In science and technology, experiments are meant for seeking new knowledge, validating hypotheses, and reinforcement. We start with something, test it against reality, observe where expectations meet outcomes and where they diverge, then iterate. The same logic applies to personal experiments—testing how our own systems work, what makes us tick, and what actually changes behavior versus what we merely think should change it.
Over time, I've realized I've been naturally running experiments on myself. Not in a structured, Brown University app-tracking way, but through lived iteration. Here are some of the silly and interesting ones.
The Distraction Problem: Inside-Out, Not Outside-In
We all face distractions—notifications, messages, calls. The inbound distractions are easy to solve: silence mode, focus mode, uninstall apps. I've done all of that. I uninstalled Snapchat, Instagram, LinkedIn, Twitter.
But I was still distracted.
The real problem isn't inbound. It's inside-out. Our System One sends interval signals: there must be some notification. The notification hasn't reached you, but your mind is pushing you to check. Expecting admission results, waiting for responses to internship emails, hoping someone saw your post—this expectation is the distraction.
Last year, I checked Reddit daily for admission updates. I'd refresh portals compulsively. I'd emailed 40 professors for TA positions and found myself mentally following up every day. The expectation itself was exhausting.
To cope, I borrowed a trick from a friend who window-shopped on Amazon to fall asleep. Except I started actually ordering things. Then I'd expect the delivery. Same loop.
The fix: treat platforms as broadcast, not seeking. LinkedIn became a journal—I share what I learned, not expecting responses. Stories became walls where I post what I ate, what I listened to. If someone relates, they'll reach out. But I'm not waiting.
Now I don't worry about whether people will mail or ping. If they need something, they'll message. As simple as that.
Alarms Don't Work. Intention Does.
For years, I'd set alarms and never wake up on time. Sometimes I'd stay awake past the alarm, stop it, then fall asleep. I tried a loud clock that made me do math problems. I tried Alexa. None of it worked—they just made me annoyed.
So I cancelled all alarms.
Instead, I write down what I'm going to do the next day. I message myself on WhatsApp—just a list. Before sleeping, that intention sits in my mind: tomorrow we're coding this, working on this project, this deadline exists.
Simple result: if something important exists, I wake up. If tomorrow is chill, I sleep in. My wake-up now floats between 6:00 and 7:30, naturally. No stress, no jarring sounds, no annoyance.
One bonus discovery: watching a crime thriller before bed gives me proper sleep. I don't wander mentally—I drift off during the movie and wake up fresh.
Food: When Place Changes, Patterns Emerge
In Bangalore, I ate whatever I wanted whenever I wanted. Biryani from Tanjore Cafe, idli and mutton from Suvai, late-night cravings satisfied within minutes. Indiranagar made everything too accessible.
New York became a natural experiment. I didn't consciously change—the constraints changed me. Cooking became mindful. Less rice, chicken only twice a week, lots of milk. The body found its own rhythmic way of eating.
Now mornings have a pattern: cornflakes one day, bagel the next, high-fiber bread with avocado after that. It's happening routinely, without someone telling me what to eat.
On cravings: If I buy chips, I can't stop eating them. Kettle chips with high sodium make me drowsy. Solution: stop buying them. Replace with walnuts, almonds, milk. I keep different varieties—grass-fed, skimmed, lactose-free, whole—so there's always something to enjoy, not just something to consume.
Cream cheese on high-fiber bread blocks sugar crashes. Developing love for fiber—corn, spinach, multigrain bread—came from realizing I could actually cook these things well.
New ranking: Fat, fiber, protein are rank one. Carbs and processed foods are rank two—weekly at most.
No bloating, no indigestion, no stomach issues. The place changed, the situation changed, and behavior followed.
Deadlines: Finding Your Own Pace
During master's applications, I watched people finish by December 1st. I was still finalizing colleges on Christmas Day 2024. On December 31st, I was with my friend Kripakar in Church Street while someone called asking if I'd submitted my SOP. I hadn't. My first deadline was January 7th for UC Berkeley.
But I was making progress in my own pace. Every day I'd visit a café, record my voice, transcribe it. By the end, I had 10,000+ words typed and over two hours of recordings. I'd feed everything into AI, distill it down to 3,000 words, then 1,500. Each application got finished in the final hours—sometimes thirty minutes before deadline.
From outside: rushing. From inside: natural pace completing on time.
The same happened with visa interviews, booking tickets, paying fees, course registration, homework submissions. Everything on time, nothing rushed.
The lesson: I have the tendency to finish things on time. No need to follow others' rules about starting early.
Though I do think counterfactually—what if I started early and let the work take its natural time instead of compressed bursts? That's the practice edge I'm working on. That's partly why I don't skip these dailies.
Presentations and Long-Term Betting
I've stress-tested this: commit to a talk, prepare nothing, use just 6:00 PM to midnight the night before. Still able to deliver well.
Last semester's project presentation—we made the deck at 6 AM for a 10 AM session. We had the results, the details. Just needed to assemble.
This works because of long-term betting. Since Mu Sigma, I've made countless slides, done ad-hoc analysis, quick prototyping. The pattern recognition is there. Stories live in my mind before I deliberately prepare for anything specific.
The advice: Keep writing, scribbling, reading, presenting. It tames your systems to know where to check, what to pull, how to structure. The parts are already there.
Learning Systems: Avoiding FOMO
I've tried following random online materials, GitHub repositories, courses. Never works for me. I need to build my own understanding: this is how things are, these are the resources we need, these are the origin texts.
I also can't optimize myself against others' achievements. If someone solved 500 LeetCode problems, I can't benchmark against that. My objective is different: how well do we understand this? What unique solutions can we think of? What alchemy can we create?
FOMO sometimes creeps in. Handling it requires recognizing what actually works for me versus what makes me tired or anxious. It's an egalitarian view—everyone's perception is kind of right. This works for me; that might not.
Attention Requires Engagement
Whenever I'm not interacting, I sleep. Noticed this since St. Joseph's College—first bench, alone, and if the class got boring or theoretical with no questions allowed, I'd drift off.
Same at Columbia. Same at meetups. If there's no engagement, no conversation, no involvement from my side—I sleep.
My professors were fine with it. "He's not disturbing the class. You're talking and disturbing."
My systems are wired for active participation. Passive reception puts me to sleep.
Systems I Resist
I distance myself from systems that suppress voice. WhatsApp groups where different opinions get deleted. Meetups where shallow presentations allow no real questions. Student councils that repeat flawed methods because that's what seniors did.
When suggestions aren't heard, when conversations aren't allowed, when the response is you don't know anything before I've finished speaking—I leave.
It's clear: my systems can find their own change, adjust to situations. But sometimes they resist because the environment isn't worth adapting to.
Mindset Tuning vs. Physical Nudging
Many things have two parts: tuning the mindset and nudging in the physical world.
If I had to rank them: mindset first, nudging second.
You can't physically nudge yourself into cooking every day. You can't set an alarm to make yourself care. The buy-in has to come from inside—let's do this—from genuine interest, curiosity, or reflection.
Experiments with mindset matter more than experiments with alarm clocks, subscriptions, or productivity apps. Behavior changes first; physical adjustments follow.
I'm not nullifying nudges—they help. But if I could only choose one, I'd change my mind before changing my surroundings.
The Perception Books Give
The books I read—Thinking, Fast and Slow, Talking to Strangers, The Artist's Way, A Book of Alchemy—show different angles, not prescriptions. A Book of Alchemy has 100 articles on how different people write every day. Not one theory, but many paths.
They give perception, not prescription. With perception, we build our own prescribing systems.
When someone says "books aren't useful," and you stop reading because of that—that's where you lose originality. When you see what works for you and figure out how to adjust rather than compelling yourself to follow pre-written rules—that's where authentic points of view emerge.
There's no free lunch theorem in machine learning. Same applies to habits and intuitions. It's up to us how we get things done, minimize waste, maximize utility, and keep moving forward instead of sabotaging ourselves.
As soon as your system reflects that something needs fixing—you fix it. Not because someone else does it that way, but because you've understood how you work.
That's the experiment. Always running.