Skip to content
Surprising Science

The Best AI Will Think Like a Baby

Computers scientists at UC Berkeley are studying the cognitive characteristics of toddlers, hoping to give computers the same ability to learn quickly and imagine creative solutions. 

What’s the Latest Development?


Even human minds still in a developmental stage can solve problems which remain huge challenges for computers. For this reason, computer scientists at UC Berkeley want to know more about infants’ minds. Psychologists believe that toddlers make subtle calculations which in turn determine their behavior, such as choosing a kind of lollipop based on how appealing its color is. The minds of youngsters are also understood to be highly imaginative, allowing make-believe objects to substitute for actual objects in the problem solving process. 

What’s the Big Idea?

Most importantly, children’s minds are highly adept at learning. They are, in fact, the most efficient learning machines we know of. Could computers be given the tools that come naturally to children’s cognitive development, artificial intelligence should be able to learn new skills on its own, rather than rely on a programmer’s input for knowledge. In principle, computers could model children’s ability to test hypotheses, detect statistical patterns and draw conclusions while constantly adapting to changes.

Photo credit: shutterstock.com



Related
The Dr. Data Show is a new web series that breaks the mold for data science infotainment, captivating the planet with short webisodes that cover the very best of machine learning and predictive analytics.

Up Next
THE CHRISTIAN SCIENCE MONITOR reports on the fast-developing tend for adult children to continue to live with their parents well into their twenties and even thirties.  Maybe 30% of so-called “young adults” […]