Sarah T. Vu


Cognimates is an open-source platform where parents and children participate in creative programming activities where they learn how to build games, program robots, and train their own machine learning models. Some of the activities are mediated by embodied intelligent agent, which help learners scaffold learning and better collaborate.

This platform originated as part of a research project at the MIT Media Lab to explore how children think about AI, since children today are growing up with home assistants and machine learning algorithms shaping their lives. Including the capabilities to train and use text and vision machine learning models, this open source project has grown to include compatibility with embodied agents and hardware like Anki Cozmo, Alexa, microbits, and Philips hue lights. Alongside this platform, we have developed projects and curriculum to help children learn how to use our drag and drop blocks and compatible hardware.

Our goal is to democratize AI and allow more people to create with it. We hope our platform will make creating with AI easier, and encourage more people to explore a topic that is generally viewed as difficult to learn or build with. The main focus of our work for now will be to implement a translation process for the platform so we can make the AI extension and the Teach AI playground available in other languages.

Official website

Sarah is an undergraduate student at MIT majoring in Computer Science. She is interested in Computer Science and Education, and more specifically, how the intersection of the two fields can help democratize knowledge and help alleviate the opportunity gap in America. In addition to her work on Cognimates, she is a swim coach, mentor, and Director of Curriculum at Amphibious Achievement, a student-led organization that serves over 70 local, under-resourced high school students in the greater Boston area.