2024 Wrapped

Written from 15th Dec '24 - 01 Jan '25

The time of turning 20/21 is a period of immense flux -- change that defies easy quantification. At the start of 2024, I had no clear roadmap for the year and refrained from making overly ambitious targets. This piece is an audit of my year: an account of how easy it is to lose sight, grow jaded and simply jump into routines that give quick dopamine hits.

On solitude and connection

3White1Black

At the very start of Jan, I wrote in my journal [^1] that I relished solitude. By circumstance, there were solitary 1km walks under the blazing sun (and Singapore humidity) that I had to make to join my batchmates at a security post. Despite my soaked inner tee amidst the oppressive heat, I enjoyed these walks -- they gave me the freedom to think my own thoughts, pray, and meditate, away from the constant company of batchmates 24/5.

Now, I could never imagine enjoying such moments. I would simply want to get from one place to another quickly, probably complaining while I'm at it. I think it was gaining some sort of leverage over my circumstances that diminished these simple joys. That meant commissioning: no longer treated as a trainee, as someone who was at the bottom of the food chain in military hierarchy.

As the year progressed, I realised solitude, while restorative, couldn't sustain me indefinitely. This became obvious in my post-mortem for running, especially when training to race a distance I'd never competed in before. I often wished I had someone to pull me through those long, early-morning tempos!

So it's not so much that I wanted to be left alone; I just needed to be surrounded by the right company. But in situations when I can't control who I'm with (pretty much as in NS), then I'd just have to adopt a positive attitude and cherish those quiet moments of solitude.

Overcommitment and Inertia

When officers commission, we have to go through parade rehearsals. On my very first parade rehearsals with drill boots (the kind that go klik-klak-klik-klak when you march), fatigue from a fitness test [^2] got the better of me. I slipped and fell on the barrel of my rifle. Dazed but conscious, I immediately stood up and wondered why so many people were staring at me and then surmised it was probably because of the rate of blood collection on the ground. I got my face stitched up in a hospital (it was a very clean, half-barrel gash on my right cheek) and was back rehearsing one day later with a plaster and some stitches on my face.

On hindsight, this was my first encounter with needing to know how to manage my energy levels. But I'd continue to make the same mistake throughout the year.

Initially, navy work didn't take up a ton of time. At the start of doing my work, there was some weird sense of accomplishment -- that I was doing something for my unit (I remember distinctively thinking that after sending out my first email, haha). I was triple checking my work -- I wanted everything to be on point, 100% correct. To save time, I rationalised that staying in (camp) made more sense to avoid the 90-minute, door-to-door commute.

My outside of work productivity peaked my first week staying in camp: I wrote one long article (broken into three parts) consolidating my thoughts on the education system in Singapore. Simultaneously, I continued development on Running Made Simple, together with Mok. March and April were largely spent doing that.

But with flexibility came the mistake of setting too many goals for myself: get a driving license, take run training more seriously (run a fast quarter marathon, then a fast half), getting good at ML, working through college apps and exploring sports science in a rigorous manner -- all these while managing a day job at camp.

I found progress on many fronts painstakingly slow. Back in Apr / May, I was barely averaging 30-40km / week. Workouts had to be in the morning and the increase in the training load left me feeling exhausted on most days. Starting my stint at Teraflop was exciting too -- but I could only give it the dregs of my energy after work and training. It was much easier to blow off some steam with a game (or two) of PUBG Mobile (oops).

The result? A sense of being spread too thin and deep frustration at myself on days that I got no work done.

On Freedom and Accountability

I would say that net-net, my posting in NS was a privileged one: I had a good boss, a great working schedule, and to top it off, I found some of the work that I did actually meaningful. I no longer had to crawl in the mud, get turned-out, suffer through long route marches or from bad sea state.

I implemented a few systems for me to remain accountable. The first was a task logger, where I would use BeFocused to log blocks of time that I was focused for, for a certain task -- following the Pomodoro technique. To add detail to my progress, I built a Telegram bot where I logged specific tasks at the end of each night. By parsing this data with regex, I could easily compare work completed week-to-week.

I like to do year on year comparisons. Unlike the restrictive training schedules in 2023, freedom in holding some power in 2024 meant free time whenever I got my work done. I used to look forward to weekends during BMT or OCS to squeeze in a bit of quick hacking, but the drive to make the most of my free time has since diminished. The tighter restrictions of 2023 also meant more rigorously quantified goals and clearer boundaries on what I could realistically achieve.

I think optimising said free time becomes much harder when there's more space to play around with. There were now various decision matrices that I had to work through daily: staying in to save time in travelling + getting more outside NS work done vs comfort at home, taking a more direct route to camp vs making a detour to hitch a ride from my boss, choosing between "resting" (aka enjoying a round of PUBG M with my friends) vs grinding through The Little Book of Deep Learning.

Status

(a personalised conclusion from my piece on Status) I'm also terribly aware of how much more I could have done with my time. I know of friends who've built and launched incredible frontend apps in NS, as well as folks who've won OAI hackathons while in NS. I'd like to think they had focus.

Some of my good friends tell me that I'm being too hard on myself -- after all, what are teenagers our age supposed to be doing anyway? There's a constant tension between freedom, accountability and ambition that I've not quite figured out yet. And I don't know whether my ambition has taken a plunge (again, because of the environment that I'm working in).

In 2023, my reaction to these feelings were more visceral: being stuck in Tekong after a month in NYC, watching friends get on with life, securing funding, etc. left me with an extreme sense of FOMO. In 2024, I've grown jaded of the constant "thrilled to announce...", "we're launching...", "new tool dropped by...". That sharp sting of FOMO has softened, but the underlying sense of comparison still lingers.

Young ambitious precocious folks will always found things, start things, and claim their share of the limelight. But when my lifestyle is heavily influenced by external circumstances way beyond my control, it pays to take a step back and think more deeply about some of the emotions and thoughts that lurk in my mind.

On one hand, I acknowledge the need for having accountability. NS is a unique experience and there's always the temptation of exaggerate its challenges or use it as an excuse for a lack of progress. On the other, I've learnt that constantly comparing myself to others only leads to frustration. Comparison is the thief of joy. In some sense, no one will really know how much I'm going through, and whether I'm telling the truth.

There are two main takeaways that guide my thinking day-to0day:

1. It's easier for me to gain traction on things that I'm interested in, rather than what others find compelling. Folks may be interested in AI / ML and building B2C SaaS ("personal AIs"). While I find them vaguely cool, they're not where I want to devote my energy -- at least not yet. Getting things rolling and gathering momentum at the start of any project is always the hardest; once feedback loops are in play, progress comes much easier. It's something I knew before, but 2024 brought this into sharper focus.
2. All comparisons should be made internally, not externally. Yet sometimes I look back at shadows of my former self [^3] when pursuing a serious undertaking in distance running. I find myself making comparisons easily to past fitness standards (objective times, often without carbon-plated shoes, etc.). And I miss how fit I was back then and the ability to keep pushing and break personal bests. And I miss winning a race. So I think it really sucks when a race like the Standard Charted Half Marathon doesn't go according to plan.

While I’ve been working on relying less on external validation, I’m still a work in progress. On bad days, when I feel unproductive (ironically, even nailing a hard workout doesn’t count), I’ve found that sleeping early is my best mood regulator. Solitary gaming—whether it’s PUBG Mobile or Civilization VI—only pulls me into a downward spiral that can be reset only by resting my mind from further dopamine hits. Resetting through rest than escapism makes much more sense.

But at the end of the day, I don’t think these struggles with status really matter. People often accumulate status through relational ties to someone influential—degrees of connection to power, fame, or wealth. I’ve even seen batchmates brag about interacting with someone on the Forbes 30 Under 30 list during reservist duties.

What if, instead of chasing status, we could rest in knowing we are friends—truly known and loved—by Someone with the highest Status?

Loading

I think loading differs from overcommitment because while overcommitment stems from what you impose on yourself, loading is driven by external circumstances and your ability to handle what’s thrown your way. Back in 2023, most of my personal growth came from sustained and intense loading from training school(s) whereas in 2024 these were a little more localised and closer to home.

In late July to mid-August, I took a break with my parents to visit Europe. We flew into Frankfurt and managed to visit four countries in about 20 days—down south to the Black Forest, east to Austria, and then back up to Frankfurt.

Returning to Singapore was a stark contrast. Within hours of landing, I dumped my things at home, took a quick shower after the long flight, and Grabbed straight back to camp for a meeting. The abrupt shift was jarring.

After clearing email backlogs and getting up to speed with Navy work, I quickly realized that September through November was going to be high-tempo. Initially, I thought I could power through by applying periods of intense focus and “ruthless execution of tasks” (prioritize and execute). But I soon hit a wall—everyone has limits. With no one under my charge to offload menial tasks, I couldn’t move things along as quickly as I’d hoped.

As time ticked by and my bandwidth stretched thin, I found myself slumping against a wall in the office, too tired to care about the ranks of people walking past. For 5–10 minutes, exhaustion took precedence over everything else.

In that moment, I was reminded of a problem I first encountered during my time at Teraflop: crashing Google Colab instances while working on large text datasets for n-gram analysis. Those crashes happened because the CPU overloaded and ran out of memory. Sure, there were ways to mitigate it—multithreading, splitting datasets into smaller chunks—but at that moment, I felt the same way. My own processing capability was maxed out.

Unlike computer systems, where parallelization was an option, most tasks in camp had to be completed sequentially. This lack of flexibility left me completely cooked—mainly from the relentless task of passing down chain-of-command messages. It seems to me that the way our brains handle stress is more like a CPU than a GPU: sequential, not parallel, and prone to overheating when overloaded.

Despite the strain, I appreciated the intensity of getting things done. But the adrenaline rush that carried me through these periods wasn’t sustainable. Once the high wore off, I was left exhausted.

This principle applied not just to work but to emotions as well. After a busy day at work and miles of running in the morning, I became irritable and snappy—a terrible person to talk to. My energy, or bandwidth, was simply spent at scale. Recognizing and managing this will be something I carry forward into 2025.So that's something for me to bear in mind for 2025.

Conclusion

I'm writing this just before 2025 starts (2235H now). I remember feeling some amount of uncertainty before 2024 began, because I had no idea what was ahead of me. I'm honestly feeling a little anxious -- the stakes are a little higher; the fever dream of auto-piloting through NS has come to a close, nothing feels as structured or straightforward anymore.

The year of 2025 will be another year of flux. Optimists couch the new year as a myriad of opportunities and having been on tech Twitter, I think I have some idea of what exists out there. But as much as I've collated all my thoughts and learning points from 2024, my mind still feels scattered and unsettled. It's both frightening and exhilarating to see what's in store for the next year.

Sometimes, even when we know a situation isn't as high stakes as it seems and that fears are irrational, it doesn't remove the weight it carries at that moment. In those moments, there are various quick fixes to make us feel better -- relativising the issue, escaping through distraction, ranting to a friend, etc. but when those methods fail, we need a permanent answer.

While we quip that change is the only constant, I'd think there's value in a certainty that grounds us during those periods of flux. In the coming year, it's my sincere hope that I can continue to find true rest in that certainty.

Things accomplished

• 10k PR (slow), HM PR (even slower)
• driving license (3 months from start of practicals)
• functional, nuclear powered aircraft carriers toured: 1
• countries visited: 5?
• times I saw Naval Ravikant irl: 1
• pieces written: 8? (word count: 20000+?)
• commits: 203
• notable events: devday, network state conf
• projects: 3? (running made simple + teraflop?? + elite endurance viz)

Footnotes

[^1]:I kept a journal on a Notion page from the start of NS until now, it has grown to 19k+ words now, and takes >10s to render on my Mac. I can't even get it to render on my phone, haha

[^2]: That fitness test involved standard IPPT, followed by max pull-ups. I was rather dazed after that to be honest but we were rushing along to make the timing for the rehearsal in a new kit. Other friends also fell at that area but none as bad as mine.

[^3]: PBs of 2.12 in the 800 and 19.42 in the 5000. Not objectively fast, but okay-ish.