Home / Tech News / Sesame Workshop and IBM team up to test a new A.I.-powered teaching method

Sesame Workshop and IBM team up to test a new A.I.-powered teaching method

Can A.I. help build better educational apps for kids? That’s a question Sesame Workshop, the nonprofit organization behind the popular children’s TV program “Sesame Street” and others, aims to answer. The company has teamed up with IBM to create the first vocabulary learning app powered by IBM’s A.I., which adapts itself the child’s current reading level and vocabulary range, then continues to intelligently adjust as the child’s vocabulary skills improve.

IBM and Sesame Workshop announced last year that the two companies would work together on a line of cognitive apps, games and educational toys. This new app is the first result of that three-year partnership.

The companies have now just completed a pilot trial for the app, where it was introduced to over 150 students in Georgia’s Gwinnett County Public Schools. Located in the Atlanta metro area, Gwinnett County schools (GCPS) is one of the top urban school districts in the U.S., and the 13th largest district in the nation. It’s also a three-time finalist and two-time winner of The Broad Prize for Urban Education.

The app itself is built for preschool and Kindergarten-aged children on the IBM and Sesame Intelligent Play and Learning Platform. This IBM Cloud-based platform is designed to take advantage of A.I. platform IBM Watson’s cognitive capabilities, which are then tied to Sesame Workshop’s understanding and expertise in the field of early childhood education.

During the pilot program, students and educators tested the app in a classroom environment on tablet computers.

The app uses Sesame Street characters along with educational videos and word games to help student enhance their vocabulary. Teachers, meanwhile, can monitor their children’s vocabulary development in real-time through a secure dashboard and then adjust the lessons, pacing, and curriculum to each child’s needs.

The app also personalizes itself to individual children using adaptive assessments to determine the child’s current vocabulary range, then delivers learning experiences that focus on specific words. As the child continues to use the app, it will also adapt further to focus more heavily on those words and other areas that require additional attention.

The pilot program was the first phase of a longer term process that will examine whether or not a learning experience like this can demonstrably improve a child’s vocabulary. However, the initial findings are promising – after collecting 18,000 assessments from multiple choice questions over a two-week period, students appeared to have acquire new vocabulary as a result of the app.

For example, kindergarteners learned words like “arachnid,” “amplify,” “camouflage,” and “applause,” which are typically considered above their grade level.

Beyond that, the students seemed to understand the words’ meaning in context outside of the application. For example, when one class found a spider in the classroom, a kid shouted “arachnid!” Another teacher said children were able to notice forms of “camouflage” when studying animals and their skin patterns.

However, the app is still in early stages of development. The feedback from the Gwinnett pilot will be incorporated into a new version of the app, which will be introduced to more schools in an expanded pilot that begins this fall.

The app and others like it will also in the future be made available on the IBM Cloud for wide adoption in schools worldwide.