The most popular and comprehensive Open Source ECM platform
Artificial Intelligence: Using Affordances can make AI Smarter
A concept from psychology called affordances is being used to improve the capability of AI algorithms. Affordances are simply object properties that list types of actions that can be taken on an object by some some agent that is capabile of performing those actions. For example, a button can be pushed, a knob can be turned, a ball can be thrown.
Don Norman, user experience researcher, said that “when affordances are taken advantage of, the user knows what to do just by looking: no picture, label, or instruction needed.”
The idea is that robots and algorithms can be taught to do actions that can be generalized to apply in different environments. A robot that knows the basics of what it means to grasp, push and throw is able to quickly be trained how to do exactly that within a new environment. Training times have been found to drop dramatically because skills learned from one environment can be transferred to another.
Karen Hao, reported for MIT Technology Review, wrote that “in typical reinforcement learning, an agent learns through trial and error, beginning with the assumption that any action is possible. A robot learning to move from point A to point B, for example, will assume that it can move through walls or furniture until repeated failures tell it otherwise. The idea is if the robot were instead first taught its environment’s affordances, it would immediately eliminate a significant fraction of the failed trials it would have to perform. This would make its learning process more efficient and help it generalize across different environments.”
Armin Wasicek, data scientist at Avast, said that “the true power of AI is in creating new affordances. Augmenting systems with AI can increase efficiency, but adding completely new dimensions of interaction will enable AI technology leadership.”