The Preservation Room

Random thoughts for posterity. But also because I'm not on Twitter.


Providing students with raw information and sparse examples followed by a single exam on the topic to assess performance is not teaching. That’s absurd.

Why on earth do people think grad school needs to be harder than any other part of life? Why are long hours and unhealthy habits the norm and why don’t more advisors actively try to fix the problem? The solution isn’t just more mental health resources for students. What if we could avoid the mental health crisis in the first place? Do something about it.

Hear me out. One of the greatest infrastructural whiffs of the United States was (is?) missing the opportunity for nationwide high-speed rail. Think of how many single-family vehicles travel the highway system and how interstate travel could be faster, cleaner, safer, more affordable, and more enjoyable if you could just hop on an overnight bullet train to zoom across the country.

Every phase of life can be made meaningful regardless of future return on investment.

For example, don’t view education as simply preparation for the rest of your life. Following through to a PhD is a serious commitment—maybe 30 years. What if you have a career change or never use your degree? Was that time wasted?

Instead, if you live strategically in the moment, redeeming the time, you will bring purpose to and embed value in every stage as it comes.

Why are grad students paid less and worked harder with fewer benefits than an industry job they could be at instead? Give them reasonable PTO, normal hours, and perks.

You are a whole person—more than your career, achievements, and academics. Life is more than these things, so don’t buy into the narrative that perceived failure defines you.

What is a Kobayashi Maru Neural Network?

A PhD does not qualify you to teach.

Stereotypes are shortcuts your brain tries to use to reduce uncertainty in the face of limited data. This is not unlike trying to find the best reconstruction of a signal from just a few samples. Adding constraints and assumptions can make the problem tractable—but come with a bias that should be acknowledged.

Fluency is like being able to dual boot in another language instead of getting stuck with a virtual machine that’s probably trying to run on top of Windows.

The more I talk with people and hear a wide range of opinions, the less important it seems to be dogmatic on secondary issues. And taking attention away from secondary issues drives it back to what—or Who—should be the primary focus.

It’s weird to think that some of “pressures” in life are totally unique to my culture—or even subculture. For example, in the U.S. (or at least my part of it) there seems to be an implicit expectation of moving out from your parents’ home after university, but all my international friends think it’s odd that I live in the same city as my parents but not in their home.

Written by Matthew Russell who follows Jesus, studies machine learning at the University of Kentucky, and interns at NASA. Get to know him or check out his projects on GitHub.