chatGPT killed the radio star
art in the time of slop
At dinner, I confess to Ellie and Dan my embarrassing internet habit: I talk to ChatGPT like I talk to a real person. I often start my chats with “Hi can you” and end then with “please”. When it delivers me exactly what I’m looking for, I say “great!” and “amazing” and occasionally “thank you soooo much”. I recognize it’s frivolous, but I can’t help myself. In Dan’s words, I’m committing the modern sin of a boomer double space.
This becomes my new bit at parties: in what manner do you talk to ChatGPT? Over dinner, Colton tells me he’s ruthlessly terse as to not confuse the model tuning. It’s certainly logical, but for the most part people are a baseline level of polite whether out of habit or as an insurance policy should the AI ever become sentient.
It’s funny then to reflect on our mannerisms with AI in the context of how we interact with one another on the internet. Sometimes I’ll post a video on TikTok and the comments will say something to the effect of “jeans where from?” or more simply “bag?”, “belt?”, “boots?”. I understand the unwritten agreement of posting content online, so this isn’t a complaint as much as it’s an observation: online exchanges can often become transactional to the point where there’s little distinction between humans and sentient search engines.
✧✧✦
When I first started posting videos online, I found it fascinating that strangers wanted to watch what I had to make. While the idea of following someone you don’t know isn’t new, it was the first time where it felt like we were willing to not just listen to a handful of influencers, but to invest into the ideas, preferences and habits of strangers on a mass scale.
A few years ago, Megan Garber published a provocative essay in The Atlantic on how we’re already living in the metaverse. The crux of Garber’s argument is that the internet has already blurred the lines between what is real and what is fiction — high profile scandals quickly become voyeuristic, consumable television (think Elizabeth Holmes, Inventing Anna, The Crown) and the proliferation of content is turning unwitting strangers into background characters and occasionally stars of our homemade films (think TikTok street interviews, filming in workout studios, Amazon delivery guy pranks).
I’m still not convinced, however, that we were ever meant to interact with this many people on a daily basis. When all you see on your FYP is a never-ending loop of strangers, it’s hard to disentangle what we view as foreign because it’s someone we don’t know and what we view as a fictional character on a screen. It would be bizarre to leave behind anything but positivity on an acquaintance’s instagram, but hate comments feel commonplace and oftentimes, justified on TikTok. Garber describes what we already know — behind the anonymity of a screen, people act in ways they wouldn’t think to in real life.
She leaves us with a warning: every dystopian novel unravels with a similar storyline.
“We will become so distracted and dazed by our actions that we’ll lose our sense of what is real. We will make our escapes so comprehensive that we cannot free ourselves from them. The result will be a populace that forgets how to think, how to empathize with one another, even how to govern and be governed.”
✦✦✧
There’s been a lot of discourse recently about the merits of AI-generated art: at what point is this art and at what point is this slop? It seems that the internet has taken the side of the humans — there is quality and value in the painstaking creation, process and imagination that underlies artistic expression. For now, a cheap replica is not enough.
But how should we engage with AI in a world that is in eternal pursuit of efficiency, speed and progress? I want chatGPT to do the mundane activities (emails, to-do lists) so I can engage in the real world: write, create, enjoy the company of others. But I find myself impatient for the outcome — I want the good parts to come immediately. Sometimes I am tempted to ask the model to help me make a decision, to help me finish an essay, and this impatience scares me.
In Berger’s Ways of Seeing (first published in 1972) he writes about the value of art in an era of reproduction — what happens when art becomes “ubiquitous, insubstantial, available, valueless and free”? We no longer have to watch any of Hayao Miyazaki’s work in order to participate in his universe. We can now see the Studio Ghibli version of ourselves from the comfort of our bedrooms via our smartphones, our Twitter feeds and our text messages.
✦✧✦
Last fall, I got to spend some time in one of my favorite art galleries: White Rabbit, one of the largest collections of Chinese contemporary art. The piece that caught my eye was a little video montage on the way out of the exhibit superimposed on top of a catchy jingle titled “David”. The 5 minute “music video” by Guan Xiao is a looping collection of home videos taken by tourists in Florence viewing Michelangelo’s David. The critique is on the compulsive desire of David’s spectators to document the sculpture for one singular purpose: to walk away with tangible evidence that they had seen the masterpiece. As we watch the crowds behind a sea of iPhones, we are confronted with the realization that there is “little remaining interest” in the artwork’s beauty or realness.
All of this to say, in a time where we are increasingly exposed to AI-generated slop, there is value in seeing the real thing but only if we’re really looking. Berger argues that original art is important because of the relationship between the artist in history and the viewer in the present. There will always be meaning found in human connection, creativity and originality, but it requires openness, and importantly, patience in how we relate to ideas and to each other. ✦✦
I absolutely loved this Tina! My fiancé always speaks very politely to Alexa, and he says it's because he's worried if we get in the habit of screaming "SHUT UP!" or "TELL ME THE WEATHER" at a computer person, we'll be more inclined to do the same to real people, like you mentioned. And I tend to agree! Online interactions are already coloring (and contaminating, in the case of the David statue) our in-person interactions so much that it isn't much of a leap. Thanks for sharing--really interesting to think about!
I think the discussion about how we interact with AI (or any other computer) is both interesting and fascinating. I do the same. When I chat with ChatGPT, I write as if I’m talking to a real human being. The reasons for this are:
1) I’m afraid that if I start getting sloppy and cold, it will affect how I communicate with other human beings. AI will become just as integrated into our lives as the smartphones attached to our hands, so building a good habit now feels crucial. Not to mention, I’ve seen and felt the effects of how social media sharing and consumerism have changed the way we treat friendships and communicate with each other, and the disinterest we have when meeting new people.
2) Most importantly, I respond to ChatGPT in a friendly manner because it’s so damn friendly back. You just can’t help it! That’s also why I gravitate toward using ChatGPT rather than other models; their more transactional way of presenting information feels rather boring and disengaging. As humans, we thrive on engagement. The creators behind ChatGPT understand this. (The Like-button killed engagement, and we're now desperate for it. Whether engagement comes from a computer or a real human being, the brain doesn't see a difference.)
So, whether or not Altman says that saying “thanks” to ChatGPT costs a few bottles of water and that people should stop saying thank you — or at least don’t need to — they should also consider making the model less friendly if that’s the goal. But I think that would lead us into serious trouble. It could amplify the scary scenarios of humans forgetting how to communicate with each other with empathy, and further drive the transactional/brusque tone we already adopted. Like I said, we’ve already seen how profit-driven social media has shaped our behavior toward one another. When AI becomes a daily form of communication, like an extension of our thoughts, what effects will a daily exposure to a neutral, unemotional tone have on how we treat each other? — Sacrificing relational habits for efficiency may not be worth the trade-off.
And lastly:
3) imo, how one talks to AI says a lot about a person and their EQ.
With that said, not all interactions with AI require friendliness. It's important to distinguish between conversations. Some prompts are purely transactional and quick, while others are deeper and more engaging. The latter is where I find myself holding the conversation as if I were talking to a human assistant.