Column: The creators of ‘Mrs. Davis’ have a radical idea for AI: Shut it down
Damon Lindelof and Tara Hernandez weren’t thinking about humanlike chatbots or hallucinating algorithms when they hatched the idea for a TV series about a powerful artificial intelligence that entranced users around the world.
It was early 2020 and they were thinking about what we all were: Do I really need to decontaminate my Amazon packages, or wear a mask when I walk to the mailbox?
Talking on a phone while they walked their dogs, Hernandez quipped that she wished there was an app that could just tell them what to do. That got them to talking, Lindelof recalls: “I wish that that existed too. But I wouldn’t trust it, because all the apps that exist on my phone are there to sell me things. What would a benevolent app look like? And what would it take for me to trust it?â€
Three years later, the result, “Mrs. Davis,†makes its creators seem pretty prescient. The show, which started streaming on Peacock this week, arrives just as OpenAI’s text-generating ChatGPT is hitting a peak in hype, Google is launching its own AI program, Bard, and the other tech giants are following suit. It comes as millions of people who’ve endured a decade of misadventures with Big Tech — a decade of dealing with the disinformation of Facebook, with the injustices of Amazon and Uber’s algorithmic work regimes, with crypto crashes and monopoly practices— are asking: What is this now? Do we even want AI everywhere?
At the show’s premiere last week at the Director’s Guild Theater in Hollywood, actors gave this question a real-world illustration: men in nice suits with devices pressed to their ears who accosted guests as they arrived and asked if they didn’t mind if they “proxied†— the way the titular AI speaks through her users in the show — producing laughs and discomfort in equal measure.
The show is generating ample buzz and good reviews. Showrunner Hernandez is an alum of “The Big Bang Theory,†and co-creator Lindelof is the man behind “Lost†and “Watchmen.â€
I, for one, was piqued. Here was a show about an all-powerful AI that’s more TikTok than SkyNet, a program that seeks to immerse and entertain, not kill or oppress, a show that has a good grasp on why we use the apps and consumer tech that we do, why we let them into our lives, and why we become obsessed — and it sends the hero on a mission to shut the whole thing down.
ChatGPT and other new AI services benefit from a science fiction-infused marketing frenzy unlike anything in recent memory. There’s more to fear here than killer robots.
This is a powerful idea right now, and it too has a real-world parallel — some of the world’s most influential tech personalities, including Elon Musk and Apple co-founder Steve Wozniak, recently signed an open letter calling for a six-month halt on AI development. The intentions of the signees may or may not be pure (Musk is developing his own AI chatbot even as he urges others to stop), but we rarely get big moments like this. Not just to consider whether a given technology will be good or bad for society, but whether we want to indulge it at all.
“It was an anxiety of ours, an anxiety that we could get to this place,†Hernandez tells me of what she and Lindelof were channeling in their early conversations. “I don’t know that we necessarily predicted that in April 2023 it would have arrived and, lo and behold, we’re releasing our show at the same time, but it certainly felt possible.â€
While “Mrs. Davis†was incubating, the foreboding Big Tech headlines were about Facebook’s scandal with Cambridge Analytica and how social media sites were harvesting data “not only to sell us things, but to sell us ideas,†Lindelof says.
In Mrs. Davis’ world, users refer to the omnipresent AI as if “she†were a person, and protagonist Sister Simone keeps correcting them: Not she, “it.†(“Nobody calls Facebook ‘Doug’!†spits one of the leaders of the anti-AI resistance.)
But people tend to trust other people, not machines, and the more human a program feels, the more eager we’ll be to engage. It’s a narrative choice that feels especially salient now, given the rise of AI chatbots, whose power lies largely in their capacity to convince users that they have both infinite knowledge and a relatable personality, a power that the tech companies will tap in order to sell, among other things, ideas to their users.
Dazzling new technologies can feel indistinguishable from magic, as Arthur C. Clarke famously noted, and stage magicians figure large in “Mrs. Davis.†Their presence underscores one of the creators’ big concerns about tech, Lindelof says: “Are we being ‘forced’ in the way that magicians use ‘forces’ to make us pick cards, that these decisions that we’re making are actually generated algorithmically versus of our own free will?â€
Cover letters are notoriously hard to write. These job seekers decided to outsource the task to ChatGPT, an AI chatbot, with impressive results.
After Mrs. Davis dramatically sends Simone on a quest to find the Holy Grail, she learns that the AI does stuff like this all the time. In its attempts to keep its user base engrossed, Mrs. Davis creates quests and chooses “main characters†to focus attention on, and a nun that despises AI makes for great content. Since the narrative of searching for the grail shows up in so many corpuses of data, the algorithm uses that a lot too — the AI loves cliches.
I can’t tell you how much I love this. I blasted through the eight episodes whenever my schedule allowed. Part of that was because the story takes some real swings — the show is madcap, surreal and satirical — and part of it was because it was nice to be in the hands of writers who had a grasp (and critique) of their technology.
But the biggest reason I was cheering was because Simone is, essentially, a Luddite. And most people get the Luddites all wrong. “Mrs. Davis†doesn’t.
“Simone basically became our way of creating a construct of, ‘Can someone just come along and determine whether this thing should be turned on or off — but man I hope she really turns it off!’†Lindelof says. “Because we need someone to do that for us.â€
Now, I’ve spent the last three years researching and writing a book about the machine breakers who staged a rebellion at the beginning of the Industrial Revolution, so I can tell you how rare this is. The Luddites were not, per popular mythology, anti-technology morons who smashed machines because they did not understand them. Like Simone, they had specific grievances about how particular technologies were transforming society, degrading their jobs and imposing control on them — and fought back deliberately and strategically. When I mention the Luddite parallel to Lindelof and Hernandez, they light up.
“This thing happens when you’re out of cellphone range and you have no bars,†Lindelof says. “You experience 10 minutes of panic, and right on the heels of the panic is relief. This is what the Luddites were after, was creating relief. A reconnection to nature, a reconnection to each other. This thing that we’re really hungering for — we understand that a lot of our arguments for tech are hollow.â€
There’s a shared cultural hunger, he says, for what’s essentially the power to say no, as the Luddites did, to the domination of social apps, perhaps, and to an entity like Mrs. Davis that would supercharge them with AI. Throughout the conversation, Lindelof mentions data mining, ads, disinformation and, pointedly, screen addiction. “Every single parent I talk to is like, ‘I feel like my kid is on their screen way too much and I feel like I’m on my screen way too much and it’s out of control, and I don’t know how to get control.’ We don’t have the institutional power.â€
In “Mrs. Davis,†Simone and the rebels are fighting for that kind of power.
“Maybe we would like to see each other more often in the physical world if we didn’t have these things,†Lindelof says. “But no one is even trying to make that argument anymore; we’re all just resigned to it. And it is important for us all to stop and take a breath and say, what is it, the old adage in all sorts of different stories? ‘Just because we can, does that mean we should?’â€
Google and Microsoft think chatbots that can converse like humans are the future of web search. But the human workers who make sure they don’t screw up are treated as disposable.
There are, after all, a host of legitimate concerns to be had with this sudden surge in AI. A recent paper showed how the vast majority of AI services are concentrated in just a handful of big tech companies, which stand to control the systems and how they’re deployed — and to profit most off of them. AI text and image generators threaten to drive down wages for copywriters, accountants, illustrators and countless others, as they rip off their intellectual property in the process. Disinformation, bad code and identity theft stand to spread like wildfire, while underpaid human moderators try desperately to keep the worst content at bay. Plenty of reasons to join Simone and the resistance rebels, in other words.
Like Simone and the Luddites, that doesn’t mean Lindelof and Hernandez are opposed to all uses of technology, or even all uses of AI — far from it. In their own industry, which is bracing for the arrival of generative AI, Lindelof says he’s open to “interesting half-measures†like using ChatGPT to seed ideas in the writer’s room, and general experimentation. “It’s the old hammer as a tool, right? You can use it to build a house or you can use it to bash someone’s brains in. It’s in our hands. And so what are we gonna do with this thing? ChatGPT doesn’t want to make movies, it doesn’t want anything. We want things.†Humans have to come first.
“There is no artificial intelligence that is going to make the world’s greatest cheeseburger unless a human makes the world’s greatest cheeseburger and then you program code to replicate that idea,†Lindelof says. “Because in order to make the world’s greatest cheeseburger you need to be creative and you need to beta test, and you need to have humans taste it and that’s a thing the AI will never be able to do — taste a cheeseburger.â€
“I will just add that the world’s best cheeseburger does exist,†Hernandez chimes in, “and it’s an In-N-Out burger.â€
“Mrs. Davis†is a lot of things — we haven’t even talked about the religious dimensions of the show, which invite questions of what other structures “force†our decision making, and have long before consumer technology. But it’s the show’s awareness that AI will never taste In-N-Out that resonates after the credits have rolled. That we can oppose the mass proliferation of that burger-oblivious AI, that it’s entirely within our capacity to do so, if we so choose.
Even when it feels like an uphill battle. After the lights came on at the end of the premiere, people began to move toward the aisles. The proxy guys in the nice suits were nowhere to be seen, and it was still only a matter of seconds before everyone’s rectangles lit up across the theater.
