Writing is no longer proof of work

For years, a lot of company writing was never really about communication. It was evidence. A risk assessment showed someone had followed the ritual. A giant consultant report showed effort had been spent. A status update showed a manager had actually understood what was happening below them.

That matters because the writing itself often wasn’t the point. In the source’s framing, these documents worked a bit like proof-of-work: nobody cared much about the content, only that someone had sat down and produced it. The system relied on trust, plus the occasional spot check. If a report looked fake, someone might read it closely. If a PR broke production, the risk assessment might suddenly matter.

LLMs mess that up. A hundred-page report can now be produced quickly and cheaply. A detailed status update can be generated without any real understanding behind it. That makes writing a much weaker signal of compliance, effort, or learning than it used to be. Schools already know the problem. So do political and regulatory bodies that use formal writing as a filter.

The awkward part is that there isn’t an obvious fix. Organizations may push toward video, live presentations, stricter bans on AI-written text, or some new way of judging effort. None of those options looks clean. So there’s a good chance the old process-heavy world stays in place for a while, even as everyone knows the evidence is easy to fake.

The practical advice in the piece is blunt: don’t spend time hand-writing documents that only exist as evidence, and don’t waste time treating AI-authored evidence as if it still means much. Save the manual effort for the rare documents that are meant to actually communicate something.