A little over three years ago, I won the lottery.

I received a job offer from Khan Academy.

After months of preparation, my dream came true – I was given the opportunity to grow alongside some of the brightest, kindest minds in the industry, to work on one of the most important problems around. I couldn’t believe my luck.

During that time, I learned and grew so, so much – personally and professionally. I am eternally grateful to KA for providing me with a safe space and supportive environment in which to grow.

KA is a truly incredible place to work, a dream company in so many ways, all due to the capable, creative, and endearingly humble people who work there.

But now, like a hermit crab, I must move on to a new shell. To the next stage in my journey.

Where’s that, you might ask?

Well, you see, when I applied, I actually didn’t submit a real resume. Instead, I submitted a kind of ridiculous “pitch deck” on myself.

In it, I included the following slide:

Slide from my "pitch deck resume" stating my longer-term goal to "build a company, nonprofit, or social movement that helps people change their mindset to achieve their goals"

After all these years (and much reflection on how specifically I want to spend my one wild and precious life), my personal mission hasn’t changed all that much – something I find deeply reassuring.

I want to equip myself and others with the tools, know-how, and support to become more compassionate and competent human beings.

I care deeply about building a more just and fun world.

So, my hypotheses about how to achieve my mission have changed, but not the target itself.1

As my mission crystallized, I realized that I needed a different learning environment to accelerate my progress towards it. I thought back on the times I was most energized at KA, when I learned the most.

It was when I worked on an experiment called the “Teams” experiment – an experiment designed to answer the question “Can we motivate students on Khan Academy by challenging them to work in small teams to achieve a difficult goal?”

It was honestly one of the most exhilarating times of my life. As the only person working full-time on the project, fueled by the boundless energy and enthusiasm of my incredible project partner Kitt (a designer, former Hollywood screenwriter, and one of the funniest people I know), I ate, slept, breathed Teams.

In four short weeks, we spec’ed, prototyped, coded, and launched the experiment to 355 excited learners picked from a lottery and watched it go. And boy, was it amazing. Some of the teams were unbelievably diverse, due to our random assignment of teams (done for the purposes of making things easier). Here’s an example:

Bios of one of the teams that competed in our first Team experiment, including a 30-year-old mom, a 20-year old Swiss computer science student, and a 9th-grade student from a school in Thailand

I played customer support, eagerly monitored the “team chat” boards we hacked together, deployed surveys, and analyzed survey and A/B test results.

At the end of it, nearly 90% of teams achieved the goal. A goal that was specifically chosen as a high-water mark from this group’s past performance. It ended up being one of the A/B tests with the most drastic (statistically significant) improvement in core metrics and social metrics (e.g., profile updates, discussion posts) we’d ever seen.

The sketch by Kitt and me that started it all…

Kitt and my super rough sketch of teams when we thought of the idea

…that evolved into this “team page”, which we designed, developed, tested, and launched four weeks after starting the project:

Screenshot of our team's team page

Despite the initial experiment’s success, there were some key challenges ahead. One of which was to understand whether this bump in engagement could be replicated over multiple challenges. That is, would students sign up for a subsequent Challenge, and if they did, would they participate to the same degree?

Due to resourcing constraints, we deployed a follow-up experiment with a set-up that I knew was non-ideal. We loosened qualification criteria to capture less engaged users and ran 3 identical challenges back-to-back 3 weekends in a row.

As I’d feared, engagement waned with each subsequent challenge.

I realized after the second experiment that getting Teams right would not be an easy win. Social dynamics between even just two people, let alone more than two, are hard to model and control. So I let go of my death grip on my “baby” and admitted to myself it wasn’t right for KA right now – that we were too far from low-hanging fruit territory to keep going.

In the follow-up experiment, we also tried Solo challenges – an almost identical experience except without the team assignment component. More people signed up for Solo challenges and more of them reached their goals, but fewer of them far exceeded the goal or went on to do another challenge. In short, the experience was kinda boring.

Later, KA invested in the LearnStorm competition – a programs-led Bay Area-wide inter-school competition on Khan Academy which has seen some really stellar participation rates and sparked a lot of energy and excitement. I’m so glad we made this first foray into social.

But part of me is still drawn to this idea of connecting people who might not otherwise know each other, who live in drastically different parts of the world and come from drastically different walks of life, but have shared interests or values. I’d love to see KA later pursue LearnStorm at a global scale, with diverse teams drawn up not by geography, but by interest or life circumstances. Think Harry Potter houses meets Khan Academy. (Okay, okay… enough of the crazy brainstorming.)


The point of this very long aside was to say that I realize that what I crave is a similar sort of learning experience where I’m the owner and driver of an idea, 100%. Where a failure means that I really do fail.

You might say, raising the stakes or something like that. But I think of it as just removing a few layers between me and feedback. Getting rid of constraints like having to model a company’s current product and/or strategy and instead just working in the global space of the most pressing and interesting problems I can think of working on.

To dream about moonshot ideas and start from first principles.

To rebuild my hustle muscle, which I admit has been atrophying over the past couple of years.

To stress and strengthen my mental, emotional, and physical fortitude, which I’ve been steadily building up and am ready to put to the test.

In short, it’s go-time.

I’m giving myself a year to run experiments and working on projects designed to bolster my ultimate odds of succeeding at my long-term mission.

A core goal in all of this is to develop expertise, measured by a deep understanding of the specific conditions under which certain interventions do or do not work.

The curriculum

Specifically, my work will include (but is not limited to):

  • Optimizing my health and habits to help me steadily level up in compassion and competence
  • Studying disparate domains to learn as much as I can, as quickly as I can, about the many avenues to behavior change
  • Developing frameworks and testing hypotheses around helping people reprogram themselves for success
  • Honing my writing and thinking chops
  • Practicing the skill of developing product-market fit
  • Sharing the best of what I’ve been learning

Concretely (hehe), here’s a list of some of the areas I’m studying:

To counter possible anxiety from having no income for at least a couple of months, I’m framing it as a year-long free Master’s program, custom-tuned to my needs and interests.

A steal, right?!

I’m so excited and grateful to have this opportunity to work 100% on a problem that I care about so much (and that I personally think is one of the most important and universal problems of our time, given how much it’s affected me, the people around me, and the students I’ve met in the course of my work at KA).

Thanks for reading this far and for being a part of my journey, anonymous nyan cat. Seriously. You give me strength, even though I don’t know you yet.



P.S. - If you struggle with setting and achieving goals, I’d really like to help you become your best self. Please consider subscribing to my personal email newsletter for the inside scoop on my personal journey or check out HabitLab, the site where I focus more on specific behavior change techniques that you can immediately apply in your life. Subscribe to the HabitLab newsletter to get entertaining pointers to the most helpful resources I’m discovering along the way.

  1. If you’re curious about my thought process, I used to assume that the only way to get to this goal was to create an incredibly beautiful, groundbreaking MMORPG that taught people empathy by letting people role-play different characters – e.g., a homeless child in a New York City shelter, an impoverished sex worker in Bangladesh, an LGBTQ person trapped in an intensely homophobic community. I have since realized, in part from reading Lean Startup by Eric Ries, that this probably isn’t the only way and that there might be leaner approaches that could be folded into such a game in the future, when the technology/medium is more mature and when virtual reality is more mainstream. I also decided that I’d rather specialize in storytelling and psychology than in game programming and design, if I have to pick a focus area – which, due to life constraints, I must.