Before every semester, a lot of talk at MIT is about which classes everyone will be taking. In past semesters, I've considered printing out a list of the classes I will be taking to show people when they ask, instead of having to repeat the phrase "six double oh four, six double oh six, six oh four two, and fourteen oh one" for the 6,000th time. But I've never done it out of fear that I would go weeks without actually talking to anyone.

It's a fundamental law of the universe that, for the month before every semester begins, all conversations between MIT students eventually contain the question "what classes are you taking next semester?" The answer is often: "I don't know... what classes should I take next semester?"

There are a lot of ways to decide on classes: MIT Course Evaluations Underground (Course 6) Guides, and planners. But these quantitative metrics aren't the only way to judge a class, and the course listings are often full of jargon¹¹This is actually a major problem: you don't know what the jargon means until you've taken the class! I'm thinking of creating a new website for the MIT class listings, with better, student-written class descriptions. I'm not sure the best way to do it though, let me know if you're interested in helping.. Personal anecdotes can also be really helpful. Here is a list of what my favorite classes were from undergrad and, more importantly, why.

The following is a list of my favorite course 6 (EECS) classes, and my favorite non-course 6 classes. The classes are in order of when I took them: oldest to most recent.

Favorite Course 6 Classes

• 6.004 - Computation Structures, with Chris Terman. Computers have lots of layers: at the lowest layer, there are electrons flowing through wires. At the highest layers, there are massive server clouds communicating with each other all over the world to send YouTube videos to your computer. This class teaches you how each layer works, and how they connect. Put simply: this class teaches you how a computer works.
• 6.115 - Microcomputer Project Laboratory, with Steven Leeb. This is a major lab class, and the class I've put the most time into. This is also the class that's been the most rewarding. These labs are long and time consuming, but once you complete it, you feel like to can figure out how everything works. If 004 teaches you how a computer works, 115 teaches you how to be an engineer. The professor, Steve Leeb, is very entertaining at lectures. You can also make incredible final projects. The best I've seen? Katy Kem's laser guitar.
• 6.034 - Introduction to Artificial Intelligence, with Patrick Henry Winston. AI is all the rage these days, but this class is special in it's structure. There are 3 lectures per week: the first two teach the fundamentals of AI, the third is a guest lecture from a new guest each week. The guest is usually a professor (either in AI, machine learning, neuroscience, and related fields), and they explain their cutting edge research. It's fantastic because you learn about the cutting edge as you learn about the basics. It's gotten a lot of recognition, 6.034 was named one of the 5 best CS classes in the country.
• 6.UAT - Oral Communication in EECS, with Tony Eng. This class teaches how to communicate, explain, and converse about technical content. Disclaimer: I TA this class. I enjoyed it so much that I will be TAing it a third time this Fall, meaning this will be my fourth time through the class. The set up is great: weekly lectures, with twice-a-week recitations. The recitations are 8 students, one instructor, and one TA (me!). That's the best student ratio of almost any course 6 recitation.
• 6.803 - The Human Intelligence Enterprise, with Patrick Henry Winston. This is a discussion based seminar class that covers the most important papers written in AI. As Professor Winston says on the last class, this is "grad school in a box". Some friends of mine might disagree with this listing: some say this class is too old-school. But it's created literally dozens of hours of debate about AI between us, and that itself is worthwhile enough of making the list.
• 6.S064 / 6.002x - Circuits and Electronics, with Agarwal/Sussman/Lam. I've written about this class extensively: it's MIT's first entirely-online-class. Check out my review of it and suggestions for the future.
• 6.945 - Large Scale Symbolic Systems, with Gerry Sussman. This is class is only for the brave: there is a lot of programming: all in Scheme. But this class is great because it's an "inspiration" class, as Sussman would say. In lectures, he talks about big ideas and rarely discussed concepts... almost at a philosophical level.

Favorite HASS / Elective Classes

• MAS.110 - Fundamentals of Computational Media Design, with Professor V. Michael Bove, Jr. This is the humanities class that pairs with the MAS Freshman Media Lab Program. It's nearly impossible to enroll in this class without the MAS Freshman Program, but both the class and the freshman program are fantastic. It's a great introduction to the Media Lab, which is one of the most interesting and exciting places on campus.
• MAS.S66 - Indistinguishable from ... Magic as Interface, Technology, and Tradition, with Dan Novy (Novysan). MAS.S66 is used a generic "special number" class... so it gets reused every semester with different classes. But "Indistinguishable from" was a truly unqiue class. I can't imagine this class existing outside of the Media Lab. Novysan and Greg Borenstein made this class fantastic. In it's description, it's listed as "focusing on the exploration of Magic as a prompt to create real world solutions to real world challenges." and that's pretty accurate. The assignments involved creating technology based magic tricks, and the class used magic as a lens to look at the world. If someone asked me for an example of "what makes MIT different", this class would be one of my first examples.
• CMS.350 - 21st Century Journalism, with Seth Mnookin. This is a fun and interesting class investigating and creating modern journalism. But the best part of this class is the instructor, Seth Mnookin, who keeps the seminar-based classes engaging, but also actually cares about his students. He was extremely engaging and helpful both inside and outside the classroom. It's a trend of CMS in general: the instructors care. Other departments could look to the CMS faculty for a great example of engaging and supporting the students.
• MAS.S66 - Human-Machine Symbiosis, with Professor Pattie Maes. This is the class version of Fluid Interfaces, the group I UROPed with early in college. I'll keep it short: this class had the most interesting and forward-thinking subject material of any class I've taken (even compared to the AI classes!).
• CMS.614 - Network Cultures, with Chris Peterson. An obvious trend of this list? The instructors really make-or-break the class: and Network Cultures is no exception. I was in the special circumstance of taking this class during the semester of 2016 election. The discussions surrounding the election and the outcome were handled better in this class than any other class I was in, by light-years. Chris Peterson was also the most aware and connected instructor I've had. If you're tired of out-of-touch instructors that don't connect the material to the rest of the world... take a class from Chris.

Phew! Ok, that's my list. I glanced back at the Facebook "Note" I wrote in 2011 titled "Best Class in FM" and it holds up pretty well. I have a feeling in 6 years this one will too.