Vol.049 — Everything Is “AI-Powered.” Nobody Tells You How Useful It Actually Is.
Date: 2026-06-23 | Newsletter
Key Summary
In Vol.049, Zenta admits he has started ignoring the word “AI” — not out of opposition (he builds AI agents for a living) but exhaustion, after one pitch for “AI-powered market research” turned out to be a staffer running a copy-paste prompt and billing for the magic word. He cites a WordPress VIP survey of 2,000 people, including 800 CMOs and enterprise buyers: 60% find “AI” in brand messaging a turnoff, 86% still want to check the original source. But “turnoff” is the wrong frame, he argues — people are simply tired. A label is meant to save work; when every product wears “AI-powered,” the proving job gets handed back to the buyer, and that verification cost is the real tax. The deeper problem for Functional AI Partners: the market gave buyers a label but no yardstick for usefulness. His fix is four outcome questions — net time saved after checking, whether it fails loudly or quietly, whether a non-expert can catch its errors, and whether it fits your actual job — none of which mention “AI.” In ASEAN the tax is heavier where AI literacy is lower. Do it simple. Stop asking if it’s AI. Start measuring if it helps.