It is officially 2026. I began planning this entry in 2025, but call with family stretched longer than expected – reminding me, once again, that I cannot multitask the way I think I can. Even so, before writing a dramatic year-end post, I wanted to reflect on 2025 in a verbose and raw form, so that I can read it later and not forget the details.
If I had to describe 2025 in one word, it would be overwhelming. The interpretation of that word changed every day, but its intensity remained constant.
The First Quarter 🔗
The year began with my return on January 2nd. I went straight from the airport to the lab and immediately started working on evaluation tasks for our paper. That phase exposed the rough edges of the submission. The paper focused on designing a new language, and my incomplete grasp of some theoretical aspects made identifying missing features and bugs extremely difficult. Debugging – and even knowing what to debug – felt overwhelming.
Still, I pushed through and completed two example case studies, which felt like a genuine achievement at the time. When the results came out, rejection seemed likely, though we stayed cautiously optimistic until the final decision. Once it arrived, we regrouped quickly and resubmitted to another venue within a week. That version felt significantly stronger. By then, nearly a quarter of the year had already passed.
On the personal side, I had already broken several promises to friends – travel plans and visits from Singapore. That was when I realized something uncomfortable: I was no longer living like an adult balancing commitments; I was simply reacting to circumstances.
The Second Quarter 🔗
After the resubmission, I hoped to finally work on things I genuinely enjoyed. Earlier in the year, I had been curious about eBPF, but without sufficient foundations, my reading felt scattered and unproductive. After discussing this with my advisor, I was encouraged to implement the Capstone processor first, as a way to build a strong base. I trusted that advice and committed fully.
By May 10th, I was close to finishing – but I had to return to India for my sister’s examination. I left knowing the debugging wasn’t complete and that it was mentally consuming me. When I returned, my routine narrowed drastically: work out in the morning, debug all day, go home, sleep.
Debugging dragged on endlessly. My peak physical condition – six days a week in the gym – slowly disappeared. Twelve- to fourteen-hour days turned into sixteen-hour marathons, often ending with overnight stays in the lab. This wasn’t hustle. I wasn’t chasing ambition. I was addicted to the feeling of closure. And debugging RTL is far more mentally draining than debugging software.
The Third Quarter 🔗
By early Q3, exhaustion had set in. Around this time, I took a short break to visit my grandparents and my university. Those few days – especially conversations with a close friend – gave me much-needed clarity.
I realized how fortunate I was to be working on something I genuinely liked. A clearer PhD direction began to form: I wanted to make the lives of other RTL designers easier. Through our work on Anvil, I had already gained a reasonable grasp of compiler design – but none of this could truly begin until I found closure with the Capstone processor.
Before that break, our March submission unexpectedly received a revision decision, compressing months of work into a few weeks. When I returned, I worked more systematically at first – planning before execution – which helped. But toward the end, I slipped back into familiar patterns: endless debugging and chasing closure. I worked through fever and exhaustion, largely as a consequence of procrastinated work.
Eventually, we submitted just before the deadline. Around the same time, the Capstone processor finally reached closure when a colleague identified the last bug.
That bug was painfully simple compared to the time spent chasing it. More than anything, that experience confirmed an instinct I had been developing: verification – despite being extremely difficult for me – is what I should pursue. There’s a saying about doing what you’re bad at; I don’t remember it exactly, but the idea stayed with me.
The Fourth Quarter 🔗
The last quarter of the year began with a search for a concrete problem statement. My routine had deteriorated, my physique had declined, and I knew the learning curve ahead was steep. I tried to reset everything at once.
For weeks, I burned mental energy producing weak ideas for weak problems while juggling TA duties, coursework, and research. Initially, being fully occupied felt satisfying. But it quickly became overwhelming.
I was a TA for a class of 150 students and had enthusiastically redesigned the assignments, hoping to teach actively during TA sessions. In hindsight, I underestimated both the difficulty and the ambiguity of the assignment. The result was weeks of nonstop student emails, forum posts, and multiple extra Zoom sessions. Much of the confusion was my responsibility, though some of it stemmed from students’ lack of background.
Balancing empathy with accountability was difficult – especially when I was already stretched thin. I soon realized my TA reviews were going to be bad. And they were (they came back a few days ago). That experience deserves a blog post of its own, but for now I’ll just say this: I learned a lot about teaching and the importance of precise specifications – useful lessons for someone interested in programming languages.
Meanwhile, coursework and research stagnated. The overload eventually manifested physically. I disrupted my routine completely – late nights, minimal sleep, excessive coffee – and developed severe reflux, dry mouth, and panic symptoms that culminated in an ER visit.
What followed was a forced pause due to phagophobia (fear of swallowing). I couldn’t eat solid food for a week or ten days, which drained my energy further. Ironically, once I diagnosed the issue using ChatGPT, I was able to recover relatively quickly.
Recovery didn’t come from anything dramatic, but from fixing basics: sleeping properly, quitting coffee, and stabilizing my routine. Within a week, I could eat normally again. By mid-quarter, my health improved. TA duties wrapped up, coursework eased, and my ideas – while still imperfect – slowly became less chaotic as I kept learning.
I committed to another research deadline, only to realize two nights before submission that my core motivating example was flawed. I aborted the submission – my first missed deadline. That failure weighed heavily. I had neglected family, missed celebrations, struggled academically, and felt stretched too thin.
Some good news followed. My QE went better than expected, and our earlier paper was accepted. I was grateful – but the happiness was muted, perhaps because exhaustion had become my baseline.
I ended the year by visiting home and taking real rest – mostly watching movies and doing very little. It wasn’t a perfect break, but it gave me the motivation to start 2026 afresh, seeing my family happy and healthy (at least by the end of the year).
That, in verbose detail, is how 2025 unfolded for me. Especially in the last quarter, I felt overwhelmed almost constantly. But I also learned a lot about myself – my limits, my tendencies, and my internal demons. And the good thing is: when there are problems to fix, life is never boring.
This is how I close 2025: humbled by colleagues; grateful for the support that allowed me to explore a new direction; grateful for my health; and grateful for tangible progress in my education, career, and learning. I’m thankful that I could visit my family, attend a close friend group’s convocation, and reconnect after a year. I’m equally grateful for the joyful gatherings at my workplace that made life more comfortable. Most of all, I’m grateful that my family is well at the end of the year.
For 2026, I hope to regain my physique, manage stress and time better, and think more systematically. But above all, I hope to talk to more people than I did this year. Life is deeply non-deterministic, and people should never be left behind.
Signing off,
Aditya Ranjan Jha