6 months in, what’s up ?

Am I humble bragging about the amount of tech books I have read in the first half of 2025 ? Certainly not. This picture is more a monument to my failure than it is a sign of success. But let us get to it a bit later.

First of all, I would like to say how excited I am in writing this blog post ! For the first time in my life (apart from hitting the gym), I have been able to stick to a habit for atleast half a year.

Granted, it is messy, my notes and summaries were all over the place, but it still feels like a great achievement to be able to have some consistent output. The modern life allows us to get away with a lot of indiscipline and chaos and this blog/website is an attempt to help bring order and discipline to my intellectual pursuits.

Now let’s get back to the picture. You see a bunch of big fat books on programming: ranging from Rust and locks to distributed systems. Notice any pattern ?

NONE

And guess what ? That is exactly what I wanted to draw your attention to !

I read “Rust Atomics and Locks” more or less cover to cover. I even bravely tried to implement the nsync library in Rust, but gave up 2 weeks in, tired of fighting the Rust type system (damn the type system got some good hands).

I switched tracks to CUDA and C, attempting to write Kernels (which I did though) using Tensara, but nothing beats having an NVIDIA GPU in your hands on a local machine and being able to debug and trace the execution. Tired again in a few weeks, I stopped after the chapter on Prefix sums, before picking up from Chapter 15 on parallel-izing Graph algorithms.

Before that I was grokking through the Data Structures and Algorithms books, learning about sketch data Structures and K-D trees (sadly, you dont get asked about them much, or pretty much never during programming interviews).

Last month I was thumbing through the Kubernetes Up and Running book to get myself upto speed working on Kubernetes and fixing bugs for customers. How about Understanding Software Dynamics ? Well nothing beyond the section on benchmarking Cache and Memory bandwidth was exciting or interesting (not that performance debugging/benchmarking wasn’t interesting but the book as such wasn’t very captivating for me) Patterns of Distributed Systems ? Got the book, read a few pages, forgot, planning to someday restart reading through it.

All of this, in a span of less than 6 months. Now extend this to the a period lasting the last 10-15 years. Welcome to Govind’s brain and the sheer chaos that it is 24x7.

Why am I writing this post ? Because tracking my reading, work and experiments over the last 6 months has finally given me the proof how despite reading a lot of material I never make it beyond the Expert Beginner stage. It is simply not possible for someone to learn all of this AT ONCE, WITHIN A SPAN OF 6 months to a year. You can argue that such people exist in life and maybe you are right, but that doesn’t mean that YOU or I can achieve it (well I certainly can’t).

Like a dog at a feast that tries to take a bit at everything, I took a bite of every snack/meal, getting to know the taste, but never getting any solid nutrition. The gains are too meagre, the knowledge too fleeting, leaving you perpetually unsatisfied with the depth of your knowledge. And it shows, when you try to prep your resume for an interview or when you get asked about your experience at an interview. You perhaps manage to weasel your way through but deep down you know that your knowledge is vacuous and what you learnt was just surface level.

So I have decided to do something brave: Drop almost everything I am learning. Everything except what I am working on at work at the moment: Platform and Cloud Engineering on Kubernetes.

You see at some point in your adult life, the bravest thing you can do is to admit that you are over-stretched and you need help. As a 30s something person with a family, life is already too hard to deal with everything by oneself and additional macho tasks like learning everything from scratch at once simply makes life insufferable and ruins the life part of work/life balance.

My pet hypothesis based on anecdotal evidence and from observing talented people is that very few people have succeeded working on a side-project that is drastically different from their 9-5 work (again, counterexamples a plenty for there are literally > 8 billion people on this planet). A lot of successful people find something interesting as a young person during studies, research or at work, tinker with the same post work hours and then succeed at building a great product or a service around it. Let’s say you are learning AI from scratch while your dayjob is that of a Frontend engineer. You get maybe 3-4hrs a day (if you are in your 20s and happily unmarried that is) after work to pursue your learning while allocating enough time for sleep and social activities. Over a week that translates to maybe 30hrs. But what if you are working on AI as a FT job and then continue tinkering with it post work ? you get 8 + 3/4 hrs = 60 hrs a week excluding the Weekend, or 80hrs or more if you grind through the weekends as well. With some decent planning, the results simply compound over the days and months and there is simply no beating the results with a part time side hustle.

This is not to say that there is no hope or that trying out drastically different things is a bad idea and must be avoided. Nuance is the key and in this case, the end decision lies on you and your priorities in life. Wanna go all in ? Go ahead. Want to pursue something at the cost of a social life and sleep ? Go ahead. You are the boss. YOU MUST PRIORITIZE

Given time and social constraints, at the moment I simply cannot afford the time to work on a side hustle totally different than my main line of work. I must hence choose my battles wisely and for now I have decided to build depth in whatever I am good at and working on full-time. It does feel a bit anachronistic though, that in the age of AI when one is learning to make machines think and do everything a human can, here I am focusing on trying to do everything by myself, whereas I could be putting in the time and money into making the machine do it for me instead.

You might be right. I certainly have not forgotten or ignored the risk of AI making my job redundant, however I have 2 caveats to add to that line of thought:

1. No matter how much AI is hyped up at the moment, it simply isn’t there yet. AI coding solutions fail as more and more context is added and by the time you add daily Sprint Stand-ups, Quarterly update meetings and customers screaming at you on Slack, AI is often no better than an average engineer.
2. I often think of AI as more of a tool and less as an intelligence. When Engineering Drawings moved from being drafted on pieces of papers to computers, the Mechanical Engineers picked up computers, it wasn’t the software engineers picking up the mechanical stuff. In a similar fashion as AI improves I will be able to use it better and better to improve the solutions I work on, but my domain knowledge of systems and computers will still be highly useful and effective.

So, in summary, I spent 6 months observing myself and what I do and how I ended up like a hamster inside a wheel, always running but never moving. For the next 6 months, I shall therefore choose to undertake a new experiment: chisel my way through a singular path with single-minded focus with confidence that things will pan out in a positive way.

Conviction and Persistence determines success more often than Raw talent or skill.

As the old saying goes “Fortune favors the brave” or put more beautifully in Sanskrit : वीरभोग्या वसुंधरा (The Brave shall inherit the Earth)