Image via Jessie Orrico, Unsplash
According to scientists the next step for smart machines is the ability to think and learn the way that humans do. Artificial intelligence has. to date been missing one important ability. That is the ability to grasp new concepts in a way that replicates human learning.
Currently, machines are capable of learning but in a very complicated way. Hundreds or thousand of examples are required to teach a machine something that even a child could learn using only one or very few examples.
After much research US scientists have managed to create an algorithm that they claim captures the unique human ability to learn. This computer model was designed based on a study involving learning unfamiliar handwritten alphabetic characters. "We aimed to reverse-engineer how people learn about these simple visual concepts, in terms of identifying the types of computations that the mind may be performing, and testing these assumptions by trying to recreate the behaviour," Dr Lake said.
What This Research Means For The Future
According to Professor Ruslan Salakhutdinov "Replicating [human learning] abilities is an exciting area of research connecting machine learning, statistics, computer vision, and cognitive science." Professor Ruslan Salakhutdinov of the University of Toronto also said he hoped this new work would help guide progress in artificial intelligence by leading to next-generation intelligent machines "that hopefully will come close to displaying human-like intelligence".