To thy known self be true.

Applying to the MIT Media Lab


Wed 13 Nov 2024

These days, I get many messages asking me for advice about applying to the MIT Media Lab. I have my thoughts on this and my advice would be similar across the board, so I am sharing it here for all of you to read. That said, these are my subjective opinions based on observations from the past year and I make no claim of their universality and factfulness. Crucially, they are not an official guide to applying to the Media Lab. Make use of the advice as you please.

To start: If you need to ask me for advice on applying for 2025 in November 2024, it may already be too late.

Disclaimer: This may be more or less true depending on the group(s) you are applying to. We will get to this.

Without wanting to discourage you, it’s competitive getting into the Media Lab. It’s just that there are so many people applying alongside the other many schools they are applying to. However, many people who get into the Media Lab’s Master’s program fostered relationships with the Media Lab in advance. That’s because the Master in Media Arts in Sciences (MAS) is not a regular Master program where you take classes. More than half your time is spent on research in your specific research group. It’s not just about your quality as a student and researcher - it’s about your aptitude and your personal fit.

That creates a hierarchy in importance of what you bring to the table:
- Grades - least important. Naturally, you need to be somewhat good. Your grades could be medium/bad and your projects/experience great. Skill > grades. Grades are just a proxy for skill. They aren’t enough to impress anybody.
- Projects - important. Varies from person to person based on age, background, history, etc.
- Personal fit - very important. Difficult to assess from an application (and interview) only. Having met the professor or anybody from the group prior to the application is a huge bonus.

Its own ecosystem

One person described it to me as “a culture of incest.” In other words, the Media Lab highly favors its own people. People who have been or are there, or people who have worked with people from the Media Lab before, or people who are friends (of friends) with the Media Lab (collaborating labs, Media Lab spinoffs, etc.). 

A large portion of newly accepted MAS students come from these backgrounds:
- MIT (or Wellesley) students who did undergraduate research at the Media Lab before
- Visiting Students at the Media Lab (and often the specific groups they are applying to) or another department at MIT
- Harvard/CMU/universities that have current MAS students

Group specificity

For example, my group, the Fluid Interfaces group, has a history of past Visiting Students or Undergraduate Researchers who became MAS students. In fact, most MAS students were previously Visiting Students. And if they weren’t, they almost always knew the group beforehand and had some connection to it.

The City Science group, I heard, is known for only taking people they previously collaborated with.

It really depends on your group. As a useful rule of thumb, the older and more famous/popular your group, the more it is connection-based. I work in a fairly popular group, so here, a personal connection is important. 

You are lucky! Because two new faculty members just joined the Media Lab and started new groups (Multisensory Intelligence and Critical Matter). You can assume that they will likely pick some students who they have not worked with before.

You need projects on a personal website

This is how things work here - in the tech scene in the US in general and the MIT Media Lab in particular. You need some combination of media, arts, technology, and science. You need at least a minimal website (like the one you are currently visiting) describing your past work. Projects are everything. Try to have real photos as visuals.

To get inspiration, check out the websites of students at the Harvard Graduate School of Design or Media Lab students. They don’t all need to be fancy. Plain HTML may suffice but get your idea across.

Be technical

Well, it’s the Massachussets Institute of Technology. You should know how to code, at least a little. And it’s the Media Lab, the place of transdiciplinary innovation. You should have reached beyond the confines of whatever discipline you were previously educated in. (Many people indeed come from a combination of academic/professional disciplines.) (But I assume you knew that, since you are interested in applying.)

However, there is an engineering bias at the Media Lab (yeah, it’s still MIT). You need the skills to make some stuff (software, electronics, bio, art, design, ...), and that’s what engineering (academic or self-taught) teaches you. Almost nobody comes from a pure social science background (something like digital humanities would work). Some people come from a pure natural sciences background, but I am not sure how or little engineering they had to do before. Many people come from an engineering background (MechE, EE, CS). Artists or designers in and between are common, but they also always know some engineering.

In other words, you may be fine as an engineer without natural science/social science/design skills, but not the reverse.

Fellowship = freedom

Each group is limited in the number of students they take by two factors:
- Funding
- PI’s time

Funding is the most crucial. The Media Lab’s Master’s program is unique in that students work as Research Assistants for more than half their time. In return, they are paid a salary by MIT. It’s similar to how PhDs work in the U.S.

For most people, the money for the stipend comes from the Media Arts and Sciences program. That means it was raised by the Media Lab, or more specifically, the PI, to support students. The Media Lab’s funding model is somewhat unique in that it relies on private “Member companies” that sponsor research in exchange for early access to ideas and Intellectual Property (IP).

If the PI did not manage to get enough money for their group, they cannot sponsor another student. Basically, a student is a “cost” to them, an employee, and they are aware that a student is costing them money from their budget. That’s why fellowships are your friend. With a fellowship, you are basically “free” for your PI. You don’t cost anything (monetarily) and you bring in research and resources. It’s a sweet deal for a PI. A fellowship also means you will have more power and freedom in your research, as you are not an employee.

With a fellowship, the main thing you cost them is time. The number of open positions for a group depends on how much time the PI wants to spend that year on mentoring.

Exceptions are the rule

So does this mean there is no point in applying? Not necessarily. However, I am writing this post to potentially save you the $85 application fee.

Many applicants think the Media Lab is like any other grad school, but it’s not. The work is often much more unstructured and self-directed than what you would have at other grad school programs. Moreover, unlike most other Master programs in the US, Media Lab students only take about 5 classes in two years (instead of 4 per semester at, for example, Harvard Graduate School of Design). The rest of the time is spent on their research and Research Assistant contributions. 

As a result, the application process is perhaps also not quite like that of other graduate schools. I just feel a little uncomfortable that many people get their hopes up and pay the application fee when the system works in a certain way and they didn’t know and it was never communicated to them.

I would think twice about applying to the Media Lab unless you have one of the following: existing connection to the Media Lab (e.g., past research experience), a unique (multidisciplinary) background/skillset/interest (e.g., computational design for textiles, or a transition from film studies to digital fabrication to bio art), or a dedicated passion and strong vision for the work of a specific group.

That said, the Media Lab is always open for surprises and these are, of course, no set rules. PIs are often looking for a diverse skillset and personalities in their groups. You might still want to take your chances. 

Ultimately, I believe the lab selects people who are a good fit (although it certainly misses out on many others who also would have been a good fit). If you know that you and the Media Lab are a good fit, you probably are.