← The AI Hype Audit — all 100 verdicts
HOLDS UP "Claude just outplayed Apple" — true story, verified to the repo. Told by a marketer under an FTC order.
The claimBen Zhang, a robotics engineer, lost his phone at the office. Find My was disabled by his company's MDM. He asked Claude, which suggested tracking the phone's Bluetooth signal strength and built him a live signal meter in about a minute. He walked the office watching the number climb and found the phone. He open-sourced it as 'findphone.'
Every element checks out at the source: the repo is live (ben-z/findphone — 'locate a nearby Bluetooth device by signal strength, from the macOS command line, for when Find My isn't available'), Ben Zhang told the story on his own X account, it hit Hacker News, and the on-screen terminal in the reel matches the actual tool's output (dBm readings, SAME ROOM indicator, peak tracking). MDM disabling Find My is standard corporate practice, and the RSSI approach is sound engineering. The story is true. The teller is the catch.
What holds up
- github.com/ben-z/findphone verified live — macOS CLI, Bluetooth RSSI tracking, exactly as described
- Ben Zhang's first-person account on X + Hacker News thread confirm the origin
- The reel's terminal frames match the real tool's UI (dBm, SAME ROOM, peak/min, ~1s refresh)
- MDM routinely disables Find My on corporate devices — the setup is plausible, not staged
What doesn't
- The poster: Anik Singal settled with the FTC (2023) over Lurn's deceptive income claims — $2.5M redress and orders against misleading earnings marketing (public record: ftc.gov). We name him for the same reason we named the #034 and #035 promoters: federal receipts
- Nothing is sold in this reel — it's authority-farming: true viral stories build the audience the next offer is sold to
- 'In one minute' is roughly right for the first cut, per Zhang himself — but it's Claude + a capable engineer; the minute is not the whole story
The catch
A true story can still be a funnel. This reel is accurate down to the terminal output — and it's posted by a marketer with an FTC settlement on income claims, building AI credibility with other people's engineering. Enjoy the story, verify the storyteller, and remember the two audits where his name already appears.
How to actually do it
- Mac users: clone github.com/ben-z/findphone and run it per the README next time a Bluetooth device goes missing
- The generalizable lesson: when the official path is blocked (MDM, lockouts), describe the physical problem to Claude — signal strength, timestamps, logs — and ask what's measurable
- File under real Claude wins for client conversations: small, verifiable, no income claim attached — the exact opposite of how it's being used in this feed
- Related entries on this poster: #034 (his '$3B AI Berkshire' pitch), and the FTC-receipts standard set at #035
- Effort
- 5 minutes to clone and run if you ever need it
- Cost
- $0 — open source
- Stack
- macOS + findphone CLI (+ Claude for your own version of the trick)
- Confidence
- High
- Posted by
- Anik Singal (Lurn) — named because the receipts are federal record
We test hype for free. We build the real thing for a living.
Thirty minutes, no pitch — and you'll leave with something useful either way.
Book a call with Todd or start with the free Business Checkup →The Verdict Weekly
Three verdicts every Friday. Free forever, unsubscribe anytime, no spam — that would be ironic.
© Schreier Group · schreiergroup.com · See a wild AI claim? Drop it here and we'll test it.