There’s been a recent blast of “I’ve told Claude to take over my socials” posts.
You will of course see the people desperate enough commenting “Prompts” or “Lucky” or whatever phrase the author has asked you to in order to get their “special/no fail prompts” to do it yourself. (I wrote about this type of bs here).
Scroll far enough through any “use Claude/ChatGPT to run your socials” post and you’ll find the comments doing something the caption never will, telling the truth.
“Congrats, now it looks like everyone else.”
“But I can tell it was made with AI.”
“We can all tell you made it with Claude too. Congrats. Now your slides look like every other person using Claude.”
“So you’ve moved from a limited amateur design tool to one-shotting a generic carousel. Wonderful.”
That’s not trolling. That’s market feedback. It’s the exact feedback we’ve been giving clients for years, just faster now.
We’ve said this before about copying competitors, and about chasing tactics instead of building strategy.
If you can find it, so can everyone else.
AI tools didn’t create that problem. They just removed the last bit of friction that used to slow it down.
Before, copying a competitor’s playbook took effort. You had to notice it, understand it, adapt it. That lag was accidentally protective. By the time you’d cloned their approach, they’d moved on, and your version looked like your version, not a photocopy.
AI collapses the lag to zero. Type in a prompt, get the output, ship it. Which means everyone chasing the same “shortcut” arrives at the same place, at the same time, looking the same way. The tool didn’t just speed up the tactic. It erased the variation that used to disguise how many people were doing the identical thing.
Sameness is not a style problem
The instinct is to read those comments as “the AI templates look bad” and go hunting for a better prompt, a different tool, a more “human” style guide. That’s still tactical thinking. It treats the output as the thing to fix.
The actual problem sits one level up, if your content strategy is “have AI make my content,” you don’t have a content strategy. You have a production method. A production method with no point of view behind it will always converge toward the mean, because the model is trained to produce the statistically likely version of “good,” not your version of true.
This is the same gap we point at when a business owner asks us to help them “do what [competitor] is doing.” The tactic isn’t the differentiator. It never was. What differentiates is the thinking underneath it, the specific read on your audience, the position you’re willing to take, the thing you know that a general-purpose model doesn’t.
What actually holds up
None of this is an argument against using AI. It’s an argument against using it as a substitute for a point of view instead of an accelerant for one.
The founders whose AI-assisted content doesn’t get called out in the comments are the ones feeding it something specific: their actual customer conversations, their actual data, their actual opinion on the topic, stated plainly before the tool ever touches it. The model amplifies whatever you put in. If what you put in is “make me a carousel about leadership,” you get the same carousel about leadership as the other ten thousand people who typed that exact prompt this week. If what you put in is a real argument, sourced from your own experience, the output has somewhere to go that theirs doesn’t.
So the question isn’t “should I use Claude for my content.” It’s “do I have something to say that Claude didn’t already say for someone else this morning.” If the answer’s no, no tool fixes that!!! It merely publishes the gap faster, in front of a bigger audience, with a comment section increasingly happy to point it out.
Strategy first. Then let the tools move faster. Not the other way around.
