There is something deeply funny about watching people argue with AI like it is a bad boyfriend, a lazy coworker, a psychic assistant, a therapist, a search engine, and a mildly haunted intern all at the same time.
I say this with love because I have absolutely done it.
I have snapped at AI when it gave me something bland. I have praised it like a golden retriever when it finally understood the assignment. I have overexplained myself. I have underexplained myself and then gotten annoyed when the response was wrong. I have said “no, not like that” with the energy of a person sending back an undercooked steak.
The more I use AI, the more convinced I am that our chats with these tools are not just about technology. They are also about communication style.
This is part of why I do not hide my use of AI. I am not interested in pretending every piece of writing arrives fully formed out of pure, untouched inspiration. I use AI to organize, sharpen, research, and pressure-test my thoughts. It does not replace my taste, my lived experience, my anger, my humor, my boundaries, or my sentence rhythm.
That last part matters. As AI becomes more embedded in creative work, emotional processing, research, and everyday life, our relationship with it starts to reveal something about us. Not because AI is human. Not because it “knows” us in the way a person does. Because communication is never only about the other party. Communication is also about how we make ourselves understood, how we react when we feel misunderstood, how we handle feedback, how specific we are willing to be, and how much responsibility we take for the exchange.
That is where things get interesting.
Some people approach AI with direct communication. They give clear instructions, state the goal, name the tone, describe the audience, and correct the output without making it a whole emotional event. These people tend to get better results faster because they understand that clarity can be firm without being harsh.
Some people approach AI with a passive communication style. They ask vaguely, accept whatever comes back, and then feel disappointed because the result does not sound like them. They may not want to be “too demanding,” even with a machine. The outcome is usually content that comes across as flat, overly refined, or strangely detached from any real point of view.
Some people go fully aggressive with it. They bark commands, insult the output, demand perfection, and treat every bad draft like a betrayal. This can produce better revisions in the short term, but it also suggests a need to control the process at all costs. If every interaction becomes a fight for control, even with software, it may be worth asking where else that pattern is showing up.
The best AI interactions tend to be assertive, not just direct. Assertive prompting knows the goal, gives context, corrects the output, and stays engaged without becoming dramatic about the fact that the first draft was not perfect. It is the difference between “write this for me” and “help me get closer to what I actually mean.”
Then there is a slightly darker version of passive-aggressive communication that I keep seeing, which is people setting AI up to fail and then acting like they caught it in a lie. They will give it incomplete or misleading prompts, strip out context, or ask leading questions designed to produce a wrong answer, and then screenshot the result like they just exposed a scandal. It is less about curiosity and more about proving a point. And honestly, it feels familiar. People have been doing this to each other forever. We ask questions we already have an answer to. We leave out key details. We test instead of communicate. The difference is that now we are doing it to a machine and still expecting it to read between lines we never wrote.
This is where AI becomes strangely useful as a communication mirror, not just in the output you get, but in the patterns you bring into the interaction. If you cannot tell an AI what you want, it is worth asking where else you might be expecting people to guess; if you accept writing that does not sound like you because you do not want to revise it, where else you might be outsourcing your voice; and if you keep demanding better results without learning how to ask better questions, where else you might be mistaking frustration for standards.
This is why I keep coming back to the whole “AI slop” conversation. Yes, there is a lot of bad AI-generated content out there. You have probably seen it. It often feels like generic, over-processed writing that prioritizes polish over substance. Everything is technically fine, but nothing actually lands. It is all “navigate the complexities” and “unlock your potential” with zero personality behind it.
I remember the first time I tried to use AI to draft something I planned to post. It gave me a clean, polished paragraph that, on paper, was completely acceptable…I hated it. It was so smooth it felt fake. I ended up rewriting almost the entire thing, but that moment stuck with me. The problem was not that the tool failed. The problem was that if I had posted it as-is, I would have been the one disappearing from my own writing.
That is why I do not think the issue is simply that AI exists. The issue is that people are publishing things without taste. They are using the tool without a point of view. They let the machine sand down the edges, remove the weirdness, flatten the emotion, and strip out the little contradictions that make something feel human, then they hit publish because it sounds “good enough.” But “good enough” is not the same as sounding like yourself. That is not really an AI problem. It is a communication problem, because good communication requires participation, revision, knowing what you meant, noticing when you were misunderstood, and staying in the conversation long enough to repair the meaning. That is true with partners, friends, and readers, and apparently it is true with the robot in the browser too.
When I work with AI, I am not asking it to become me. I am asking it to help me hear myself more clearly. Sometimes it gives me a structure I could not find on my own. Sometimes it spits out a line that makes me physically cringe, and weirdly, that’s what snaps me back into what I actually want to say. Sometimes it hands me something so generically polished that I immediately know what I do not want to sound like.
That has value. A bad AI draft can be useful if you have enough taste to reject it. A decent AI draft can be useful if you have enough voice to reshape it. A good AI exchange can be useful if you understand that prompting is not magic. It is communication. The people who are going to thrive with AI are not the people who ask it to replace their thinking. They are the people who learn how to collaborate without disappearing. They know how to say: this is close, but softer. This is too polished. This sounds like a brochure. This lost the anger. This needs more lived experience. This is structurally useful, but emotionally wrong. This sentence is cute, but it is not mine.
I think our communication styles are going to matter more as AI becomes more normal. The tool may be artificial, but the way we relate to it is still deeply human. We bring our habits into the chat box. We bring our fear of being misunderstood. We bring our perfectionism. We bring our people-pleasing. We bring our control issues. We bring our longing to be helped without having to explain every little thing.
So yes, I use AI openly, curiously, and critically. Sometimes joyfully, sometimes with suspicion, and sometimes while muttering “absolutely not, George” at my screen.
AI did not make me a writer. It simply made it harder for me to abandon the writing I was already trying to do.
I had a voice before a model could imitate one badly. I had opinions before I had prompts. I was a writer before I used AI.
That is the part I keep coming back to.
The future of AI and creativity is not just about what machines can generate. It is about how we use the tool. AI is closer to a paintbrush than a painter. You cannot just press a brush to canvas and expect a masterpiece. You have to know what you are trying to create, how to guide the strokes, when to adjust, and when to start over.
AI works through communication. And right now, a lot of people are using it without knowing how to communicate clearly, specifically, or with intention. That is why so much of what we see feels flat. Not because the tool is inherently bad, but because the skill of using it is still underdeveloped.
Until we get better at that, AI slop will keep showing up everywhere.
But that is not the future I am interested in.
I am interested in what happens when people learn to use powerful tools without abandoning their own voice, and how developing deeper connections and communication skills shapes that process. The better we understand ourselves and others, the more intentional and precise our interactions become, whether with people or with technology. That same clarity leads to more thoughtful and personalized outputs.
Because the machine can generate language, but it cannot feel the weight of what you are trying to say, or the quiet urgency behind it. It cannot sit with your contradictions, your memories, your instincts, and choose what matters enough to be spoken.
Only you can decide what you mean.
Originally posted on Medium.com
Leave a Reply