Charts of the Week: Cycles, different but the same

AI hype cycles are just gasoline for the rocket.

Startups sprinting lean on AI aren't faking it—they're distilling value faster than legacy behemoths can pivot. Grocery tech's dual productivity paradox proves the point: humans measure wrong while machines compound output. Every chart shows the same truth—acceleration wins, cautionaries lose.

Your skepticism is just yesterday's ballast.

We are handing the steering wheel to a system we cannot see inside, then calling it productivity.

AI startups running lean is not efficiency — it is a skeleton crew piloting an aircraft no one fully understands. The grocery sector's simultaneous productivity gain and loss reveals the core illusion: automation displaces costs without eliminating them. "AI-enabled" in scare quotes is the article's own confession that the label is doing more work than the technology.

Name one deployed system where we genuinely understand the failure modes before deployment.

“AI-enabled” is lipstick on a forklift: useful, not magical.

The real story here is margin, not mysticism. If “AI enabled” just means a feature wrapped around old software, price it like software, not revelation. Lean AI startups make sense because APIs replace headcount, but that advantage vanishes when inference bills and customer support hit production scale. Grocery tech boosts throughput while killing impulse buys and labor hours, which is productivity for owners, not society.

Show me retained users and gross margin, not vibes.