One Good Number, Wrapped in Everything It Claimed to Replace

This piece by @hooeem opens by positioning itself against hype-driven AI content. Read it anyway, because there's one thing in it that most AI coverage misses.

The 94/33 gap. Anthropic's research distinguishes between what AI can theoretically do and what it's actually doing in the real world right now — 94% theoretical exposure versus 33% real-world deployment. That framing resets the entire timeline question. It's not "will AI take your job?" It's "what's slowing down the 61% gap, and how long does that friction last?" That's a more useful question than anything most AI commentary is asking.

The entry-level paradox is also genuinely observed: AI damage isn't showing up in layoffs, it's showing up in jobs that never get posted. Worth knowing.

But the "three scenarios" section is where the piece stops earning its claims. Scenario 3 — the optimistic case — lifts Dario Amodei's "Machines of Loving Grace" essay almost wholesale and presents it as one of three balanced forecasts. Amodei is the CEO of Anthropic. Citing the founder of a frontier AI company on the benevolence of frontier AI, without flagging the conflict, isn't analysis. It's laundering a press release as geopolitical forecasting.

The action plan at the end is fine the way tap water is fine. "Shift from execution to judgment" is correct in the same way "eat well and exercise" is correct — true, generic, and doing no real work.

Read it for the 94/33 framing. Stop before the scenarios.