Games always have been a great way to measure ones abilities. In 1997 Deep Blue beat the world chess champion, recently DeepMind developed an AI that can play StarCraft II.
Now there is the FruitPunch AI-esports competition to spark innovation and provide a fun learning environment!
Artificial Intelligence (AI) and esports might seem like a strange combination at first but actually they go hand in hand. Games have been used to test and advance AI research for a long time. In 1997 Deep Blue, an AI developed by IBM, beat the world chess champion Kasparov at chess. This was a huge achievement for the field of artificial intelligence and it did not stop there. Recently DeepMind developed an AI that can play Starcraft II, a complex real time strategy game, at human levels which is also a great achievement.
Creating AI’s that can play video games might not seem very useful, but the advancements made in creating these AI’s are also applicable to other domains. In addition to that, video games are basically simulated environments. In other words creating an AI that can control a robot in a video game is not that different from creating an AI that can control a robot arm in a simulated factory. If this simulation is good enough an AI trained in a game should even be able to control a real robot arm using only its training from within the simulation.
There are even more exciting developments in the field of AI that are relevant to esports. DeepMind is currently trying to prove that reinforcement learning, a branch of machine learning, can be used for actual neuroscience research. And the very same reinforcement learning is one of the most suitable techniques to create an AI for video games.
The FruitPunch AI esports competition is an activity where the cutting edge of machine learning gets implemented by passionate students in a fun and engaging way. Teams from different universities will compete to create an AI that solves the challenge presented best. This year the theme of the challenge will be collaboration between human and machine. In this challenge a human player will work together with its AI partner to defeat the other teams.
The total price pool for the competition is:
“I woke up from hibernation with the systems going wild. It was too late, the display read ‘brace for impact’. Everything went black. Next thing I remember Pathy pulls me out of my chamber. My heads-up is indicating ‘power too low for long range transmission’.
‘Why would we need long range… Anywhere inside the cluster system should be safe.’, I look through the small window of my quarters and through the dust & smoke I see… A MAZE
What a sick machine mind would make such a trap… Playing games… There must be more out there.
All I know is, I need to make contact with the central community fast. Pathy will help me, as long as we’re working together we can get out of this mess.”
Enrollment for next year’s competition will open after June.
You can start preparing today by looking into Unity ML-Agents, the technology we use to interface your AI with our game. You can also start doing research into flexible and goal-oriented navigation in procedurally generated dynamic environments with limited information. An interesting place to start can be a paper by DeepMind about navigating with grid-like representations in artificial agents. Good luck and have fun preparing!