MIT Course Hack Yourself Teaches Students to Test Their Own Habits

A course built on evidence, not machine learning

The course was developed by EECS Senior Lecturer Ana Bell, ESG Visiting Instructor Carter Jernigan and ESG Senior Lecturer Paola Rebusco. It is not focused on machine learning, microprocessors, cryptography or quantum computing. Its pitch is that students can use computing tools to put together a personal toolkit of more than 60 positive habits and improve their college experience.

It is offered under the Common Ground for computing and AI education, an initiative of the MIT Schwarzman College of Computing that brings multiple departments together to develop new courses that blend computing with other disciplines.

The instructors say MIT students are skeptical unless they see evidence, so the class is built around research-backed ideas. They describe treating well-being and learning almost like systems problems: gathering meaningful data, breaking a big question into manageable parts, testing an intervention and evaluating the result.

From surveys to pre-mortems

The goals fall into two groups: personal interventions students can use in their own lives, and analytical skills to check whether those habits work. Bell calls the class "an introduction to the data science pipeline and evidence-based thinking." Students learn to design surveys, conduct interviews and think about bias and measurement.

Student projects vary. Jernigan says reducing smartphone use and getting more consistent sleep come up often, while juniors and seniors tend to focus on work habits, productivity and the move into the workplace. Students have also explored relationships, sports performance, music and motivation. They develop hypotheses and an analysis plan, though they do not run large-scale studies.

Many assignments center on teamwork. One small intervention replaces default small talk with a prompt such as "Tell me something good going on in your life." Another is the pre-mortem, in which a group imagines a project has already failed, lists what could have gone wrong and works backward to prevent it. Reading assignments are also collaborative, using the shared annotation app Perusall.

According to Jernigan, no single intervention emerged as universally best, and students' answers about the most impactful strategies varied a lot.

Intention cards and student experiences

A distinctive feature is the set of "Intention cards": professionally printed, playing-card-sized collectibles. Each distills a research-backed habit into a few words and an AI-generated image. Students earn a few each week and finish the semester with more than 60, stored in a booklet.

Rebusco says students reorganize the cards, revisit old ones when they get new ones and carry the booklets around. In her words, they end up engaging in spaced repetition and recall, "the same learning principles we discuss in class." The instructors say the cards address a basic problem with experiential courses, that knowledge alone usually does not change behavior.

Junior Sarah Hopp, a 6-3 Computer Science and Engineering major, says the cards work as both a memory tool and a souvenir. Her favorite assignment, originally developed in 2001 by Laura King, had her imagine her "best possible future self" for 20 minutes and then write that self a letter.

Student Amitoj Singh, also majoring in 6-3, applied a lesson on flow state by scheduling longer, more coherent blocks of time for his goals. He is now working on "satisficing," or learning when an outcome is good enough.

Evaluating AI tools critically

The course also covers large language models, with a focus on judging the quality of both the prompt and the response. Students may use generative AI to explore ideas or analyze data, but must reflect on the process and critique what comes back. Bell says the durable skill "isn't fluency with today's AI" but knowing how to evaluate any tool that comes next.

The instructors hope students leave able to ask better questions, a lifelong skill. They say they deliberately avoided making this another machine learning course.