Introduction

Welcome to my blog! This is where I share my ideas and projects that I've been working on.

The Axe of the Era (01/20/2022)

The world of AI is rapidly advancing, and one of the most popular models currently is GPT-3, a 175 billion parameter language model that can generate text, code, and even images. With this powerful tool, it's no surprise that people are eager to explore the new and exciting opportunities it presents.


During my winter break from NYU, I decided to take advantage of this new frontier and began experimenting and hacking together projects that I found useful in my day-to-day work. The concept was simple - my friend Ethan and I were in the library, and he mentioned off-handedly that it would be great if robots could read the news for him. It then occurred to me that OpenAI's GPT models were likely sophisticated enough to generate news article summaries. With this idea in mind, I spent the next week building a web application that could generate news articles.


I considered a few different strategies for hosting the application, such as Elastic Beanstalk or even serverless Lambda functions. However, I also had to consider the fact that the model was 175 billion parameters and that hosting it on a free tier would be difficult. I also wanted to make sure that the application was accessible to everyone and easy to use. So, I decided to build the service on Heroku, which offers great student plans for developers. Despite the cost of API and Da-Vinci tokens, I plan to monetize the application in the future.


You can try out the app for yourself, but be warned, it's addictive! I initially built the app for my own enjoyment, but I was curious to see if others would find it as entertaining as I did. I posted the service on a few different subreddits to gauge interest, and one of them (this one) received a significant amount of traffic. I received valuable feedback on which features worked best, which didn't, and what users wanted to see next. The most common request was to convert the app into a chrome extension.


So, I turned the app into a Chrome Extension and shared the news on Reddit again. To my surprise, over 1,000 people downloaded the app in just a few days. While the number of users drained my GPT da-vinci tokens, I am taking a break from the beta to strategize on how to monetize the app using a freemium plan and handling payments through Stripe. The best part is that I now have a working HTTP client that can translate and summarize all kinds of text to and from GPT-3, and I'm excited to see what other possibilities I can explore with it in the future.

What Impact Do I Want to Have (KP Fellowship App) (01/18/2022)

Recently, I applied for the Kleiner Perkins Fellowship. As part of the application process, I was asked to reflect on my purpose and passions and answer the question: "What impact do I want to have?" After some thought, I came up with the following answer:


Let's be honest, AGI is inevitable, and it will be integrated into every facet of our lives. As we approach it, we are faced with one of the biggest challenges of our generation: making sure it works for us. Having witnessed the rise of technologies like smartphones and social media, I have seen firsthand the ways in which they have both improved and hindered our lives. As AI becomes more prevalent in our daily lives, it is crucial that we develop AGI in a way that aligns with the values and goals of humanity.

That's why I want to make an impact on the world by developing AGI applications in a responsible and ethical manner. I have been working towards this goal for some time now. This year, I developed an AI-based news summarization tool that helps over 1,000 users consume news more effectively and save time for other pursuits. Additionally, I joined an EdTech startup, Kunduz, that provides accessible online tutoring. By building proprietary AI, we helped students without access to personal tutors in countries like India and Turkey gain access to education.

These projects have been incredibly fulfilling and I believe they have provided enduring value. I am eager to continue working on these issues, and I hope to join or found a startup where I can broaden my impact even further. The development of AGI is the most important issue of our time, and I want to be a part of the solution.

A Summer of Suburban (09/08/2022)

Herndon, VA, is a quiet town an hour out from D.C. It isn't the typical place you would expect to find a 20 year olds who spend 90% of his time coding, going to the beach, and dancing on weekends. Nonetheless, if you went down to Herndon this summer, you would have found me, alongside a handful of other students busy doing just that. I can't say that we were the most productive bunch, but we were certainly the most fun.

Coming in to the internship, I was a little nervous. For one, I had little safety net if I didn't find friends or find fun in the work I was doing. Luckily, I was able to find both. I was able to find a group of friends who were just as passionate about coding as I was. We spent our days working on projects, and our nights going out to the beach, or dancing at the local clubs. I was able to find a group of friends who were just as passionate about coding as I was. We spent our days working on projects, and our nights going out to the beach, or dancing at the local clubs. I felt like I was part of the frat the NYU never offered, except instead of drinking alcohol, we were drinking coffee, and instead of getting drunk, we were getting high on the thrill of coding.

The summer taught me a lot about how to operate on my own. During my bike rides along the Potomac, I was able to reflect on what motivates me to act in ways I do. Getting away from New York for the first time in over a year, I realize how acclimated I've become to constant stimuli. Being able to take a breather was beyond needed for me and helped me reorient towards a long-term view.

How I'll Spend My 20's (03/10/2022)

Today's a big day! Not because it's my birthday, but because it's the day I turn 20. I've been looking forward to this day for a while now. I've been thinking about what I want to do with my 20's, and I've come up with a few ideas. Here's a shortlist of ways I want to spend what has been heralded as the seminal decade of many people's lives:

1. Move to San Francisco. Why San Francisco? San Francisco is the Schelling point of high openness, smart, optimistic, energetic people in the world. (Inspired by advice from Patrick Collison)
2. Learn how to do a backflip. Mostly because it's something on my buckt list, and I dont think people start doing backflips after 30.
3. Start/Join a startup. While FAANG engineers have a lot of stability, the highest aspiration in my eyes is building impact on ones' own. While that's not to say that I won't work at a FAANG at any point at all, I do believe that ultimately it is important to harness the skills you gain from that experience and channel it into real entrepreneurship. Life is to short to not do what you love.
4. Develop an area of extreme expertise. Now, this isn't to say I don't maintain expertise in a variety of different fields already. However, I will say that I have yet to reach 10,000 spent on any specific domain. During my years in college, I have taken advanced coursework in NLP, AI/ML and even multi-threaded computation. While I know enough to be dangerous in any of these fields, the most hallmark developements are made by going deep. By age 30, I want to have spent 10,000 hours in one such specific domain, where I know I am capable of providing value and breaking boundaries.
5. Run a 5 minute mile. I've always been a runner, but I've never been able to break the 5 minute mile. I don't think I will get any faster in my 30's either.
6. Travel 30 countries. While I already consider myself fairly well-traveled, I find that understanding cultural nuances of other countries and customs at a microscopic level helps me better understand my own identity.
7. Read, a lot.