AI Alignment Crash Course

2/8/2025

Here are a few quick insights into the state of AI alignment research.

Lessons in the Green

1/20/2025

I usually set up an army of plants anywhere I stay. Most likely, they're satisfying a biological urge to nurture living things. Traditionally, that would be a baby. But currently, it is plants.

Shock and Awe

1/1/2025

A beginner's brief to cold plunging.

Bonfire from First Principles

9/5/2024

A beginner's guide to setting up a big fat bonfire. You need three ingredients.

AI Regulation Risk

11/4/2023

It's important to think deeply about guidelines around AI safety and risk. It's critical that we continue to build in the open to protect against unilateral control of AI.

Why I'm working on an AI startup

7/22/2023

I decided to start Khoj to build a trustworthy, accessible personal AI assistant for everyday users. I outline some of the reasons why in this post.

Ciudad de México

4/30/2023

After three months of living in Mexico City, I'm firmly convinced that it's one of the most livable cities I've ever visited.

Machine Learning for Imbalanced Datasets: A Guide for Engineers

4/7/2023

Machine learning models are often trained on imbalanced datasets. This leads to poor performance on the minority class. This post discusses some techniques to address this problem.

Ciudad de Buenos Aires

12/5/2022

Brief thoughts after one week in Argentina's capital.

Doing Good Better

10/7/2022

How do we do good in a more quantitative, data-driven method?

Sailing Mombasa

9/13/2022

Recounting the dynamics of a sailing stint in Mombasa, Kenya.

Meditations on Meditation

9/11/2022

Sometimes, our brains need a little pruning.

Wander, Forecast, Iterate: Thoughts on Better Decision-Making

7/3/2022

We suck at making decisions. Our brains are adept storytellers that dislike being measured against reality. Enmeshed in our biases, we actively reject information that proves us false. Making better decision is of existential importance to humanity. Systems that safeguard against these common failure modes are vital to improve our collective decision-making.

Theory of Religion

3/6/2022

Since the dawn of humanity, we have searched for meaning. Civilizations have turned to mythical explanations borne of the imagination to understand how the world came to be. As time passes, we reflect on what these stories say about people themselves.

Progressive Civility

1/17/2022

In our imagined annals of justice, we need to stand up for what is right to create a more equal society. The romantic ideals of humanity and equality need to be actively pursued.

Time Broker

12/31/2021

We have to be extremely frugal over our time, less the world keep extracting more from us. Do one thing at a time, and get more value for your effort.

Career

10/5/2021

Our life plans are often fraught with loopholes. We look for ways to find meaning in our fleeting blink of a life, taking uncertain twists and turns along the journey.

An Ode to Programming

9/19/2021

Being a programmer gives me a sense of wonder and self-efficacy. It helps me see the world more clearly, and allows me to engage with my surroundings in meaningful ways. This is basically a love letter to writing code.

Artashastra: Thoughts

9/10/2021

My thoughts of the book Artashastra, written by Kautilya sometime around year 0 AD. This text is best described as a state's manual for Ancient Indian Kingdoms, but provides some useful insights into industry and society today.

Economically Vegetarian

8/12/2021

Sometimes I like to imagine that my pseudo-vegetarian diet is saving the planet, but I wonder if the decision really holds economic and impact-based merit.

Earthnauts

8/7/2021

We are all explorers of our home planet. Is there anything that makes us special in this regard? Is there something we were engineered to be best at?