ChatGPT might very well take my job one day. Which is why it was probably a bad idea to give it a job interviewing the creators and star of Peacock’s offbeat limited series “Mrs. Davis.” But the gimmick was too good to pass up: “Mrs. Davis,” created by Tara Hernandez and Damon Lindelof, stars Betty Gilpin as Sister Simone, a nun who takes on an algorithm that has infiltrated most of the world with the promise of solving humanity’s problems.
So, I figured, why not have an AI grill Hernandez, Lindelof and Gilpin about their show. The three of them immediately got it when we talked over Zoom in late April. We’re all victims of the algorithm and fascinated by the hold it now has on our lives and our jobs.
“On days where my screen time is a billion hours and I’ve filled my brain with my specific robot puppy algorithm, doom scrolling and wish fulfillment montage on my screen, it’s been harder and harder to access the stillness that I need to be creative,” Gilpin sighs. “When I open my phone, I don’t use it to connect and get smarter. I use it to disconnect and get dumber, usually. I leave my phone feeling worse than when I picked it up. What makes me afraid is that as these AI capabilities are suddenly tripling and quadrupling; our emotional maturity and respect for each other and the planet is not multiplying at the same rate. I don’t trust us to use these tools for good… I think that we are sort of scrolling and clicking ourselves out of some vital parts of the human experience.”
Gilpin shares some of the same concerns as Sister Simone. In the show, while most of the world embraces the AI named Mrs. Davis — and chases after its clout (in this case, “wings”) — Simone doesn’t buy into it. She’s the product of jaded magicians, an upbringing that led her to religion. Desperate to win her over, Mrs. Davis offers Simone anything she wants, but first the nun must locate and destroy the “holy grail.” Simone accepts, but with one condition: If successful, Mrs. Davis — whom Simone blames for the death of her father — must turn off for good. And so begins the show’s wild journey, as Simone travels the globe with the help of her childhood best friend (and ex-lover) Wiley (Jake McDorman) and her husband Jesus Christ (Andy McQueen)… Told you this show was unorthodox.
“Mrs. Davis” premiered this spring to a world utterly confused by the rapid growth of artificial intelligence. The threat of AI has become a priority for the Writers Guild of America as its members strike over a new contract — and members of SAG-AFTRA worry how it might impact their careers. AI-generated music produced to emulate artists like the Weeknd and Oasis is starting to spread. And programs like ChatGPT are threatening to further disrupt already decimated industries like journalism.
So why not put it to the test? To begin, I asked ChatGPT to come up with 10 potential taglines for “Mrs. Davis” — and sorry to spook any marketing executives, but many weren’t bad. Among the better ones were No. 6: “A nun, an AI device and a whole lot of holy mischief”; No. 8: “She may be artificial, but her power is real. Meet Mrs. Davis”; and No. 9: “Sometimes, you need a little help from a higher algorithm.”
“Number nine is a little frightening,” Lindelof says. “It’s using the concept of a higher power. It’s mixing and matching the idea of the faith versus tech concept in a very frightening way.” Adds Hernandez: “She’s good. See, I’m already anthropomorphizing. This is terrifying.”
But when it came to asking ChatGPT to come up with questions for the series creators and star, it was a bit of a relief to see AI fall short. It had some good conversation starters, but also asked banal questions that often got the characters and plotlines wrong. Here’s what ChatGPT had to ask:
ChatGPT: What was the inspiration behind the premise of “Mrs. Davis,” and how did you come up with the idea of pitting faith against an all-knowing, all-powerful AI?
Hernandez: Accurate — that is the number one question we get asked. Damon and I were first linked up early on in the pandemic. I wrote a script that made it to his desk, and he responded to it. This script featured nuns as the main protagonists, but was very different than the “Mrs. Davis” we know. We talked a lot about how scared we were of the unknown, the big questions of how we were going to navigate this unprecedented time and how reliant we were becoming on technology to help us. I had pitched the idea of, ‘I just wish there was an app to guide us through this, to tell us what to do.’ So, we had the nuns script percolating, and that would be the faith component. And then this new idea of, if there was this technological higher power, that’s going to be our opposition to traditional faith, which is this omnipotent, powerful God-like figure. Somehow, we landed on an app and a nun.
Gilpin: I often play either wry, sardonic, know-all-the-answers, arms crossed women or vulnerable, open, hands-out women. I think that Simone, because of her upbringing and then [having] an experience that brought her into her faith, vacillates between both of those people. I had never really read somebody like that. So, I was pretty damn excited to get to play her.
ChatGPT: How did you approach the research and writing process for the show, particularly in terms of incorporating real-world examples of AI technology and exploring the ethical implications of this technology?
Lindelof: We were listening to a lot of podcasts, like “Rabbit Hole,” about the YouTube algorithm and how it basically could change someone’s belief system, particularly in terms of adapting conspiratorial thinking. This idea of, “OK, an algorithm is now getting people to believe in stuff.” The other primary text in our research was the book “You Look Like a Thing and I Love You” by Janelle Shane. Her thesis statement was machine learning, these things are actually pretty dumb. And we watched a lot of “Black Mirror.”
ChatGPT: The character of Sister Pauline is depicted as having a unique perspective on technology and its relationship to faith. How did you approach developing this character? [Editor’s note: There is no character named “Sister Pauline” in “Mrs. Davis.”]
Hernandez: My guess is there’s a surprising number of people with that name registered in the U.S.
ChatGPT: Mrs. Davis is depicted as an all-knowing, all-powerful AI. How did you approach creating a character that is both relatable and yet almost God-like in its abilities?
Hernandez: I refuse to engage in this one, because I’m starting to sniff that ChatGPT is really steering us in a pro-powerful AI direction. I feel like this is a trap.
Lindelof: The trap is for Tara and I try to imagine how we approach creating a character that is a God. You know how that works out for the Nazis at the end of “Raiders of the Lost Ark!” Suddenly, your faces are melting off. That’s a hard pass. Nice try, ChatGPT!
ChatGPT: One of the themes of the show is the potential dangers of AI technology. How do you balance this darker portrayal of AI with the more lighthearted and comedic elements of the show?
Hernandez: OK, you’re showing some nuance. I like that. We felt like we weren’t aligned with depictions of AI that we had seen in the past, like HAL, “The Matrix” and Skynet. They always want to destroy, overthrow humanity. What if instead, it wanted to help us and be benevolent. It becomes a world that is getting this beneficial piece of technology, and it’s making their lives better — or at least perceived improvement in the quality of their lives. But to balance the dark, more thoughtful meditations on technology, we can’t just mask over everything with a joke. This show had to find a tonal balance and say something, but do so in a way that doesn’t feel heavy-handed. The show’s a big swing.
Gilpin: I often get told in comedies, “Why are you crying?” And then in dramas, they’re like, “Let’s cut the pratfall, let’s remember you just murdered six people.” So this felt like coming home. I got to have murder in my eyes while doing a pratfall. I think that life changes genre every five minutes. People are farting at funerals and sobbing at prom. That’s more true to life than having to stick to one genre.
ChatGPT: The show explores the idea that technology and spirituality are not necessarily at odds with each other. How did you approach exploring this theme, and what do you hope viewers take away from it?
Lindelof: It’s a slightly frightening question from an algorithm, because it’s taking the faith vs. technology premise and litigating the “versus.” And that’s something that we did constantly, where in any “versus” story, the real creative juice is in this idea that neither one of these things is ultimately going to win or lose, they have to find a way to coexist. The premise of the show is faith and technology are sort of interchangeable, and where the overlap is in the Venn diagram. But ultimately, the one thing that faith will always have going for it that tech doesn’t is that it embraces this sense of the unknowable, unseeable and inexplicable. What we hope viewers take away from the show is, not enough television talks about faith. The television or movies that do very often proselytize. We just want people to be thinking about these things. If they are talking about their spirituality through the prism of this goofy “nun versus AI” show, then mission accomplished.
ChatGPT: Do you believe that shows like “Mrs. Davis” have a responsibility to explore these issues and raise awareness about potential ethical concerns?
Lindelof: This is the second time we’ve seen the word “ethical” in one of ChatGPT’s questions. Methinks it doth protest too much. It knows that it’s being accused of being unethical. Therefore, by continuing to say, “I really care about ethics,” it’s expecting to divert our attention. Does ChatGPT really care about whether or not it’s ethical? Of course not, because it doesn’t care about anything. But we begin to anthropomorphize it. That said, what is the role of AI technology in society in the coming years? The cat is out of the bag.
What ChatGPT also failed to do was ask any pertinent questions about Gilpin (or any of the show’s other actors, for that matter). When I asked Gilpin about preparing for “Mrs. Davis,” she said two things were important: “I wanted to honor nuns and I wanted to honor magicians.” She pauses. “I have never said that out loud.”
Gilpin had a long chat with Teller (of Penn & Teller fame), who (a) can speak in real life and (b) holds no illusions about his craft. Even the wrestlers that Gilpin researched for her role on Netflix’s “GLOW” credited some hocus pocus for their craft. But not Teller. “Magicians are all about, ‘no, my skill is in tricking you. There’s not some outside magic orb that we’re also drawing from. It’s all about sleight of hand.’ I felt immediate reverence for him and, talking with someone who is so generous and so passionate, a parallel with the nuns that I spoke to. He’d probably get along with those nuns.”
One more reason we can breathe easy about A.I. — at least for now. As the series ends, we discover that the entire idea of Mrs. Davis is based on its misunderstanding of material it gleans from the Internet — and in particular, from the marketing materials coming out of Buffalo Wild Wings. That’s not unlike what happened when I asked ChatGPT to write a review of “Mrs. Davis.” Rather than understand what the show is all about, it plucked random incorrect facts from the web — and wrote an entire essay on how the show was set in a retirement home. It is not.
“My mom would love this show, about an older cast in a retirement home run by A.I.-proxied nurses,” Hernandez muses. “I really want this.”
ChatGPT’s entire review reads like someone who hasn’t watched the show. Which reminds Gilpin of something else: “My ninth grade essays when I wouldn’t do the homework, but would just try to bullshit run-on-sentence-paragraph my way into the required word count. ‘You know, this, this question about the War of 1812 reminds me of a personal story,’ as an example. So maybe I was ChatGPT as well!”
(This story combined one interview with Hernandez and Lindelof and another with Gilpin, and also condensed several answers for space. And it clearly was actually constructed by a human, since there’s no way ChatGPT would put together something so clever.)
