Dear Allison and Amy Rose,
Multiple times a week, one of my best friends texts me AI-generated memes based on inside jokes, or she’ll tell ChatGPT to, like, turn us into Drag Race contestants. I hate it and mostly ignore it, but it’s not slowing down, and it’s getting awkward: When I see her, she’ll ask what I thought of that ’80s music video she “made” of “us.”
I’m surprised — she’s progressive, so I figured she’d be put off by AI’s environmental consequences. I don’t want to judge her, but I also believe tech companies want us to be lazy thinkers so we rely on their products. All this doom and gloom comes to mind whenever she texts me AI-generated cartoons of our dogs hanging out at the bar. (It doesn’t help that this stuff is never funny.)
My friend is trying to be affectionate, and I hate leaving her on read. She’s really sensitive — she sometimes talks about how my other friends are cooler than she is, so I want to be careful about not hurting her feelings or scolding her. But I also don’t want to quietly resent her, and I miss her actual sense of humor. How do I kindly tell her to stop sending me AI garbage?
— My Phone Is a Slop Bucket
Dear Slop Bucket,
For three and a half decades, birthday cards from my dad have starred “Snoppy.” Snoppy looks about how he sounds. He’s a cartoon dog, but he’s also sort of a mutant letter “P” with a beady eye pierced heavily enough into his face to have dented the table he was drawn on. Snoppy usually stands alone, but not always. “What is … that part?” I asked the first time I saw quotation marks floating beside Snoppy’s head. My dad was surprised I couldn’t tell: Woddstock, obviously.
If my dad had been able to ask ChatGPT to make his Peanuts cards, given that he’s not much of an artist, I’d have said thank you, chucked them, and instantly forgotten about them. Instead, I have a lifelong collection of ballpoint-on–printer paper masterworks — something sweet and tangible to associate with my dad. When I imagine life without Snoppy, it feels like a significant loss.
Surely your friend, like my dad, is just trying to relate to you by making you a personalized token of their affection. Plenty of smart, funny people find AI memes novel, which makes sense. We’ve been primed to identify with one another this way over years of BuzzFeed quizzes, face-swapping apps, “tag yourself!” memes, and other digital expressions of our tastes and friendships. Those had their own problems, but the tradeoffs felt more legible. We were served ads and got a little loose with our data. In return, we got to laugh about which Mad Men character we were with our Facebook friends — and to become comfortable with using technology for kicks without thinking too hard about what we were exchanging for that.
Given the long-standing popularity of identity-driven shareable content, there’s a fortune going into convincing us that generative AI is cute, fun, sweet, thoughtful, and personal. One ChatGPT Images ad I get every day compares a photo of a pet rabbit with a charmless AI-generated cartoon of that same rabbit next to a sign reading, “Hoppy Home.” Another ChatGPT-sponsored post features the prompt “Give my dog the nobility portrait she deserves.” It’s a photograph of the dog, followed by an illustration of the dog in one of those Bridgerton wigs and a corset that, yes, gives the dog huge tits. A different kind of fun with pets! All of it is bait, offered so you’ll learn how to use it, then build a habit of AI.
The question you’re asking, Slop Bucket, is very much present in my life. Plenty of people I know are tickled by the “creative possibilities” of this “amazing new technology.” (And everyone loves their animals, sometimes enough to overlook their awful new breasts.) Like you, I’m worried about the environmental and intellectual consequences of relying on AI — for art, for friendship, for war, for recipes. Also like you, I think it’s a corny way to relate to your friends and that the memes aren’t funny.
Take heart. When I’ve had candid conversations with AI-curious friends, it’s gone well: no hurt feelings, no more spirit-deflating “Look, we’re 80s businesswomen!” notifications. The first thing you should try is replying directly — do this one-on-one, never in a group chat with an audience. Say, “I see the vision, but AI isn’t as funny as you.” Add a “lol” if you need it. Then say, “Wait, that reminds me” and redirect to a recent joke they made, or something you two love to dunk on. Appeal to your sensitive correspondent by emphasizing their singular charm, implying (or saying outright) that ChatGPT could never.
What if they push back or get defensive? It happens! Sometimes, when people get a sense that you don’t like AI, they want you to affirm that it’s not that bad that they’re using it, despite the ethical and environmental issues they’ve absolutely heard about by now. Your friend can tell that you’re uneasy, if not that you’re fully spiraling about cooling centers. That’s why your friend is nervously asking about memes when you hang out. People get tripped out and huffy because they think they actually made whatever generic thing they told Chat to spit back. They identify with it. However: Proof of creativity is the “creation” part. (Often, someone counters with “I had the idea, but I’m not artistic!” Good! Go Snoppy mode.)
This approval-seeking sucks for everyone involved. Recently, a friend had a perfect idea for a custom gift for a mutual. When she gave it to them, the design was clearly AI. No one said anything, but she rushed to say she was worried we’d knock her for making it on ChatGPT. We didn’t — it was Christmas! — but why are we suddenly having to assure someone we love we’re not judging her? Why is she having that feeling, when she just wanted to give someone a gift? It brought unnecessary angst into a sweet thought that could have been pulled off another way.
I don’t recommend coddling or condescending to your friend because you think they’re sensitive. That pretty much always backfires. But it’s easier to make your point in moments when a friend doesn’t think they’re giving you something and being rebuffed. On a walk in my neighborhood, my sister asked ChatGPT where we could have a glass of wine. (It showed her the worst, most dunce-infested bars in the area.) After I suggested a place, my sister asked if I ever had these problems with ChatGPT and I said I didn’t use it. She asked why; I explained, without shrieking at her about why she was bad; and she was like, “Oh! That’s horrible, yeah, wow,” and deleted the app.
If your friend is curious about your position, go ahead and talk about it. It’s hard to come by reliable data about AI, but here’s what I know and what I say: Environmentally, habitual generative-AI use isn’t like buying a plastic water bottle when you forget a reusable one at home or whatever other concessions we occasionally make despite our largely good intentions. MIT Technology Review’s analysis of the available research about the environment and AI includes a helpful illustration: Using AI to generate an event flyer easily costs as much energy as running your microwave nonstop for three and a half hours.
The report isn’t optimistic from there. “All bets are off in the coming years,” says Sasha Luccioni, a leading scientist studying AI and climate. “Generative-AI tools are getting practically shoved down our throats, and it’s getting harder and harder to opt out or to make informed choices when it comes to energy and climate.” Recent research published by Morgan Stanley also found without qualification that AI is sucking up our drinking water, especially in places that are already vulnerable to declining water quality and droughts. Defending your memes by saying you’re environmentally conscious otherwise is like saying it doesn’t matter if you apply sunscreen every day because you use wrinkle cream — even though, of course, what you more urgently need is skin-cancer protection. You can’t recycle your way out of this one, homie!
Beyond the environment, there’s a lot more to loathe about AI. If there are ethical tradeoffs that justify AI’s starring roles in ICE kidnappings, the genocide in Palestine, racist policing, and invasive surveillance tactics, I can’t think of them, and I’m not about to ask ChatGPT for help. So you could also try something like, “You’re killing people by using the computer for yuks, murderer.” It’ll go perfectly, as scolding people always does! Do your best to take your friend in good faith: It’s easy to lose sight of how bad AI is, given that we’re bombarded with the “Cute rabbit! Japan vacation tips!” applications, not the destructive ones.
I sometimes say that I’m not against every single practical application of AI — just that I don’t think it has a place in creative expression, casual communication, and most day-to-day contexts. In certain workplaces and fields of study, with specific guidelines around its uses, though? Fine! I know a scientist who uses AI to understand genetic disorders better and help develop treatments for them. His programs do this by sifting through impossibly large sets of code to pick up on tiny aberrances. Work like that is a rare example of AI justifying its energy use. I think we can agree that saving lives isn’t the same as putting a golden retriever in a lusty, period-appropriate bustier, then texting it to a friend whose patience is thinning by the day.
If your friend is obstinate in the “But I made this!” or “I’m environmentally A-OK!” or “I just thought it was nice!” or “I’m like a scientist — I embrace innovation!” modes outlined above, you know what, that’s fine. They can text someone else about it. Say, “It’s not about you. AI bums me out and I think your jokes are funnier,” and leave it there. If they want to have a debate about why ChatGPT is good, hold firm and end the conversation: “Okay, but I just like you better than ChatGPT,” or “I know you might not agree with me, but it’s how I feel.”
No matter what: Don’t be afraid to lead by example. Make bad art for your friends. For years, my go-to gift has been a portrait of a friend (or their child, Chihuahua, or favorite landmark) on a ceramic plate I found on the sidewalk, washed off, and stashed for this precise purpose. Despite what you may have heard, I’m not Finnish master portrait artist Helene Schjerfbeck, but many of my disgraceful modernist works are displayed proudly in friends’ homes.
After I painted my friend Chrissy’s cat, Chrissy drew mine on a paper plate in response, writing some inside jokes beside it. Now, I see it on my fridge every day, lightly thinking, Man, I love Chrissy! every time I reach for a seltzer. To all the AI-meme artists out there: Make something of your own. Something that takes a little more time, doesn’t work out totally how you intended, and lasts much longer. You have great alternatives to causing your friends existential despair by way of boneheaded fake album covers. If you’re using AI for relationships, what are you even “saving time” for?
Snoppy forever,
ARS
Have a question for Going Through It? Email goingthroughit@nymag.com or share using this Google Form (and read our submission terms here.)