Hacker News Reader

I think there's kind of a "Water flows downhill towards the sea" rule that anything fun eventually gets the fun sucked out in a race to the bottom.

If something is fun, then people will take a pay cut to do it, until it's either not fun or not a career. Or both.

Look at indie game dev - You can't give games away. It takes serious skill to make a game that's worth the cost of the player's time to launch it.

Funnily enough, I had the exact opposite experience: I played with computers all day long as a kid but somehow it never occured to me that I could be a programmer, I used Scream Tracker instead and wanted to be a film composer.

After my first internship in a sound production studio, it dawned on me that what I was watching was really cutthroat assembly line work where "composers" had about an hour per episode and were constantly recycling and taking shortcuts. Outside the top 1% of Hollywood composers, that life seemed to mostly suck (later on, I realized that's true for the bottom 99% of almost all art related jobs).

It was only when I took a random job to pay the bills and started automating my job away that I discovered the joy of automation with code. Was uphill from there.

I feel this way about machine learning. Back when I was a high school student around 2010-2011, I really fell in love with classical AI and classical NLP, which led me to focus on these topics in college.

By the time I graduated in 2017, deep learning had taken over, and while the results speak for themselves, I really don't have the same passion for it as I did working with tangible concepts like HMMs, probabilistic graphical models, Bayesian nonparametrics, etc.. It's tough falling out of love with your field.

I had a similar experience; after only three years into my career I was very disillusioned with the career choice I had made, because although I loved programming, everything else about the job was not matching up to expectations (although we did have cubicles).

So I made the decision to leave industry and get a PhD so I could become an academic and teach. By the time I finished that, I had seen enough of what academia was like and became very disillusioned with the prospect of continuing on that path, because although I loved research and teaching, everything else about the job did not match up to expectations (although we did have private offices). So I went back into industry.

I've had a range of different software jobs, some good and some not so much. Fortunately I've ended up now in a job that is super interesting and that I love.

However - and this is the important point I want to make - I only got this opportunity because of the work I had done during grad school (and released publicly). What I learnt is that is that in order to get the really interesting jobs, I had to choose my own projects, outside of paid employment, and spend tons of time and effort on them to build up the experience necessary to find a full time job working on them. That's something I continue to do today, with a view to making myself eligible for other interesting jobs in the future.

After a long life you will get involved directly or indirectly in many industries/areas. Summary: All suck. Why? Because human suck, we are greedy,selfish and status-oriented and for many people they use or have to use those traits at their job. Want to be happy? Sacrifice as much as income as you are comfortably with in exchange of 1) Autonomy (you take the decisions, especially if only about you). 2)Meaning. You dont have to cure cancer, but at least dont drown in red-tape and bullshit.

You can get a cubicle if you stick to old tech software development. I had an employer who had nice 6 ft tall cubes with two desks per office to store the accumulated piles of equipment we had for firmware dev. Some genius had the idea to shrink us down to smaller cubes with half height walls to "modernize". Thankfully that was shot down by someone with enough sense to realize the chaos this would cause.

You may have a rad at this post by Alex:

https://thedailywtf.com/articles/programming-sucks!-or-at-le...

If I had to choose one sentence to quote, it would be this:

_There’s a term for this type of boring software: information systems._

It matches perfectly my experience that developing software is great when the value that you are creating is more than writing the software itself.

Developing a new algorithm and finding the best way to express that in software is nice; writing a program that helps you test a new hardware module you are developing is nice; testing different layouts for your blog can be nice.

On the other hand, working as a software scribe, as in receiving detailed, non-questionable input and being tasked with converting it to a function, becomes tedious pretty fast. Your first CRUD web application may be some fun, but the third will be a déjà-vu.

Sorry to be blunt, but the software development industry can be fine only is your master skill is something other than software development.

I've yet to be on a software team where there wasn't at least one dev that always wanted to make video games.

This is me - I took up computer science wanting to make computer games, but also didn't want to commit to a single career path.

Turns out computer games jobs are quite few and far between, if you don't live in a big city so I ended up in the more common web dev role instead. Less fulfilling, but also probably less stressful and more stable (and better paid).

Funny enough I played video games, wanted to write video games, but got tired of video games before graduating.

I graduated college shortly before you started and the landscape wasn't all that different when I graduated than when you started. Yes, a good part of any job is just data/crud plumbing but there are interesting and challenging problems to solve even if you're not at a FAANG. Sometimes you got to find and fix them yourself. The one thing that I love now that I'd didn't have early in my career is the tooling! Just having the right programming and debug options is huge and I feel like I was in the Dark Ages in the early to mid aughts compared to now. As for the cubicle/office change, that might be moot now with move to WFH.

> Gone is the dream of going to work and solving novel problems, instead most of what I do is data plumbing.

This was always true. I think the quicksort paper estimates that half of all CPU time is spent sorting.

Do different programming? I never took a data-plumbing job in my life.