The researchers at Google-owned DeepMind have used an AI technique that unlocked several unexplained features of the human brain. A recent development in computer science related to reinforcement learning can be applied to understand how the brain's dopamine system works.
This research was published in the journal "Nature" which aims to understand mental health, learning and motivation disorders. This found evidence called "distribution reinforcement learning" used in AI algorithm mimics the dopamine reward system in the brain.
The probability of future rewards for the brain to use and distribute is possible with this technique, rather than immediate rewards that are action-focused. Researchers explained the discovery, "We found that dopamine neurons in the brain were each tuned to different levels of pessimism or optimism. If they were a choir, they wouldn’t all be singing the same note, but harmonizing – each with a consistent vocal register, like bass and soprano singers.
"In artificial reinforcement learning systems, this diverse tuning creates a richer training signal that greatly speeds learning in neural networks, and we speculate that the brain might use it for the same reason. The existence of distributional reinforcement learning in the brain has interesting implications for both AI and neuroscience,” the researchers conclude.
“We hope that asking and answering these questions will stimulate progress in neuroscience that will feedback to benefit AI research, completing the virtuous circle.” The AI research can be used to create algorithm and machines capable of replicating the human brain. Google's DeepMind previously created neural networks inspired from biology capable of mastering Atari computer games to a superhuman level.
DeepMind also said that the latest AI research findings can benefit neuroscience.