The Heuristic Is Good. The Rest Is Staging.

Claire Vo, writing on X, with the kind of energy that makes you want to both forward it to your CEO and gently edit it before you do:

I have a simple heuristic for identifying who has been kicked out of the arena and it's this: They can't do anything the same day.

The heuristic is good. When engineering was the bottleneck, slow meant constrained. Now slow means broken. AI didn't create your organisation's dysfunction — it just made it legible.

Everything before and after that line is a different article.

The diagnosis is sharp: mid-market software companies mistaking motion for speed, slapping natural-language interfaces on calcified processes, telling the board they're "cursor for X" while the renewal pipeline quietly empties. Real. Recognisable. Worth the read.

The prescription is where it goes sideways. Fire resistant executives. Fire resistant engineers faster. Rebuild from scratch in under two weeks. Rearchitect the company in markdown over a weekend. Bring everyone back to the office — or open Claude Code at the all-hands and declare that the new office. These aren't a plan. They're a mood board.

The deeper problem: the post is written to make you feel you are probably in denial, probably doomed, probably not as far along as you think. That's a FOMO structure dressed up as a diagnosis. The insight doesn't need it — and the audience Vo is actually trying to reach will use the anxiety as a reason to forward the post rather than change anything.

The grief-cycle framing, the "ngmi" shorthand, the startup romanticism (DM their best engineers! clone their product for 20% of the price! be obviously Having More Fun™ online!) — this is the packaging around a useful observation about organisational speed.

Strip the packaging. Keep the heuristic.

If your company can't do anything important the same day — not everything, anything — that's not an AI problem. It's a structure problem that AI has made too expensive to ignore.

That's the post. Everything else is staging.