A widely read essay on LinkedIn use and AI writing has resurfaced through the event’s Hacker News listing, arguing that people who rely on large language models to generate posts are making themselves sound generic and untrustworthy. The piece, published on The Observation Deck and originally posted on LinkedIn in November 2025, says the platform has become a place where AI-written content is increasingly obvious and increasingly difficult to ignore.
The author’s core complaint is not simply that AI-assisted writing can be bland. The argument is that it strips away the cues that help readers tell whether a person is speaking in their own voice. The essay says many LinkedIn posts now follow predictable formulas, with emojis, short fragments, and overused constructions that read as machine-assisted even when the topic itself is real. Once that pattern becomes visible, the writer argues, readers stop trusting the person behind the post.
That concern is framed as more than a stylistic preference. The essay says LinkedIn itself encourages users to "rewrite it with AI," which lowers the barrier to polished but impersonal content. The author argues that the result is a loss of credibility: if a post does not sound like the person who supposedly wrote it, the audience starts wondering whether the substance is also synthetic. The piece treats authenticity as a precondition for attention, not a bonus feature.
The essay does not reject AI outright. In fact, it gives the technology several legitimate uses. The author says LLMs are useful for brainstorming, understanding text, and editing. But the line is drawn at authorship. The piece argues that AI is good at helping people think and revise, while being poor at replacing the writer’s own perspective. That distinction matters because a reader may accept a tool-assisted draft, but still expect the final voice to belong to the human who stands behind it.
A recurring theme in the essay is the social cost of obvious AI use. The author suggests that once a reader spots what looks like generated language, they may stop reading altogether. More importantly, they may stop believing that the speaker is telling the full truth. In the essay’s view, that is how a writing shortcut becomes a reputational liability. The output may be competent in a narrow grammatical sense, but it can still fail as communication if it sounds detached from the person making the claim.
The piece also makes a larger point about professional networks. LinkedIn is a place where people share work updates, career milestones, and opinions meant to signal expertise. If those updates are increasingly machine-shaped, then the platform loses some of the human texture that makes it useful in the first place. The author argues that readers are not just consuming text; they are trying to judge whether a person is thoughtful, credible, and real. AI-heavy writing makes that judgment harder.
In that sense, the essay is less a complaint about bad prose than a warning about mediated identity. It says people should trust their own voice because that voice carries the credibility the message needs. For readers who have watched AI writing spread across social and professional platforms, the argument will sound familiar: the issue is not whether the text is passable, but whether it still feels like a person wrote it.
