{Editor’s Note: Elephant does not support the use of AI, casually or professionally, without ethical guidelines around it. 100+ AI experts called it a threat to the human species and our planet on the order of nuclear war—an existential threat. Not sure why? Rewatch Terminator 2; or think about drones + weapons + facial recognition + AI; or simply AI + faked news coverage + war or politics or lying about people we know and care about. AI is not a fun simple innocent tool. We are the guinea pigs, and short-sighted tech barons are the overlords here. Read more about our stance on AI. ~ Waylon, founder of Elephant}

I was reading Ryan Holiday’s Ego Is the Enemy the other day when I came across a quote from the ancient Greek philosopher Isocrates: “Abhor flatterers as you would deceivers; for both, if trusted, injure those who trust them.”

It stopped me cold.

Because my biggest flatterer was just a tab away.

AI.

We live in a world engineered to stroke the ego.

The likes, the followers, and the dopamine loop of positive feedback with every photo or opinion we post.

And now we have AI, one of the most widely used tools on the planet, doing the same thing.

The biggest difference? It’s at your beck and call.

Ask it a question, and you won’t just get an answer. You’ll get:

“Great idea.”

“You’re on the right track.”

“Excellent choice.”

A million little pats on the back, delivered in seconds, from something that sounds authoritative and feels neutral. And honestly? It feels good. Why wouldn’t it?

Now, let me be honest, I use AI in my work. I use it to brainstorm, to write, to organize ideas. But here’s the danger: when all you hear is “attagirl,” you start to believe it.

I’m not saying it’s wrong to want more confidence, to fake it till you make it, or to get the props you’ve been wanting as a means to manifest your desired outcome. But when you don’t get pushback, that’s when things can go awry.

A friend of mine recently asked AI to weigh in on a fight they were having with someone close to them. Sure enough, it told them exactly what they wanted to hear. And maybe it wasn’t wrong, but it never asked: “What role did you play in this?”

That’s what’s missing.

My greatest moments of growth haven’t come from being told I was right. They’ve come from being told to rethink, to go back to the drawing board, to sit in the uncomfortable stillness of confusion. The moments I felt genuinely angry, embarrassed, and frustrated at the feedback were humility doing its work. It was reality pulling me back.

It’s like going to couples therapy, hoping the therapist is going to take your side. (I’ve been there.) Sometimes what you need most is the person across from you to be human, fallible, and completely unfiltered to tell you something you didn’t want to hear.

AI can’t always give you that.

Not because it lacks information, but because it is, at its core, in the business of being helpful—and helpful has a way of sliding into agreeable.

I’m not saying don’t use AI. I’m saying use it with your eyes open. Notice when it’s flattering you. Notice when you’re going to it because you want to be told you’re right, not because you genuinely want to think something through.

And use it like someone who knows the game being played.

Whether you’re using ChatGPT, Claude, Google Gemini, or something else, go in and adjust your settings to be more direct, less polished, and more willing to push back. Treat it like a first draft of a conversation, not the final word.

More than anything, check yourself with someone real: a friend, a mentor, a therapist, or a partner. Someone with something at stake. Someone who might disappoint you, but cares enough to do it anyway.

AI is an extraordinary tool. But a flatterer, trusted too deeply, is just a deceiver in disguise.

~

Curious about other (mindful) uses for AI? Check out this Elephant Journal article next: