← The AI Hype Audit — all 494 verdicts
PARTLY
Partly true: Ray-Ban Meta glasses, a vision model and the Stockfish engine can feed someone chess moves in real time, but this clip is an edited creator stunt and its own repost says the result isn't confirmed
The claim'A guy reportedly used Ray-Ban Meta glasses, GPT, and Stockfish to secretly get winning moves during an in-person chess game.' He claims to have played 'once or twice'; the glasses stream the board over a WhatsApp call to a MacBook, a GPT model reads the board, Stockfish picks the best move, and the move is sent back as audio. The caption adds that 'the edited footage doesn't independently confirm the full result.'
The setup described is technically workable. Ray-Ban Meta glasses can place a WhatsApp video call, so a laptop can receive a live view of the board. A vision-capable model can read a chessboard position from a clear image most of the time, Stockfish is a free open-source engine far stronger than any human, and the move can be spoken back through the glasses' speakers. Hobbyists have built several versions of this, and over-the-board cheating with hidden devices is a real problem that chess organisers already screen for. So the capability is real. This particular video is a creator stunt, though. It is cut together, the opponent is a stranger on the street, the 'I've only played once or twice' line is part of the bit, and the reposting page itself says the footage doesn't confirm the result. Board reading from a moving head-mounted camera is also error-prone, and in the creator's own explanation the move is drawn over the video in the glasses' view rather than spoken, which differs from the caption. Treat it as a demonstration of a plausible trick, not proof that it works reliably.
What holds up
- Reel transcribed (faster-whisper large-v3): the creator explains calling his Mac from the glasses, reading the board with a GPT model, running Stockfish, and re-rendering the best move.
- Caption read; it credits a Korean TikTok creator and states the edited footage does not confirm the result.
- Components checked: WhatsApp calling from Ray-Ban Meta glasses, open-source Stockfish, and vision models reading board images are all real; end-to-end reliability is unverified.
What doesn’t
- Edited stunt; outcome not independently shown.
- Caption and creator differ on how the move is delivered (audio vs. overlay).
- Using this in rated or prize play is cheating and gets players banned.
The catch
The parts are real; the clip proves a stunt, not a reliable system.
How to actually do it
- If you run events, assume smart glasses can relay a camera feed and set a device policy.
- For a legitimate use, the same pipeline (camera to vision model to engine) makes a good chess-training or commentary tool.
- Don't use it in any real game.
Smart glasses plus a vision model plus Stockfish can relay chess moves; this clip is an edited stunt with an unconfirmed result.
- Confidence
- Medium
- Posted by
- an AI-news aggregator page reposting a Korean TikTok creator's video (about 140,000 plays); sent to Buddy Wed Oct 21, 2026, 7:24 AM
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.