This Unfair Rock, Paper, Scissors Robot Wins 100 Percent of the Time
A team at the University of Tokyo have ensured we'll never win again.
In what will surely ruin many innocent games at recess, a group of Japanese students at the University of Tokyo have developed a robot that will beat human opponents at rock, paper, scissors 100 percent of the time. In hindsight, it’s amazing that humans stuck around so long after AI mastered chess and Jeopardy.
The “Janken robot” is the third version created by a team at the school’s Ishikawa Watanabe Laboratory. The first incarnation of the bot had a rock, paper, scissors reaction speed of about 20 milliseconds after the human opponent threw up their sign, while the second could defeat human opponents nearly instantaneously.
To ensure maximum human defeat, the third version incorporates a camera called the 1ms Auto Pan-Tilt. The tech was first developed by the University for camera operators shooting sporting events to capture perfect slow motion replay by keeping a focal point in the center of the image at all times. The third Janken robot also uses another piece of University of Tokyo tech called the Lumipen 2, a high-speed projection system that can map images on targeted objects.
Diagram showing how the robot plays rock-paper-scissors.
A combination of the two makes sure the Janken can track a human hand playing rock, paper, scissors in high speed wherever it moves to ensure a winning hand in the game.
While innocent enough (we’ll be really impressed when they come up with a robot that can beat us at hopscotch or double Dutch), an explanation on the team’s site foresees a bright future for the fast-reflexed robots (all sic here):
Considering from another point of view, locating factories oversea has been advantageous in labor-intensive process that requires human’s eyes and hands because it is difficult to make the process automatic, or it is not worth the cost. However, by realizing faster process than human’s working speed, the productivity can be improved in regards to cost. Currently, although the cost-cutting of the robot is difficult, it is possible to change the location condition of the factory fundamentally by increasing the speed of the robot including visual function.
The connection between how a robot that can beat us lowly humans in rock, paper, scissors parlays into replacing us in the workforce doesn’t make complete sense at the moment, but neither does a rock, paper, scissors robot with a 100 percent success rate. The only scarier robot to challenge is the one that wins every game of mercy.
