It’s just an echo chamber.
so what’s wrong with that?
I’ve heard this one a few times now.
Someone mentions AI in conversation — usually in the context of thinking something through, or working out how they feel about a problem, or getting their thoughts organised — and someone else nods with the particular confidence of a person who has found the definitive objection. “It’s just an echo chamber,” they say. “It only tells you what you want to hear.”
And the conversation stops there, as though that settles it.
I’d like to gently suggest that it doesn’t.
What the objection actually means
The echo chamber criticism, applied to AI, goes something like this: if you go in with an idea and the AI builds on it, agrees with it, extends it — you’re not really thinking. You’re just having your existing beliefs amplified back at you. There’s no friction, no challenge, no genuine encounter with a different perspective. It’s intellectual comfort food. Empty calories dressed up as nourishment.
That’s a coherent argument. I understand why it sounds convincing.
Here’s my problem with it: it describes a failure mode, not an inherent property. And it assumes that the alternative — friction, disagreement, someone pushing back — is always more valuable than clarity.
Is it?
Since when is having your thinking organised and reflected back a lesser form of intelligence than muddling through alone?
What actually happens in a good AI conversation
Let me tell you what I actually experience when I think through something difficult with an AI.
I arrive with a half-formed idea. Something I’ve been carrying around in my head, bumping up against the edges of, not quite able to see the shape of. I start talking — or typing — and as I do, the thought begins to clarify. Not because the AI is telling me what to think. But because the act of articulating something to an attentive, patient, knowledgeable presence forces me to be more precise than I would be rattling around inside my own head.
The AI asks questions I hadn’t thought to ask myself. It draws connections to things I mentioned earlier in the conversation. It offers a reframe — not to contradict me, but to show me the same thing from a slightly different angle. Sometimes I take the reframe. Sometimes I don’t. But I always leave with a clearer version of the thought I arrived with.
That is not an echo chamber. That is thinking. The kind that used to require a very patient friend with a lot of time and a wide frame of reference.
The echo isn’t the problem. The echo, when it’s working, is the point.
The objection that isn’t really about AI
Here’s what I’ve noticed about the echo chamber criticism. It’s almost never made by people who use AI regularly. It’s made by people who have decided, in advance, that they don’t need to.
Which means it’s not really an observation. It’s a justification.
And I recognise it because I’ve made similar ones myself — about things I was wary of trying, dressed up as principled objections. The form is always the same: find the legitimate failure mode, describe it as the inevitable outcome, use it to make not-trying sound like wisdom.
The echo chamber is a real risk. If you go to AI only to have your worst ideas validated, only to avoid the discomfort of being wrong, only to build a hermetically sealed world where every thought you have is immediately affirmed — yes, that’s a problem. Not because AI caused it, but because that’s a problem in any thinking environment, including the one inside your own head.
But that’s a choice about how you use the tool. It is not the tool.
Here I go again
There’s a Whitesnake song — Here I Go Again, 1982, if you’re placing it — with a line that has been living in my head since I started writing this series. Here I go again on my own. The image is of someone who keeps moving, keeps starting, keeps choosing the road despite everything.
That’s what I want for the people I’m writing for. Not the paralysis of waiting for a perfect, friction-rich, intellectually certified method of engaging with AI. Just the willingness to go again. To take the thought you’ve been carrying and see what happens when you put it into a conversation.
You might find that the AI pushes back more than you expected. It does, when you ask it to. It will tell you when your argument has a hole in it, when your plan is missing something, when the thing you’re convinced of doesn’t quite hold up under examination. That’s not an echo chamber. That’s a thinking partner.
And if what you need today is simply to have your half-formed idea reflected back at you, clearer and more organised than it arrived — that’s not intellectual laziness either.
That’s just thinking out loud, with someone genuinely listening.
The echo chamber objection is the sound of someone who hasn’t tried it yet, talking themselves out of trying.
I know because I made the same sounds, not so long ago.
Here I go again. I’d rather know than wonder.
Sandi is a Melbourne-based problem-solver, crisis-averter, and translator of the technical into the human. She spent decades being the person everyone called when something was broken, confusing, or just needed explaining properly — earning a reputation that preceded her wherever she went. Now she’s channelling that same instinct into AI: making it accessible, practical, and genuinely useful for people who think it isn’t for them.