The meeting that finally broke me was an hour long. The whole team, pulled off work and into a call, to read out answers to a list of questions so that someone a few levels up could tick the matching boxes. That was the entire function of the hour. We spoke, they filled in a grid. No decision came out of it, because there was nothing to decide. It was data collection with a dozen salaries in the room, and it should have been a form.
I’ll be upfront that this is a rant. But there’s an argument under it, and the argument is that the case for not holding the meeting has quietly gotten stronger.
Meetings Were Always a Symptom
Most recurring meetings were never really a tool. They were a symptom of one of two things: status that should have been written down, or a decision nobody was empowered to make, so a group got convened to share the weight of not making it. That was true long before any of the current tooling existed, and it’s the older half of this argument.
What changed isn’t the diagnosis. It’s what the alternative is now worth.
Text Is the Only Version Your Tools Can Use
Writing things down was always better for humans. It’s searchable, skimmable, readable on your own schedule, and it quietly forces the writer to think the point through instead of waving at it on a call. None of that is new.
What’s new is that text is now also the only version of the conversation your tools can do anything with. A written update, a design doc, a decision recorded in a thread, all of it is a corpus. You can summarize three days of discussion in thirty seconds, ask what got decided about a thing and get a real answer that points at where it was decided, hand a newcomer a digest instead of a calendar. A meeting produces almost none of this. Even with a transcript, an hour of meandering talk is low signal, and the 200-word decision someone wrote down afterward is high signal. The written one is what your tooling can actually work over. So the thing that was already the right call now pays a second time.
The Sync That Resolves Nothing
None of which stops people from defending the daily sync to the death. The one where the same status gets recited for the same fifteen minutes, the same open questions stay open, and the meeting ends roughly where it began except it’s now later in the day. The slot fills itself regardless of how much there is to say, which is just Parkinson’s Law showing up on the calendar one more time.
If that content is status, the written version is faster for the people and legible to the machines, and the live version is pure overhead. Keeping it anyway is a preference for the ritual of meeting over the result of communicating.
The Form Test
Then there’s the category my broken hour belonged to, which is larger than anyone admits: meetings whose only job is to move information out of a team’s heads and into someone else’s spreadsheet.
There’s a clean test for these. Look at what the meeting produces. If the output is a filled-in set of answers, a grid, a status rollup, a list of numbers, then it was a form, and you made twelve people attend a form. If the output is a real decision that needed argument to reach, fine, that one might have earned the hour. Most recurring meetings produce the first thing and present it as the second.
Where AI Makes It Worse
Let me turn the rant on my own side, because AI enables a fresh mistake too.
The notetaker bots are everywhere now, and they’ve quietly become permission to keep meeting. “Don’t worry about notes, the AI will summarize it” sounds like efficiency, and is often just a reason to hold a meeting that shouldn’t have happened, now with a summary attached that nobody opens. A summarized pointless meeting is still a pointless meeting. The summary isn’t the fix. Not holding it is the fix.
I’m aware of the irony, because I’ve built one of these tools, something that transcribes and summarizes meetings on-device, and I still think most of the meetings it would summarize shouldn’t exist. The tool earns its place on the few that genuinely needed to happen. It is not a license for the rest.
And to be fair to the other side, some things do still want a live room. A contested decision with real nuance, early ambiguous thinking done out loud, anything with friction between people. Async isn’t free either, and it only works if a team actually writes and actually reads. The point was never zero meetings. It’s that a meeting should be the rare exception you reach for when text genuinely can’t do the job, and most of them fail that test on the first question.
The Takeaway
So before you put an hour on a dozen calendars, look at what comes out the other end. If the answer is information that was always going to end up in a document or a grid, you didn’t need a meeting, and in the AI era you don’t even need everyone in the room while it gets written. The doc can be read, summarized, and queried by the whole team, tools included, on their own time.
The hour I sat through produced a filled-in grid. It should have been a form, and the genuinely insulting part is that a form would have handed everyone upstairs a cleaner, searchable answer than the meeting ever did. So, what’s the last meeting you sat in that was just a form with extra steps?