
Fancy a game? AI rock, paper, scissors
Rock, paper, scissors… with AI
A little side project to show what's possible in a web browser: hand tracking, a model that learns from you, live statistics and predictive analytics, all in plain HTML, CSS and JavaScript.
- Enable your camera and allow access when your browser asks.
- Hold up rock, paper or scissors inside the on-screen guide.
- Press Ready and keep still for the three-second countdown.
No webcam? Use the click-to-throw buttons under the camera instead.
Play full screen
A hands-free classic
Read the room.
Throw your sign.
Show your hand to the camera. The arena reads the gesture, the computer throws back, and the scoreboard keeps the receipts.
Hold your throw in the frame.
Enable your camera to unlock the first throw.
No throw captured
Your next captured frame will appear here.
Wrong read? Tune it.
Your corrections stay in this browser and become JSON training examples. Build a small personal gesture library, then export it whenever you like.
Play a few rounds to see your tendencies.
Accuracy appears after your first predicted round.
- After a win, you repeat
- —
- After a loss, you switch
- —
- Most common follow-up
- —
- Longest run, same throw
- —
What's going on behind the scenes
The same building blocks power real business tools: reading documents and images, learning from feedback, and keeping data safe.
- HTML5
- CSS3
- JavaScript
- MediaPipe Hands
- Web Audio API
- Markov chains
- Local storage
- JSON
Imagine what this could do for your business.
If a browser can read a hand gesture, it can read an invoice, a form or a dashboard. Let's talk about yours.