Not the flood

When everyone sounds the same

Something gave way on LinkedIn this spring. The feed filled with posts nobody quite wrote, comments that only restated the line above them, and applications polished to a shine and hollow underneath. So LinkedIn built a filter. The company says it now flags generic AI content with about 94 percent accuracy, and rather than delete that content, it buries it, capping the reach to the author's own immediate network. The Next Web reported the rollout in May, along with the name the platform chose for the problem. AI slop.

The filter does not stop at posts. LinkedIn extended the same detection to job applications, where a generic AI application costs a person the most.

The flood was the goal

It helps to remember how the feed got here. A generation of tools promised to make applying effortless. Point them at a thousand listings, let them write a thousand letters, and apply while you sleep. The pitch was volume. More applications, less work, better odds.

The odds did not move. The flood did. When every application reads the same, none of them reads as a person. Resume Now's 2026 survey of hiring managers found that 62 percent reject AI-generated resumes that show no personalization, and 36 percent name generic content among the top reasons they pass on a candidate. The tools meant to open doors taught employers to distrust whatever came through them.

What we chose not to build

CoBlack could have shipped a spray tool. It would have been the easier product, and for a while it would have sold. We did not, and the reason is the whole of what we believe.

Automation should make a person clearer to an employer, not blur them into everyone else. It should carry the truth of what someone can do, faster than they could carry it alone. The moment it starts manufacturing a generic version of you to flood a system, it has stopped working for you, whatever the dashboard says.

Proof you do not have to fake

So CoBlack points its automation at the opposite of volume. It matches on capability, the scope of work a person has actually handled and the problems they have actually solved, drawn from validated employer career pages and applicant tracking systems rather than the open boards. It tailors each application to the one opening in front of it, never to a thousand at once. And it shows the reasoning behind every match in plain language before anything is sent, so the case it makes for you is one you could defend in a room.

That is slower than a spray. It is meant to be. A real signal takes a little more than a fake one, and it is the only kind that still travels.

More you, not less

The slop filters are not the enemy of the job seeker. They are a rough, overdue correction to a market that confused motion with progress. They will catch honest people alongside the lazy ones, because a 94 percent filter sorts by trend more than it judges with care. That is the cost of the flood. Everyone pays for the noise.

The way out is not a better disguise. It is to stop disguising. Send less, send true, and send it straight to the people doing the hiring. The candidates who come through what follows will be the ones who still sound like themselves, the one thing the machines were never able to mass-produce.