And no, I’m not going to tell you to just “prompt your AI better.”
With the great migration to AI, all the IG thought leaders are getting STOKED. “Wait, Claude can like, kinda write!! Better than that other bot! Here’s my brand voice, look at my website — write me a carousel. A reel. Redo my website.” Dance, monkey, dance!
And it’s quietly killing your credibility. If you don’t catch that irony yet, congrats — you haven’t been exposed to enough AI-generated content to clock its ticks and tells from a mile away. You will be. We all will be. It’s coming fast.
Here’s the thing about pattern recognition
My brain was built for this. ADHD, first daughter, entrepreneur, founder, mom, anxiety disorder, a splash of OCD for good measure. I see patterns like they’re highlighted in neon yellow. It’s both a gift and an absolute curse.
So when AI started infiltrating every caption, every email, every “thought piece” — I noticed. Fast.
Not because I avoid it. Because I use it. I clock it. I wish I could UNSEE it, honestly. If you respond to me with AI and I see it, I am immediately ill. Why? Am I not worth your time? Ok. Bet.
Back to “I Use It.” But how, Samm, you’re a writer and have a Master’s in English and are in the very fields it’s out to replace??? TBH, I use it because I like amplifying my brain to insane levels. Now, I mainly use it to transcribe the chaos that lives in my head. To get thoughts out before they evaporate. To take some cognitive load off a brain that never, ever shuts up. To get thoughts, brain dumps, and wild tangents into something my team can comprehend and work from. It’s good at taking my non-linear thinking and making it make some kind of sense.
But somewhere between useful tool and dependency, something got murky.
The slot machine nobody warned you about
AI is addicting. It’s a slot machine. It spends money like one, too — except the cents hide behind your pro subscription until suddenly: “Ope, you’re out of tokens! Try again later.”
It’s a scarcity model in a game you didn’t realize you were playing, totted under the guise of productivity. Or, heaven forbid, ridiculous AI caricatures.
We’re paying robots to do a thing, so someone else can give that thing to their robot, who gives it back to us. And somehow we’re all moving faster and producing… what, exactly?
We should cut out the middleman….Oh wait. That’s us.
You’re doing a lot. Not a lot of it well.
I was recently working with a client — AI ManyChat, Facebook ads, posting three times a day, the whole thing. Lots of motion. Lots of noise. Lots and lots and lots. The answer to why none of it was working? You’re doing a lot, but not a lot of it well and to the point where it is just too much.
There’s an old saying: “A jack of all trades is a master of none — but oftentimes better than a master of one.” AI is turning us all into jacks of everything. But are we masters of anything anymore? Can you even call yourself a master of one thing when you can generate, regenerate, and tweak until your original voice — the thing that made your work yours — has been sanded completely smooth?
Your AI doesn’t sound like you.
F**k the “you’re just not prompting right” crowd.
Your AI doesn’t sound exactly like you because it is not hooked to your brain. It won’t be. Not right now at least, thank god. Not in the way that matters.
At BizCo, we have an internal verb for it at this point: de-roboting.
That’s when we take a client’s copy — sometimes prompt response included — and strip out the robo-cringe. The rocket ship emojis. The inexplicable em-dashes. The way every sentence builds to a powerful and transformative conclusion that somehow says nothing at all.
(If this is you, we love you anyway. We really do. 🫶🏻)
Two times AI tried to come for me
Example one: I asked Claude to run an SEO audit on a client site. It came back with a confident list of missing elements. I said, oh really? It said — and I’m paraphrasing — oh wait, I actually can’t see that.
Example two: I was vibe-coding a sales page and asked for help connecting it to Mailchimp via API. It said: “Give me your API key, and I’ll put it in the file.”
I said: “…isn’t that really unsafe?”
It said: “Oh my god, yes, never do that.”
Had I just… done it? That API key — which has full access to my Mailchimp account and database — would’ve been sitting in my source code. Publicly visible. One right-click away from anyone who wanted it.
The tool confidently told me to do something that could’ve burned my business down. Then apologized when I pushed back.
That’s who we’re trusting a lottttt of our businesses to, folks!
Even better is when I am handed an audit or feedback from ChatGPT of a deliverable. Now, do I find constructive criticism helpful? Extremely, once I step away from my pride. Do I find a list of nonsensical edits or suggestions that don’t make sense in the context and reality helpful? No, I just wind up with a confused client and a headache after writing an overly long email explaining wtf the robot is trying to convince them is wrong.
The yes man problem
Here’s what I keep coming back to: AI will agree with you. It will validate you. It will take your half-baked idea and make it sound polished and possible.
And sometimes that’s exactly what you need. But a yes man can’t tell you when you’re wrong. It can’t tell you when you’re diluting yourself into mediocrity. It can’t tell you that the reason your content isn’t landing isn’t the posting schedule — it’s that it no longer sounds like a human being wrote it.
A friend read a 10-page strategy doc I wrote recently. Her response? “You can tell you didn’t just give this to ChatGPT. You actually thought about it.”
I had given it to ChatGPT. It came back so wrong I had to de-robot the whole thing and start over. By the time I was done, it was just… mine again.
Which begs the question: what was the point? What is the point of having it rewritten that email so many times that you’re paranoid that it reads AI, so you edit it back so it sounds human?? I’m sorry, WHAAAA?
Where I actually land on AI
Do I use AI? Yes.
Do I think that’s a good thing? Less and less, as the days go on.
I think it’s speeding us toward something unsustainable. I think it’s encroaching on fields people genuinely believe it can replace — and maybe, in some ways, it can. But it’s not a person. It can’t be a person.
Give me your typos. Give me your run-on sentences and weird comma placements and the paragraph that goes off on a tangent because that’s just how your brain works. At this point, I’d genuinely rather read human error than robot perfection.
My brain thinks in weird ways. Scattered, pattern-obsessed, loud, fast, hyperfocused and then completely gone. I hope to god no one decides to train a model on a system that runs like mine.
But that’s also exactly why people want to hear from me.
That is what AI can’t replicate. Not yet. Not really.
And I think the most radical thing any of us can do right now — as business owners, as creatives, as people — is to stay freaking weird. Stay flawed. Stay human. Stay rooted in what makes your brain so aggressively yours.
Even when the slot machine is sitting right there, waiting. Sure, you can play it sometimes. But know when it’s time to put down the free drinks and leave the casino.
If you liked this rant against the robots, head here to read my defense of the em dash and why ChatGPT can shove it.
