Machine Learning AI Has Beat Chess, but Now It’s Close to Beating Physics-Based Sports Games as Well
Machine Learning AI Has Beat Chess, but Now It’s Close to Beating Physics-Based Sports Games as Well
By Kyle Barr
January 20, 2023
Summary
Hell, the most sophisticated AI systems have a very good chance against top players in the incredibly complicated game of Go. But, in the uber-complicated car-based soccer game of Rocket League, can an AI do a boosted 360 aerial bicycle kick power shot from the midline? Can it pinch a ball off the side ramp so precisely it sails into the goal at 145 km/h? No, at least not yet, but AI can apparently dribble like a madman.
For more than a week, players have been driven up the wall by machine learning-based AI that's been hacked into games of Rocket League.
The game is close to 8 years old, but the frenetic, high flying, high speed car soccer game is still enjoyed by close to 90,000 players across most platforms, according to ActivePlayer.io.
As videos of the bot online show, the capabilities of machine learning AI have gotten so advanced they can actively dunk on real-life players in an incredibly complicated, physics-based environment.
From the outset, it uses strange physics that make it hard for a beginning player to comprehend.
"Constructing a learning algorithm and environment that can lead to a bot as good as Next to requires intimate knowledge of distributed computing, reinforcement learning, a solid intuition about why good rocket league players are good, and the ability to accurately predict how a change in the environment will lead to a desired change in the agent."
Players are certainly worried about encountering the bot in their games.
Reference
Barr, K. (2023, January 20). Machine learning AI has beat Chess, but now it's close to beating physics-based sports games as well. Gizmodo Australia. Retrieved February 2, 2023, from https://www.gizmodo.com.au/2023/01/machine-learning-ai-has-beat-chess-but-now-its-close-to-beating-physics-based-sports-games-as-well/