Two Computers Play Rock Paper Scissors
The demo is simple. The interesting part is the strategy question underneath it: are the machines just randomizing, or are they actually learning how opponents behave?
The Direct Answer
When two computers play Rock Paper Scissors, the demo usually proves that machines can exchange throws, resolve outcomes, and keep score. The more interesting question is what strategy those machines are actually using, because random play and modeled play are very different things.
What The Basic Demo Shows
Two machines connected over a network can send Rock, Paper, or Scissors choices, compare results, and continue for as many rounds as the protocol allows. That part is straightforward. The rules are trivial for a computer to implement. What matters is what happens after that mechanical layer is solved.
Random Machines Versus Learning Machines
A naive machine can simply generate random throws with equal probability. That is already stronger than the kind of fake randomness most humans produce on demand. But a more interesting machine does something else: it watches how opponents react to wins, losses, and repeats, then updates its own choices accordingly.
Why Human Research Matters Here
The Zhejiang research matters because it showed that humans do not respond randomly after outcomes. They repeat and shift in patterned ways. A machine opponent that tracks those response habits would be doing something much closer to actual strategic play than a machine that just samples from a uniform randomizer. If you want that paper-level version, go next to The Rock Paper Scissors Study from Zhejiang University.
How This Differs From The Tokyo Robot
The University of Tokyo robot wins by speed. It reads the human hand during formation and responds fast enough to finish with the winner. That is a timing exploit. Two computers playing over a network raise a different question: not reaction speed, but strategy selection. The interesting comparison is therefore not fairness, but intelligence and adaptation. For the speed-based case, see University of Tokyo Rock Paper Scissors Robot.
Why This Still Matters To Players
Competitive players should care because machine play exposes what humans try to hide. Fatigue, tilt, overuse of Rock, and predictable post-loss shifts all become legible to a system that never gets bored and never forgets the last hundred rounds. That is why computer play keeps mattering in both research and online practice environments.
The Useful Short Version
If someone asks what it means when two computers play Rock Paper Scissors, the clean answer is this: the demo itself proves the rules are easy to automate, but the real question is whether the machines are only randomizing or actually learning the human biases that make RPS strategically interesting.