I don’t spend time on LinkedIn anymore. I open it, scan the feed and feel like everyone is now posting in a language they don’t actually speak. It used to be that I could reliably find something that made me stop because the person sharing it had clearly wrestled with the idea before publishing. The posts felt real – grounded in personal risk, missteps and discovery. Those ingredients still show up today, but increasingly, they feel engineered rather than drawn from lived experience.
You know the posts. The reflection on a new certification that promises to drive performance but never explains how. The summary of a video labeled fascinating without a word about why it’s fascinating. The carefully staged announcement about being honored and humbled, complete with professional photography and tagged executives.
None of this is inherently wrong. We should celebrate milestones and what we’re learning. But too often the point of the post isn’t the insight. It’s the image. It’s about looking polished, current, impressive and thoughtful all at once.
So, what happened?
The Problem Isn’t AI
Data from Originality.AI suggests that more than half of longer-form LinkedIn posts were likely AI-generated or assisted, a massive spike since the arrival of accessible LLMs. That’s not a failure of technology, but rather a symptom of how AI has lowered the psychological barrier to publishing. Writing is hard; generating a polished, middle-of-the-road essay is easy. But volume does not equal authority, and constant exposure to AI-generated output is eroding credibility.
AI is not the enemy. I use it every day. It’s a powerful tool for organizing thinking, strengthening structure, and improving clarity. So, the issue isn’t the innovation itself, but how casually we’ve handed it the work that should belong to us.
AI should function as a content assistant, not as the source of our convictions. We’re supposed to form the point of view. When we hand our thinking off to AI, we become the blandest version of ourselves. By design, AI produces the statistical middle. It synthesizes what’s most likely to resonate across the broadest audience. But meaningful ideas rarely come from consensus. They come from proximity, experience, our mistakes, and from caring enough about a topic to sit with it before publishing an opinion.
AI can help you articulate what you already believe and refine your language. It should not decide what matters to you. That substance has to originate with you.
If we’re going to use AI well, then we need a higher standard.
First, earn the right to speak. Thought leadership does not begin with a trending topic. It begins with sustained interest and lived experience. The issues you care about enough to study deeply are the ones you’re equipped to speak on credibly. It means you’ve navigated something long enough to understand where the nuance lives. It does not mean skimming a headline, prompting a summary, and publishing a polished reaction 15 minutes later.
Second, form your point of view before you open the tool. AI can refine an argument, but it cannot originate one. If your first encounter with an idea happens inside a prompt, you aren’t leading the thought. Real opinions are shaped by friction. They come from sitting with an idea long enough to notice where you agree, where you resist, and where your experience complicates the narrative. If you skip that step, what you publish may sound coherent, but it will not be anchored in anything you’ve lived or tested.
Third, protect your voice. Feed the tool your language such as past original writings and your natural cadence. Do you tend to be direct or reflective? Do you write in long, layered sentences or short, declarative ones? Do you rely on data, story, or both? Do you use superlatives, or do you avoid them?
Those patterns are your voice. AI can mirror those tendencies back to you, but it cannot invent them. You still have to recognize yourself in the output. Edit until the piece sounds like something you would say without a script. If it doesn’t, it’s not ready to publish. And if it cannot be disagreed with, it likely does not go deep enough to matter.
LinkedIn was built on earned perspective; it will only survive if we protect that. If we want the platform to matter, we must be willing to do the work ourselves – to write with a voice that is recognizably human and an argument that is recognizably ours.
