Since ChatGPT took the world by storm in 2022, large language models and generative artificial intelligence have upended how we learn, how we work and how we relate to each other.
As we push further into an AI-powered world, for TC alumnus Luyen Chou (M.A. ’07), an ed tech founder and investor of nearly 40 years, this new class of AI tools represents an opportunity to finally center the expertise of educators while iterating on a technology that, “at its root, learns.” As part of his commitment to human-centered ed tech solutions, in early 2025, Chou launched DeweyLearn, a multimodal, AI-powered learning platform that’s supporting instruction through expert-driven insights for teachers and learners.
Named after TC’s own John Dewey — the seminal philosopher, psychologist and educational reformer, who encouraged Chou’s grandmother to study teaching at the University of Chicago during his trip to China from 1919 to 1921 — DeweyLearn goes beyond chatbots and leverages agentic AI, autonomous systems that can execute complex tasks with minimal human supervision, to help instructors and learners master the science of education through analysis of classroom instruction. While it was initially used in K–12 classrooms, DeweyLearn can support education in all contexts. “The ability to democratize access to human expertise and mentorship at scale is something that I think is really exciting [about] DeweyLearn,” explains Chou.
The best applications of AI are the ones that continue to put humans in the loop.
Just over a year since its launch, DeweyLearn is already being leveraged across multiple disciplines. More than 2,000 culinary students are receiving targeted feedback; consultants are leveraging the tool to identify corporate competencies and measure skill attainment; clinical therapists-in-training use DeweyLearn to evaluate their application of therapeutic interventions; and clinical assessors are getting support from DeweyLearn to conduct behavioral assessments.
Driven by a desire to expand access to a high-quality education, Chou has been developing innovative ed tech solutions since 1989 in his first role at the Dalton School, his alma mater, working under the late Frank Moretti (Ph.D. ’83), who served on the TC faculty as Professor of Communications. Moretti was an early ed tech pioneer who — alongside his colleague Robbie McClintock (Ph.D. ’68), Professor Emeritus in Historical and Philosophical Foundations of Education at TC, with support from Chou — brought computers into the school and created a groundbreaking local server network that connected classrooms to Dalton’s multimedia library as well as to NYC cultural institutions.
Luyen Chou (M.A. ’07) demonstrating the art of creating human-centered AI learning applications (Photo: Paulo Basseto)
Chou credits this experience working alongside two “luminaries in the history of ed tech” as a major inspiration to pursue his own vision for education in the modern world and “accelerate the value delivery of these paradigmatic shifts in technology to learners.”
When Chou enrolled in the master’s program in educational leadership at TC’s Klingenstein Center in 2005, he was the founding associate head of the newly opened School at Columbia University, an endeavor he embarked on after more than 10 years at the helm of Learn Technologies, Inc., an ed tech company founded by Moretti. While Chou was already deeply immersed in the world of education technology, his time at the Center — led at the time by the late Pearl Rock Kane, a former Dalton teacher and a beloved mentor for the entrepreneur — provided Chou with a valuable “competitive advantage” that he uses to craft learning solutions that stand the test of time.
“People coming out of TC have an instinctive understanding of what this complex market needs and what will actually be possible,” says Chou, who also founded the Brooklyn Prospect Charter School network in 2009 with fellow Klingenstein alumnus, Daniel Rubenstein (M.Ed. ’07). Chou’s TC education and an early-adopter mindset are central aspects of his approach to “engineer solutions for better learning.”
People coming out of TC have an instinctive understanding of what this complex market needs and what will actually be possible.
The advent of powerful large language models (LLMs) like ChatGPT is another force transforming the field of education, and there is considerable debate about what role, if any, they should play in the classroom. For Chou, however, the key to unlocking the value of this new generation of AI tools is to center the expertise of educators.
“The best applications of AI are the ones that continue to put humans in the loop,” he says. “If we continue to have that human-machine relationship, we can work together to build a model of learning that is better than either humans or machines could do on their own.”
Chou’s latest solution, DeweyLearn, comes at a time when ideas about education are in flux, something he sees as an opportunity to move the field forward. “The world is beginning to think of education as a lifetime continuum rather than as something that’s delivered within the walls of particular institutions,” says Chou. “The value is in the learner’s achievement of their learning goals; it’s not in the success or failure of institutions.”