Does Beauty Care How It was Made?

Last Sunday, I stumbled across an essay about “the quiet grief of adult friendship.” It was a quick read that I enjoyed, especially given how my wife and I were discussing the same topic just the night before. I shared the link with her, retweeted a screenshot of my favorite part of the essay, and then I moved on.

When I checked my phone a few hours later, my X notifications were blowing up like crazy. This random, nearly mindless retweet was going viral.

Apparently, this essay — and in particular the excerpt I reposted — was striking a real chord with people. Adult friendship is one of those hard to describe and deeply emotional feelings that’s rarely captured well. Yet this essay did it justice with a call towards choosing friendship. It clearly resonated with so many people on the Internet who have experienced the quiet grief of gradually losing your friends as you progress through adulthood.

As the day progressed and the tweet spread, I saw another topic brewing. People were calling out that the essay might have been written by AI. This was deeply upsetting to many.

So much so that some folks questioned how I could call this essay “beautiful” when it might have been written by AI. Some accused me of not reading enough books or writing enough to recognize what good writing looked like. Others went after me for being an AI tech bro pushing an AI-first propaganda.

I try not to take the Internet too seriously, especially when the pitchforks come out. But this entire episode made me think deeply about what was upsetting people.

In my mind, the essay had done its work by becoming an artifact that millions of people experienced. It was the ultimate creative act because it did something real, it touched people’s lives. 9K reposts, many of which included people tagging their own friends or saying they texted them to say they were choosing that friendship.

If the author of the essay had used AI to write it or help them write it, would knowing that it was AI-enhanced undo the emotional impact? Why was it preventing certain people from altogether reading the essay?

A Tale as Old as Time

This pattern in rejecting changes to our creative process is not new.

When the printing press arrived, scribes worried that mass-produced books would devalue the sacred act of copying manuscripts by hand. When photography emerged in the 1830s, painters declared it the death of art, because it was allegedly a mechanical process incapable of capturing a soul. When recorded music spread in the last 75 years, critics argued that something essential was lost when you separated a performance from the living moment it was created in. Each time, the argument took the same shape: a new tool that enhances human creation is not real creation, it is a shortcut that desacralizes the whole thing.

Yet each time, the tool became part of a new process and a new norm, and beauty kept taking on a new, evolved shape.

Of course, AI does something to the creative process that raises the stakes in a way worth explicitly acknowledging.

LLMs can accelerate human creation so dramatically that it can start to feel less like a tool and more like a total delegation of the creative act. That is a real concern, and a different one than the printing press posed.

But even many of the norms we’re reaching to evaluate creative work are younger and in need of evolution, than we think. Take for example, the issue of plagiarism. Danny Crichton points out, plagiarism is considered the cardinal sin of creative life — but the concept is only about 200 years old. It emerged out of a specific moment in society’s transition into print culture and copyright law. Before that, ideas moved freely, were remixed and retransmitted without attribution, and no one was deeply upset about it. The defense against plagiarism feels ancient and sacred, but it was only invented around the same time as the steam engine.

So that begs the question of what we find so upsetting about the prospect of humans using AI to help them write? It isn’t exactly plagiarism, but it isn’t exactly original. It isn’t exactly pure, but it isn’t exactly inhuman. Why do today’s norms dictate that it matters whether was something was written by AI? What exactly are we protecting when we reject AI-assisted writing?

Soul Over Source

In my opinion, there is a coherent, honest criticism of AI-assisted writing that we should fight. Ethan Mollick named it explicitly in his latest blog post as AI surrender

AI need not undermine your ability to think, but it can do so if used badly and badly is often the default. My colleagues at Wharton call this “cognitive surrender,” and they documented how people would stop thinking about problems and just let the AI do the work, even when the AI was wrong. I think part of the problem is the way these tools are designed.

When you hand over more than the packaging of your ideas and also the thinking, you get what he calls “meaning-shaped attention vampires”. I see a lot of this all throughout my LinkedIn feed, and more so increasingly on X…all contributing to the dead Internet theory. That’s text that sounds like it means something because it is carefully constructed and optimized for engagement, but carries no emotional weight or human meaning because the intent is soulless.

We should reject content that feels soulless.

When Rick Rubin talks about the act of creation in a general sense beyond writing or music, he describes creators as being a vessel. It’s not that we create anything new out of thin air, sitting tranquilly and in a pure flow state, contriving greatness with words, images, videos, pixels, or brush strokes. Any person who creates knows it’s quite often the opposite. The act of creation is about letting something pass through you. A song, an essay, a painting — it already exists somewhere. And our only job as creators is to get out of the way and let it come through. Some tools help us do this better than others. But they still remain just tools, as part of a larger creative process.

The process can vary, the tools can evolve, the artifacts might change. But the reason for your creation, the soul you breathe into your work, your “why” — must always come from you.

Bad AI writing exposes the fact that many people have very little to say — no real feelings, truth, or genuine observations about the world. When the tools lower the barrier to creation, then the obvious temptation is to create. But before you can create, you must first learn to listen and see and to become a vessel.

Increasingly, it feels that most are writing to play their part in the attention economy game. To release their little vampires into the digital world to do their bidding. These writers have mutated a purer version of the creative process by ignoring the cardinal rule around finding their soul.

I am not dogmatically against AI-assisted writing, but I am against soulless acts of creation. And I do believe we need new norms and behaviors to discourage meaningless attention vampires from clamoring for our attention.

Let It Land

I couldn’t help but obsessively look at people’s various tweets about the viral essay. One of my favorite takes was from Joy Bhattacharjya’s take in the aftermath:

If AI allows one to package authentic thoughts better , I don’t see the problem.
I wear glasses to see clearly. If AI helps someone put together their thoughts better, I shouldn’t have any intrinsic issues.
I was surprised at first, and perhaps a little repelled that AI could manipulate me emotionally. But how is that different from specific chord progressions that an enormous number of chart topping songs use, or the common tropes in a hit teen romance drama?

There were those who read the essay and liked it, then learned it was AI and felt repulsed, and then reframed their reaction to accept it. That’s what updating norms in real-time looks like.

And yet the people most upset about the essay seemed to have nothing to say about the actual content. It was like a mental block that prevented them from even engaging with it because the question of its origins was too loud.

Is it not strange?

These people let their theory of beauty prevent them from potentially experiencing something beautiful. Or perhaps, the larger issue around the rise of AI threatening our humanity invoked a mental resistance to enjoying the piece.

While I understand the psychological fears, my goal is to brace myself for what’s ahead and to adapt. There will only be more AI-generated writing in our future. Most of that work may be exactly what we don’t want to see: hollow, generic, meaningless pieces that don’t offer much.

But the ones that shine will do the same thing that good writing did before we had generative AI. Did it strike a chord with you? Did you feel something? Could you sense there was someone with something real to say, and did that person succeed?

Experiencing beauty has always required a willingness. A kind of surrender from the viewer to receive it. Beauty cannot be engineered and curated in a predictable manner. So when it exists, you must be willing to feel it first and interrogate it later.

I find that so often, we resist the encompassing nature of beauty all around us because we are so focused on what beauty should feel like and how it should arrive.

But I think refusing to receive beauty because of how it arrived, whether through AI-assisted writing or something else, is a cost not worth paying. Beauty in the age of AI will keep coming in unexpected forms, through unexpected hands.

The only question we should keep asking is whether it lands. And for me, this one did.