What my wife taught me about care, writing and AI slop.
"It’s not the form that matters, it’s the content!"
I used to say this to my wife (then girlfriend) when we had an argument and I wanted to make my point.
Early in our relationship, she’d tell me the way I said things hurt her. My response was always the same: focus on what I’m saying, not how I’m saying it. I thought I was being rational, analytical, and “efficient,” since—I believed—form was decoration and content was truth.
Over the years, I realised I was wrong.
Growing as a husband and as a dad, I see it differently now, and I appreciate that form does matter! A lot.
Form is the evidence of something deeper: it’s evidence of care. Form is the texture through which meaning travels, and that texture does carry meaning too.
Writing is thinking
A colleague posted on Slack arguing that long-form writing helps remote teams think together and better collaborate. I initially resisted because it felt inefficient for someone like me, used to operating in startups where all communication is about speed, effectiveness, and cutting the b*s. Then I asked myself: why not? And started seeing the value of his main point, that writing is thinking.
Writing is the act of thinking itself, made visible to others (and to us, too).
This relationship between thinking and writing matters more than I realized because if writing is thinking, then the form of your writing reveals the quality of your thought. Rushed writing reveals rushed thinking and careless writing reveals careless thought.
And probably that’s where AI breaks down completely.
I’ve spent nearly two years experimenting with AI writing tools:
building custom GPTs, feeding them my knowledge base, tuning them to match my voice. I’ve tried every workflow:
Bullet points to AI, full article out;
Dictate thoughts, let AI polish them;
Pen and paper brainstorming, AI for cleanup.
Some pieces worked well while many flopped miserably.
After some time, I noticed the pattern: articles where I wrote 70% or more and AI contributed 30% or less got real responses, comments, newsletter replies, true engagement. I’ve met new, wonderful people that way.
Articles where AI generated 70% or more of the content led to silence, zero resonance, no meaningful interaction.
People can tell the difference. We coined the words “AI slop,” in fact, because AI writing feels fundamentally different. Pure AI output is tasteless, careless and thoughtless.
Tasteless because there was no taste behind it.
Careless because there was no care behind it.
Thoughtless because there was no thought behind it.
One word captures it: soulless. And readers sense it instantly.
"AI" is the new "stupid."
When my kids hear or read something they don’t like, something that feels cringe, nonsensical, or just off, they developed a simple dismissal: “Oh, that’s AI.” Which actually means, “that’s stupid.” They use it for everything now: bad writing, stupid arguments, even human-written content that lacks thought. “That’s AI” has become their shorthand for absence of care.
Kids see this clearly because they have fresh eyes for what most adults have buried: the difference between words arranged cleverly and words that carry genuine thought.
Philosopher John Haugeland wrote in the 1980s (during the first AI hype wave) that AI systems fundamentally lack something. He called it “giving a damn.” The capacity to care about what you’re doing, to take responsibility for it. To have it matter. And I think Haugeland would stand by his words today, even looking at systems far more advanced than anything he imagined.
I’ve been building machine learning systems since 2012. I’ve built ventures and great products on this technology. Traditional machine learning techniques and deep neural networks are incredible, the engineering is remarkable, but calling it “intelligence” misses what’s actually happening and misdirects the general public — themes I go into at length in The Ethics of AI.
Synthetic Agency
I prefer how contemporary philosopher Luciano Floridi calls it: artificial agency.
These systems have agency, they’re built to do things, accomplish goals, produce outputs. But agency isn’t intelligence. Agency is engineered capability. We’ve built artifacts that perform increasingly complicated tasks that computers historically struggled with. We’ve found clever techniques and engineering tricks to get acceptable (or even exceptional) results in a variety of contexts.
And that’s indeed artificial agency. And I’d dare to suggest a better term: "synthetic agency."
I like looking up definitions, and I think there is no better word:
Synthetic
/sɪnˈθɛtɪk/
adjective
(of a substance): made by chemical synthesis, especially to imitate a natural product. “Synthetic rubber.”
(of an emotion or action) not genuine; insincere. “Their tears are a bit synthetic.”
Similar: artificial, fake, false, faux, imitation, mock, simulated, pseudo, sham, bogus, spurious, counterfeit, forged, pretended, so-called, man-made, manufactured, unnatural, fabricated, replica
Opposite: real, genuine, natural
Logic (of a proposition): having truth or falsity determinable by recourse to experience.
Linguistics (of a language): characterized by the use of inflections rather than word order to express grammatical structure.
A machine that can write coherent sentences, draw an image, plan some work and execute it is an engineering artifact synthesized from something with a more advanced agency—us, our knowledge, our intention. This matters because it establishes the natural hierarchy: creator and tool, user and machine, person and artifact.
Form actualizes potential
There’s an old philosophical principle that clarifies why this matters: Aristotle’s hylomorphism. In his view, everything consists of matter (potential—what something is made of) and form (the actualizing principle that gives it structure and definition). Matter alone is indeterminate. Form actualizes that potential, making something what it is.
This applies directly to communication. Raw ideas are matter, potential meaning waiting to be actualized. Form is what makes potential meaning actual and real. When you write carelessly, you’re leaving meaning in its potential state. When you write with care, choose words deliberately, structure thoughtfully, respect your reader’s time, you’re actualizing that potential.
Care is the form that actualizes communication.
AI writing lacks this form because it has no care behind it. No stakes. No possibility of being wrong in a way that matters. Allow me one more philosophical digression here.
Borrowing some concepts from Floridi’s philosophy of information, information exists at multiple levels: statistics, syntax, semantics, pragmatics, and apobetics. In other words, raw data (syntax) isn’t information until form is applied, until it becomes meaningful (semantic) and purposeful (pragmatic) to an actual receiver, hence achieving its higher purpose (apobetics).
Information channels require relationships and flows between systems. You can’t have meaning without a sender who cares and a receiver who’s being thought about. (Read here for more on that.)
Meaning, therefore, requires relational presence. Information only becomes meaningful through the relationship between sender and receiver. Someone had to think about someone else.
When AI generates text, there’s no one thinking about you, the reader.
There’s pattern matching trained on billions of examples, but no actual consideration of whether this specific message reaches this specific person effectively.
Form matters because it’s proof that someone actually considered the person on the other end.
Wrapping up my learnings
When I told my wife, “content over form,” I thought I was being efficient. Eliminating unnecessary friction. Getting to what really matters. So why did my words hurt her?
Because bad form causes friction. On the contrary, well-designed and carefully executed form is how meaning arrives without damage. Form is the care you take to make your message land softly, clearly, completely. Form is evidence that you thought about the person receiving your words.
There are words and there are colors. There’s texture and depth. It’s not just thoughts transmitted, transacted. It’s the care of thought traveling from writer to reader.
That’s what pure AI writing lacks. I emphasize “pure” because I do use AI to polish form, as an expression of care. AI helps me express my thoughts better. Pure AI output can achieve grammatical correctness, vocabulary range, or even logical structure, but it deeply lacks the evidence of someone giving a damn.
When I hear claims about “superintelligence” — machinery that can replace, achieve, and even surpass the care, thoughts, and tastes of its human creators — it sounds silly. Care, thought, and taste, perhaps, can be simulated or engineered, but they ultimately land as shallow or blatant stupidity. Like the kids would say, “That’s AI.”
True care, thought and taste emerge from something having stakes. From someone who can be wrong, who can fail, who has something to lose by not caring.
So, form matters because form carries care.
My wife wasn’t being sensitive when my words hurt her. I was being careless with how I delivered them.
Pure AI writing falls flat because there’s no one behind it who gives a damn. Writing is thinking made visible, and form is how that thinking reaches another person intact.
I still reach for AI tools. They’re helpful for specific tasks, like polishing this article (it shortens my sentences too much, though! You’ll tell me if that’s good or bad) — see my AI disclosure for exactly how much AI assistance went into this one. But I’ve learned the ratio: when my care, my thought, my taste dominate the output, people respond. When the synthetic agency dominates, people sense the absence.
I hope you felt the care in this article. Let me know—honestly!
FAQs
What does it mean when form carries care?
It means the way something is written, structured, and delivered is evidence of whether real attention went into it. In writing, good prose is how care reaches the reader without being damaged or diluted.
Why does AI writing often feel like slop?
AI writing often feels flat because it can produce fluent text without genuine stakes, judgement, or responsibility behind it. The words may be grammatical, but the reader can still sense when no one truly cared about the message.
Can AI still be useful for writing?
Yes. AI can help with polishing, structure, grammar correction and suggest copyediting, especially when a human has already done the thinking. Rather than focusing on using AI at all or not, the problem is letting it dominate the substance so the result loses taste and care.
How can I tell if writing was created with care?
Look for specificity, clear judgement, and a sense that the writer understood the reader. Care shows up in the choices around tone, pacing, examples, and restraint. If the piece feels generic or interchangeable, or when you read a sentence and can't recall what is the main message/argument, care is probably missing.
What is the main argument of this article?
The core argument is that form is not separate from meaning. Form is the vehicle through which care, thought, and responsibility arrive. When form is sloppy, the message may still exist, but it may fail to land or lands with less force and less trust.