Stop Approving AI Content That Sounds Fine - But Doesn’t Sound Like You

I choked on my coffee when I read the opening line of the post ChatGPT had just written. "I've seen this movie before." Not because it was surprising. Because I'd seen that exact line before, in a different campaign, for a different person.

This client was a senior technology executive. Twenty-five years of real enterprise transformation. AWS, HP, IBM. We had spent two hours rebuilding his brief from scratch. The whole point was to sound like him, not like every other person on LinkedIn suddenly discovering they have thoughts about AI.

And still. That line.

So I pushed back. Told ChatGPT exactly what was wrong. It apologised, agreed with everything I said, told me it understood completely.

Then it served me the same post again with different words.

I've had arguments with ChatGPT that got pretty heated, which is a strange thing to admit about a piece of software. But there is something genuinely maddening about a tool that tells you it understands exactly what you're asking for and then can't deliver it. It's not being difficult. It's structurally weighted toward content that gets approved, and your instructions barely register against that gravity.

The shape is always the same. Clean hook. Relatable contrast. Punchy stacked questions. Tidy conclusion. CTA with the answer already baked in.

It looks fine. You could publish it tomorrow and nobody would question it.

That is the problem.

ChatGPT isn't trying to sound like you. It's trying to produce something you'll accept. Those are different jobs, and it's very good at the second one.

Don't trade your authenticity for convenience. The platform is already full of people who did, and most of them don't know it yet.

I'm curious, if you went back through your last ten posts, how many actually sound like you?

hashtag#PersonalBranding hashtag#AIContent hashtag#LinkedInStrategy hashtag#Authenticity