4 ms·

>I spent all day every day at CMU either programming, doing mathematics, or thinking about one of those things (to great effect: often the trick to prove a theo

>I spent all day every day at CMU either programming, doing mathematics, or thinking about one of those things (to great effect: often the trick to prove a theorem would pop into my head while showering or while taking a walk). I would fall asleep while programming in the middle of the night, dream about programming, then wake up and continue programming just where I left off. This is such a great experience. I wish I could study at CMU on-site and experience all this. I'm an old horse still kicking :P >I've since reformed my ambitions, instead of looking to start a company or get a PhD in mathematics, I've decided that hard work is not the love of my life and instead I should focus on my hobbies while looking for a career path that can be simultaneously fulfilling but laid back. Glad that you figure it out. Guess the study burned you out :(

helen___keller 5y ago
> This is such a great experience. I wish I could study at CMU on-site and experience all this. I'm an old horse still kicking :P In truth, it was a great experience, it was just also the worst experience of my life. It's a little tough to explain but, in the end it comes down to personal expectations and mental health. I often miss the work itself. A single problem set at CMU felt more interesting, impactful, and substantial than a multiple months-long project at a corporate job. Professors typically give you starter code that's ready to start working on "the meat" of the problem - here's a C++ project that loads a 3d geometry and renders it but the key function in rendering isn't implemented. Now go write a raytracer. Now go write a typechecker. Implement "malloc". Take this efficient sequential algorithm and design a parallel version that's provably more efficient, and implement it in functional code. The big problem is that the pacing was brutal and inescapable and for many - such as myself - failure was not an option. When you don't get to sleep on sundays, wednesdays, and thursdays, every single week for a few months, all while dealing with the anxiety that maybe doing your best still isn't going to be good enough, you start to daydream that maybe you'll get hit by a car finally have an excuse to take a break.

helen___keller 5y ago

> This is such a great experience. I wish I could study at CMU on-site and experience all this. I'm an old horse still kicking :P In truth, it was a great experience, it was just also the worst experience of my life. It's a little tough to explain but, in the end it comes down to personal expectations and mental health. I often miss the work itself. A single problem set at CMU felt more interesting, impactful, and substantial than a multiple months-long project at a corporate job. Professors typically give you starter code that's ready to start working on "the meat" of the problem - here's a C++ project that loads a 3d geometry and renders it but the key function in rendering isn't implemented. Now go write a raytracer. Now go write a typechecker. Implement "malloc". Take this efficient sequential algorithm and design a parallel version that's provably more efficient, and implement it in functional code. The big problem is that the pacing was brutal and inescapable and for many - such as myself - failure was not an option. When you don't get to sleep on sundays, wednesdays, and thursdays, every single week for a few months, all while dealing with the anxiety that maybe doing your best still isn't going to be good enough, you start to daydream that maybe you'll get hit by a car finally have an excuse to take a break.