept hearing other comedians make jokes about technology that misunderstood how it worked. The group’s roster of around 55 comedians all have day jobs in tech. (Mr. Patel is an engineer.)
“The credentials back up the jokes,” he said.
Even as A.I. threatens to replace workers, including engineers, Mr. Patel said, a common theme among his peers is how limited the technology’s abilities actually are.
“We’re all dealing with the same crap where a nontechnical manager says, ‘Can we use A.I. for that?’ And we say, ‘No.’ And they say, ‘Try it anyway,’” Mr. Patel said.
Lately, tech companies have invited Mr. Patel’s comedy group to come roast them. The setup tests whether Silicon Valley’s famous “radical candor” extends to delicate egos of those at the top. But comedians can get away with saying things others can’t, Mr. Patel said. “It’s almost like us being a bunch of court jesters.”
The post Stand-Up, Drama and Spambots: The Creative World Takes On A.I. appeared first on New York Times.
