Massachusetts Institute of Technology (MIT) researchers recently reported that consistently writing essays with the help of a large language model (LLM) reduced the levels of neural, linguistic and behavioral output compared to those who used a search engine or only their brain.

Concerned about what this could mean for the future of AI and its role in learning, Mateo Morales Gómez ’29 wanted to find an alternative way for students to use AI while still actively engaging in the learning process. His solution was Koan AI.

“The use of artificial intelligence, or generative artificial intelligence, like Claude, Gemini or ChatGPT, affects the way that we understand and learn.… [Koan] is a startup focused on solving the problem of cognitive deficiency [with AI],” Morales Gómez said.

Morales Gómez is not alone in this mission. He works with a team of college students around the world to make AI tools that prioritize education, each serving a different role in their enterprise. Morales Gómez is the Chief Marketing Officer of Koan, responsible for the operation’s strategy as well as developing connections with investors and customers.

The team met at Equitech Futures at Oxford University, a technology and ethics-focused leadership program for young adults, and came up with their idea for a learning management system with a personalized AI assistant called Aidan.

“Aidan is there 24 hours, seven days of the week, and he will help you with all of your tasks. But instead of solving your math problem or writing the essay for you, he uses the Socratic method,” Morales Gómez said.

Rather than giving a student direct answers to queries, Aidan guides the user towards a solution using questions and reflective thought processes while adapting to the individual needs of the learner.

The name of the startup comes from the Zen Buddhist word “koan,” a paradoxical anecdote, question or story used to push students beyond logical or analytical thought.

“Koan means teaching and learning with purpose. I think that encapsulates the essence of our work really well,” Morales Gómez said.

Since the startup’s founding in June, the team has secured $10,000 in funding from Oxford and made significant progress on Koan’s development.

“I think that we’re doing really good work so far because we have been around for less than a year … and all of the people that we already talked to—professors, students, staff—find Koan useful,” Morales Gómez said. “I’m just excited to see what the feedback will be.”

Koan is not Morales Gómez’s first experience in an entrepreneurial position. In addition to founding Koan, he also created a nonprofit organization called Impetu. Morales Gómez began working on Impetu with a group of friends when he was 15 years old and has developed connections with many high schools in Ecuador over the past few years.

“We create sustainable development projects focused on young people. Instead of creating a project and just giving it to the people who need it, we engage young people in being part of these projects or proposing their own ideas to solve the issues that they are living with,” Morales Gómez said.

Through these previous collaborations, Morales Gómez aims to test and implement Koan in high schools in Ecuador—where he is from—and expand student access to AI ethically.

“In Latin America, we don’t use a lot of these virtual classrooms.… [They don’t] have all of the features that Canvas offers, for example,” Morales Gómez said. “And of course, there’s no AI included in [virtual classrooms] that will be specifically helpful for the teachers there.”

Koan recently launched a pilot program in several schools in Ecuador and Chicago, and the team hopes to expand to other schools in the upcoming months.

“We are right now in the second phase of testing [in Ecuador]. Our last phase will be working on all of the feedback implemented, and then I think that we will be able to open to the public in general,” Morales Gómez said.

When reflecting on the future of Koan, Morales Gómez spoke about the importance of restructuring education as AI continues to evolve rather than dismissing its usefulness as a tool.

“Let’s be honest, we’re gonna still use generative artificial intelligence from here to the rest of our lives, probably. So we need to create new strategies to interact with these AIs ethically,” Morales Gómez said. “We need to update the way that we learn and teach.”