When I told people my intentions to go to grad school, the reactions fell firmly in 2 camps - "Don't wait too long, you won't want to do it when you're older" or "Why do you want to do this now, it's pretty pointless at this stage of your career".

In a sense, it was true. The degree itself wouldn't be "that much" of a value add. I've been fortunate to work at really cool places (Uber, Google) where I've never felt that the lack of a graduate degree is an impedance on my growth as an engineer.

So then why?

There are a bunch of contributing reasons here, but I think the biggest one is the raw itch to return to a habit of "deliberate deep study". This goes beyond new things I would passively pick up in day to day work.

The type of learning at work, or in general "learning by doing" is usually top-down learning. You have a higher level goal, and try to break that into subgoals and find the best ways of accomplishing those.

I've empirically seen that while top down learning gives me a superior starting point - it works best when I augment it (soon) with a holistic bottom up study of a subject. Top down learning can give a "big" initial starting point of reference, then bottom up foundational study can act as "branches" from that strong starting point. Every new piece of knowledge attaches itself to the knowledge base, and slowly you develop a dense graph which represents your understanding of the subject. The bigger the starting point (/frame?), the more chances for that graph and those edges/bonds to form. Without that strong starting point, the branches have nothing stable to connect into - they just become loose floating nodes, eventually scheduled for deletion from the brain.

But at this point, I kinda don't want to leave my day job to become a "fulltime student" again either (though I did strongly consider it in the past). Instead, I wanted to find a way to incorporate a deliberate active learning habit into my current work routine. That's the only way I foresee this to be sustainable (plus, I'm really enjoying building real things at work).

So I started scouting universities, and I found Georgia Tech's online MSCS programme to be the most compelling.

  • It had a pretty good selection of courses, growing every year - especially in the domain I'm interested in (Artificial Intelligence).
  • It was remote + part time, which let me fit it around my own schedule and take as many or as few courses as I wanted to fit into my workload.
  • It is outrageously cost effective. Essentially < $10K for the entire degree (compared to > $100k at some other places...)

And the more I researched it, the more I felt that it was a programme which was designed rather than being an afterthought. Reading The Distributed Classroom by Dr. David Joyner and Dr. Charles Isbell made me feel like this was a true experiment in scaling higher education. I was excited.

But aren't there other ways?

Sure. If you're motivated enough, it's really not necessary to go to graduate school to do this. You can try to stick to a cadence of reading foundational books and new papers (tried this), or take your pick from any of the excellent university courses that are available on YouTube (tried this too). My biggest problem with all of these was that I stopped.

So in a sense, I'm sorta taking the easy way out. In the same way as one might choose to get a personal trainer for their fitness (despite basically everything you need being freely available online), I'm choosing to use the external drumbeat of grad school to avoid having to maintain intrinsic motivation of a long term study habit. Let's see how it goes. ¯⁠\⁠_⁠(⁠ツ⁠)⁠_⁠/⁠¯