College: What went wrong, what went right
Fri Sep 21, 2018 · 1226 words

In 2014 when I started college I dreamt of being a great student. I picked computer science as my major because I loved the field. I was happy, confident and excited. After all, I had straight A+s in class Xth and 90% in class 12, what could go wrong ? After 4 years of college, I have 7 backpapers with a history of at least 15 failed papers (along with a couple of research papers though) to my name. My name is Abhai Kollara and I’m an SOE ian.

I write this for two reasons, one, there are few mistakes that I want to admit to myself and somehow writing it out seems to be a good way to do it, two,

Mistakes should be documented

Not just for your future reference, but for other people’s as well. Granted, no one goes through the same life as you, but there is always something to be learned from mistakes, be it your own or someone else’s.

What I did wrong

The general learning culture in my college (like many others) is passing exams first, understanding concepts later (“later” generally means never). I should’ve picked this up but instead, I chose to go with the crowd. Especially for subjects that I didn’t enjoy, I too started “studying for exams”. I went from actually reading textbooks to skimming PPTs, cramming modules of syllabus on the day before the exam. Within a single month, I had destroyed the study pattern that I maintained for years. And I sucked at “studying for exams”. I had barely passed my internal exams (this, by the way, would become a joke by the end of 2nd year). You would expect this to be depressing, but it wasn’t. Because I accepted it. This is was my second mistake. Instead of inspecting what went wrong I somehow convinced myself that I was average and that this was the best that I could do. There are a couple of things that contributed to this; one, back in my school I was always one of the brightest students and I believed that, only to come to CUSAT and realize that there are tons of people like me who are just as good. As a result, I attributed my poor performance to my “averageness”. Secondly, failure was the norm, failing in your internals was “normal”, having a couple of backpapers every semester was “okay”.

Being resilient to failure is one thing, being insensitive to it is a whole different story.

When you put in your 100% and fail, you should feel a sense of disappointment because it’s not fair. It’s not fair that you did your part and you don’t get success (whatever that means btw). And if you still go out there and try again, you start developing a resilience. Resilience is good. But if you don’t feel any pain in your failure, it’s a strong sign that there’s something off.

It’s safe to say that even though I loved learning, academics had taken a back seat.

How the system contributed

In 2014, when I walked into my first college classroom on the A-block, I too was happy about college like everyone else. After all, I could finally learn just the things I loved, right ? No. Strike one (pun intended), our syllabus was as flexible as concrete. In our first year of college, we had to learn

It didn’t matter if your stream was CS or EC or mechanical, everyone learned the same things for the first two semesters. Now, the argument is, to be a good engineer you must know the basics of all relevant fields of engineering. But this lack of choice of curriculum was not limited to the first year, we never had any choice of what we got to study. Even in semesters where we had electives, our choices were limited due to the unavailability of teachers.

It’s a pity how mediocre institutes like MIT, Harvard and others produce below average talents by giving them control over what they wanted to learn ¯\_(ツ)_/¯

This is apart from the fact that written exams are ill-equipped to test for a student’s understanding. Going from words to understanding is easy, the reverse is pretty difficult. Words somehow don’t fully capture one’s understanding, nor are the right words the proof of understanding. And yet this is exactly what exams demand. Maybe I am generalizing a bit too much, you can design exams in a way to check for understanding. But the university exams were definitely not those.

On the other hand, assignments exist to augment exams. Considering that it is an engineering course, one would expect practical assignments. But no, we basically had homework, just like in school. The way we did/copied those, I used it as an example to explain peer-to-peer networks for an in-class seminar. On the day of the submission, we would come to class a bit earlier than usual and build the entire assignment from multiple sources (from those who actually found the answers). Thus, the source of all assignments could always be traced back to 3 or 4 original authors (God bless them !). Towards the end of the course some teachers dropped the whole idea of assignments, while some experimented with some novel ideas that could not be spoofed like seminars, group discussions and what not. Over the entire course of 4 years, we had a sum total of 2 coding assignments, which I wish I had done more sincerely.

Again, though the circumstances contributed, I was complacent enough to not give a damn.

What went right

The story is not as bad as I made it seem though. Despite the less than sincere learning that I did at college I still loved this science. I loved computer science. Plus, I loved to code. As a result, I always tried to keep up with the field by reading as well as talking to other “techies” in my class. SOE has a tech subculture in there, you just have to find it (Credits to CJ for introducing me to that). This was a couple of years back when AI and deep learning was beginning to explode to its full fury that we see now. After a few months of self-learning, I started contributing to the development of an open source deep learning project called Keras (shout out to Fariz). This helped me land my first job at datalog.ai which was an AI startup based out of silicon valley (A huge thank you to my CTO Malai). This was just before my 4th year of college began. In September 2017, I started as an AI research engineer (or “intern”, if you want to get technical) at Saama Technologies. And a year in, I have two research papers with my name on it #shamelessBragging.

Did it finally go right ?

I don’t know. I’ve taken a break from my job to make time to study. The backpapers are baggage which I need to clear out. Was it a good move to take the break ? I’ll tell you soon enough. But the fact is I could’ve had both this job and my degree if had tried better.


back · home · blog · projects · research · who am I ?