Decks About Decks
I.
The meeting invite landed late on a Friday. Urgent red exclamation mark. “Re-calibration session.”
I flicked my colleague a Teams message. Didn’t we have that session this morning? Melting-into-the-floor emoji.
He’d already replied to a five-message thread on the outcomes of the morning’s session. We were, apparently, lacking clarity. The exec still had gaps that needed filling.
What gaps? We went through it in detail.
Yes, but X wasn’t in attendance.
I see.
II.
If you took an honest audit of a week in most businesses, mine included, a serious portion of it would look like that. Meetings to align on meetings. Decks summarising decks. SOPs to organise the SOPs. Pre-reads nobody reads. A weekly cadence of artefacts produced for no one, reviewed by no one, acted on by no one, yet somehow driving our collective stress and energy.
We’ve built whole careers around being good at this. Around feeling busy in a way that reads as productive. Around being the person who walks into the recalibration of the recalibration with a tidy deck and a confident tone.
III.
The LLM conversation has been running for two years now (woah, where did that go) and it’s mostly two camps shouting past each other. One side says the work will be automated. The other says only humans can really do it. Both sides are arguing over who gets to do the work. Almost nobody is asking whether the work deserved to survive in the first place.
A serious slice of what we call knowledge work was already noise before any model could write an email. Status reports nobody read. Quarterly decks built to demonstrate effort. Process documents inherited from someone three restructures ago. AI didn’t do this. But it’s accelerating it.
IV.
The cruel thing is that the people doing this work were doing exactly what their jobs asked of them. I want to be careful here, because I’ve been on both sides of it. They were rewarded for it. Promoted for it. They optimised their careers around being excellent at what the system demanded, and the system demanded a lot.
I’ve asked for the deck. I’ve sat in the recalibration session and nodded. The point isn’t that those people were robotic. The point is that the system rewarded robotic behaviour, and the humans inside it adapted, the way humans always adapt. The cruelty isn’t AI replacing them. The cruelty is that parts of the job had already become a dead end, and the system was happy to let them keep walking down it. In a large part, perhaps it had overtaken the actual crux of the things we were once doing.
V.
The question I pose now isn’t whether AI does the work. It’s whether we keep doing the work at all.
If a model can produce the quarterly deck in nine seconds, the honest response isn’t to feel threatened. It’s to notice that the deck was the problem all along. The recalibration session was the problem. The five-message thread on agenda outcomes was the problem. The artefact was never the point; we just got very good at pretending it was.
What’s left, once you strip the theatre out, is the actual work. The judgement. The taste. The decision someone has to own. The difficult conversation. The call to stop a project. The decision to make something new. To Create.
That’s work I’m excited to sink my teeth into.
