Google’s AI Robot Faces Pro Table Tennis Players, Wins 100% Of Beginner-Level Matches –

Are robots developed with artificial intelligence capable of competing against professional athletes? To address the topic, Google’s AI company, DeepMind recently set up a table tennis competition between their robot and 29 players of different skill levels. The influence of AI has been evident in many sectors in recent times. Technology has already raised eyebrows by doing many things beyond our imagination. Despite the notable progress, AI-created robots are yet to demonstrate human-level performance in terms of accuracy, speed and adaptability.

As these qualities can only be assessed in sports, Google DeepMind decided on the table tennis competition. To assess the proficiency of their AI robot perfectly, a professional instructor was assigned to set the levels of its opponents – beginner, intermediate, advanced and advanced-plus.

The matches were played following the standard rules of the game. Some modifications, however, had to be made as the robot was unable to serve the ball physically. After all the tests were over, it was seen that the AI robot recorded a win percentage of 45 percent in all matches while emerging victorious in 46 percent of games, as per a statement by Google DeepMind.

Based on the skill levels, the AI robot won all matches against beginners and had a win percentage of 55 percent against mid-level opponents. It, however, failed to notch a single victory against players of advanced and advanced-plus levels.

Barney J. Reed, a Professional Table Tennis Coach, was present there during the test. About the performance of the AI-developed machine, he said, “Truly awesome to watch the robot play players of all levels and styles. Going in our aim was to have the robot be at an intermediate level. Amazingly it did just that, all the hard work paid off.”

“I feel the robot exceeded even my expectations. It was a true honour and pleasure to be a part of this research. I have learned so much and am very thankful for everyone I had the pleasure of working with on this,” Reed added.

According to DeepMind, it is the first instance where a robot has played a full competitive game against previously unseen humans. Their AI agent featured a high-level controller that could promptly select the most effective skill. It also had several low-level skills which were specialised in some basic aspects of table tennis like forehand serve, backhand targeting and forehand topspin.