Scientists, including one of Indian origin, have developed a new artificial intelligence (AI) system that can achieve the maximum possible score of 999,990 on the popular video game Pac-Man. The team from Canadian startup Maluuba used a branch of AI called reinforcement learning to play the 1980s arcade game Ms Pac-Man perfectly.
Doina Precup, an associate professor at McGill University in Canada said that is a significant achievement among AI researchers, who have been using various video games to test their systems but have found Pac-Man among the most difficult to crack.
To get the high score, the team divided the large problem of mastering Pac-Man into small pieces, which they then distributed among AI agents.
This divide-and-conquer method could have broad implications for teaching AI agents to do complex tasks that augment human capabilities. The method is similar to some theories of how the brain works, and it could have broad implications for teaching AIs to do complex tasks with limited information.
The method, which the Maluuba team calls Hybrid Reward Architecture, used more than 150 agents, each of which worked in parallel with the other agents to master Pac-Man.
For example, some agents got rewarded for successfully finding one specific pellet, while others were tasked withstaying out of the way of ghosts.
Then, the researchers created a top agent – sort of like a senior manager at a company – who took suggestions from all the agents and used them to decide where to move Pac-Man.
The top agent took into account how many agents advocated for going in a certain direction, but it also looked at the intensity with which they wanted to make that move.
For example, if 100 agents wanted to go right because that was the best path to their pellet, but three wanted to go left because there was a deadly ghost to the right, it would give more weight to the ones who had noticed the ghost and go left.
Figuring out how to win these types of video games is actually quite complex, because of the huge variety of situations you can encounter while playing the game, said Rahul Mehrotra, a program manager at Maluuba, which was aqcuired by Microsoft earlier this year.
“A lot of companies working on AI use games to build intelligent algorithms because there’s a lot of human-like intelligence capabilities that you need to beat the games,” Mehrotra said.