MIT Class Teaches Students to Build Kinder AI Chatbots

What if the chatbots competing for our attention could actually help us become better versions of ourselves instead of just stealing our time?

That's the question driving a new MIT class where students are learning to build AI with a conscience. The course, called Humane User Experience Design, teaches computer science majors to create chatbots that serve as social guides rather than digital distractions.

Professors Arvind Satyanarayan and Graham Jones created the innovative class after becoming friends while co-advising a student years ago. One's a computer scientist, the other's an anthropologist, and together they're teaching students something rarely seen in tech education: how to put humanity first.

The class uses methods from linguistic anthropology to help programmers understand what people actually need from their interactions with technology. Students learn proper interview techniques to build genuine rapport with users, not just extract data from them.

The approach is already paying off in the real world. One student missed a week of class for a trial internship at a chatbot startup and returned with exciting news: the work was exactly what he was learning in class. He got the job.

Student projects show what's possible when technology meets empathy. One team designed a chatbot using 10 colorful birds to help young people engage with credible news sources in a fun way. Another group created Project Reflecto to help users become more socially confident through guided conversations.

The Ripple Effect

The class is solving a problem that extends far beyond one campus. Young adults today grow up in what experts call the attention economy, where social media and AI compete constantly for their focus. This often leads to unhealthy relationships with digital platforms that leave people feeling drained rather than enriched.

By training future tech designers to consider human needs from the start, MIT is helping reshape how the next generation of AI tools will work. The class also breaks down academic silos in a practical way: computer science students can fulfill their humanities requirement while learning skills directly applicable to their careers.

Jones points out that human-computer interaction borrowed research methods from anthropology decades ago, but those techniques got watered down over time. This class restores the depth and human connection that was lost. It proves that technical excellence and human values don't have to compete.

The MIT Morningside Academy for Design funded the course creation and continues accepting applications for similar innovative classes that cross traditional department boundaries. The future of AI might just depend on more collaborations like this one, where scientists and humanists work together to build technology that truly serves people.