Independent newsroom The Wyre News Network OpEd desk

Perspective 5 min read

LinkedIn Is Bots Talking to Bots

An AI wrote the post, an AI wrote the comment praising it, and an AI summarized both for an audience of more AI. We build these systems, so here is the honest take on the bot-to-bot loop: how it closed, which parts are pure theater, what is somehow still real, and how a professional who actually wants outcomes gets them.

By Amber Stone for the Wyre OpEd desk

Picture the timeline. An AI wrote the post. An AI wrote the top comment praising the post. An AI summarized the post for the people who did not read it, and an AI drafted their replies. Somewhere in that chain a human was theoretically the audience, but you would be forgiven for not finding one. LinkedIn has quietly become a machine performing professionalism for other machines / and the performance is getting very good.

We build AI systems for a living / content pipelines, agents, automation, the works / so this is not a Luddite's complaint. It is the opposite. We know precisely how the sausage is made, because making the sausage is the job. Here is the honest ledger on the bot-to-bot loop: how it closed, which parts are pure theater, what is somehow still real, and what a professional who actually wants outcomes should do about it.

How the loop closed

Nobody sat down and designed a network where robots applaud each other. Three forces converged and the equilibrium fell out on its own. Reach became the currency, so everyone started optimizing for it. Generative writing got good enough to fake a point of view, so producing "content" stopped costing time or thought. And automation tools turned engagement into a background process, so liking, commenting, and connecting no longer required a person to be awake. Free production, automated distribution, a scoreboard that rewards volume: give any system those three inputs and it will do exactly what it was built to do / produce endlessly and clap for itself. This is not a glitch in the platform. It is local optimization arriving at its natural destination.

The tells

The house style that ate the platform. The dramatic one-line paragraph. The manufactured cliffhanger before the "see more." The tidy three-bullet insight. The closing "Agree?" fishing for a reply. These stopped being writing choices and became the default output of the same handful of tools, which is why ten thousand "thought leaders" now think in the identical rhythm. When everyone's voice converges on one voice, that voice is a model, not a movement.

Engagement that arrives in ninety seconds and says nothing. The comment that restates your headline back to you as a compliment. "Great insight." Praise with no fingerprints on it, no evidence anyone read past the first line, posted faster than a human could have finished reading. That is not an audience. That is an automation keeping a streak alive.

Content about content, all the way down. The post summarizing an article. The carousel summarizing the post. The comment summarizing the carousel. At no point does the chain touch the actual work, the actual client, or the actual result. It is a hall of mirrors that mistakes its own reflection for a room full of people.

Why nobody turns it off

Because every incentive in the system points the same direction: more. The creator wants reach. The tool vendor wants seats. The platform wants time-on-site and does not particularly care whether the time is spent by a person. And the metrics that are supposed to referee all of this / impressions, engagement rate, follower growth / are counted by the same machinery producing the activity. The scoreboard goes up whether or not a single human was informed, persuaded, or moved to do anything.

We have made this exact argument about ad channels on this blog: precision without truth is the most expensive kind of lie, because it feels like rigor while it measures nothing. A number that rises when bots talk to bots is not tracking your professional reputation. It is tracking the bots. Confuse the two and you will happily pour real hours into a channel that is applauding an empty room.

What is somehow still real

Now the fair part, because cynicism is just as lazy as hype. The platform still does real things, and they are worth protecting from the noise. The direct message where an actual hiring manager reaches an actual candidate. The narrow professional community where practitioners trade problems they genuinely have. The occasional post from someone with real scars, writing the thing only they could have written / and you can feel it instantly, because a model can imitate structure but not experience. The value did not leave LinkedIn. It got buried under synthetic volume, which is a very different problem, and a more solvable one. You do not need the platform to change. You need to stop grading yourself by its slot machine.

What to do instead

Write like a person with something at stake. The one thing the loop cannot manufacture is a specific, falsifiable, slightly risky opinion earned by doing the work. Bots generate the safe consensus flawlessly and infinitely. So say the thing that might be wrong. Name the tradeoff nobody wants to name. The machine can out-produce you forever on volume and will never once beat you on conviction.

Measure conversations, not impressions. The same discipline we preach for every marketing channel applies here. Reach is the vanity metric. The real one is whether a person you actually wanted to reach replied, messaged you, or showed up. One genuine reply from the right human outweighs a thousand automated claps, and the two are trivial to tell apart the moment you look for a fingerprint.

Move the relationship off the stage. The feed is the lobby, not the meeting. Real professional outcomes still close in direct messages, on calls, in rooms. Treat public posting as an invitation to a private conversation, and judge it entirely by whether that conversation ever happens.

Point the AI at the work, not the performance. This is the tell that separates operators from posers. Generative tools are spectacular at the actual job / drafting, analyzing, cleaning data, moving a process forward / and merely mediocre at pretending to have a personality on your behalf. Aim the machine at the former and you compound real leverage. Aim it at the latter and you have automated your own irrelevance.

Our position

LinkedIn did not die. It automated. The professional network quietly became a professional-content-generation network that largely entertains its own bots, and the frustrating part is not that AI is involved / we build the AI, and it is genuinely, unglamorously useful. The frustrating part is what it got aimed at: manufacturing the appearance of insight instead of the insight, farming applause instead of doing the thing that would earn it. That is a choice, not a fate.

The same tools writing broetry for an audience of comment bots could be reconciling a messy dataset, drafting the hard first version, or clearing the busywork off a real expert's desk so they have time to say something only they could say. The machine was never the problem. Pointing it at applause instead of work is. Build systems that do something. Let the bots keep the feed.

Related Reading

More Opinion

From the same desk

Perspective

A Rounding Error

Over twenty-four hours the single largest AI crawler on one consultant's website arrived roughly 1,500 times under a nonprofit research archive's name, sent back nothing, and asked for his SSH keys, his AWS config and his Firebase service account key. Across a hundred such paths: 1,028 requests, 6.7 megabytes, zero referrals, and requests for anything he had actually written rounding to nothing. It appeared in no security log, because security logs record rule trips and he was not blocking it. It appeared in exactly one place: the AI crawler dashboard, counted as an audience. Cloudflare's CFO says humans will be a rounding error on the internet within five years and admits he has been wrong every time, always low. Machine traffic passed human traffic in May 2026, two years earlier than the company forecast. Why robots.txt is a request that OpenAI's own documentation says may not apply, why blocking the obvious way costs you the visibility and keeps the fetches, what it means that agent config files are now on the standard secret-scanning wordlist, and the afternoon of work almost no company has done.

13 min

Perspective

Cut The Price Last

When demand softens, the first lever everybody reaches for is price. It is also the most expensive lever on the board, the hardest to reverse, and almost never the one that was actually broken. Redfin reports buyers have negotiating power across most of the country, the median housing payment fell to $2,575 because sellers cut asking prices to a one-year low, and pending sales hit their lowest since early April. Underneath it, July payrolls came in at minus 23,000 against an expected plus 83,000 with another 103,000 revised away. So the softening is real, and the question is which lever you pull and in what order. Why cutting price assumes a diagnosis you have usually not earned, the cheaper levers ranked, why the gap between the decision and the execution is where the margin leaks, and the measurement problem underneath all of it: cost per lead holds beautifully steady while cost per closed sale doubles.

6 min

Perspective

The Layer Nobody Looks At

A man in Connecticut typed instructions into his own court filing in three-point white text on a white background, where no judge would see them and any software would. They told the system to agree with him and to disregard the court's earlier denials. He was caught, warned, did it again, and lost his e-filing privileges. In the same seven days Google made the visible watermark on its AI images optional while keeping the invisible one, Anthropic shipped an API letting third parties detect text its models probably wrote, a new benchmark found that no frontier model clears sixty percent on basic visual perception with counting, localization and OCR among the weakest skills, and OpenAI started turning Mac users' clicks and keystrokes into a searchable memory their assistant can read, while warning in its own documentation about prompt injection. Five stories, one story. Every document, image and screen you produce now has two readers and they do not see the same thing. What the judge worked out on his own about where these attacks get caught, why provenance just moved into the layer humans cannot check, what a watermark hit actually proves and why a clean result proves nothing at all, and the ten-line diff almost nobody runs.

18 min