I quit my job.

This was tough, because I liked my job. I was good at it, it was intellectually stimulating, it paid well, I found the company’s mission meaningful and important, and being a rocket scientist is still as cool as it was in the ’60s. As someone who has always tied an uncomfortable-to-admit amount of their sense of self worth to the work they are doing, this was not so easy to step away from. Because you see – there isn’t any next step lined up.

Back when I bought my house 3 years ago, I was looking at the county map of property lines and noticed that my lot seemed to cut right through the middle of the neighboring apartment building’s dog park. Turns out they built it on my property. After a year of playing bureaucratic tennis with the city and our two title companies, they resolved the mixup by buying the land under my half of the dog park for $60k. This gift from the universe got squirreled away until it could find its purpose. Now it has – 12 months for me to take off work and attack the world of possibilities. The dog park sabbatical.

Inspiration

Foundation

This idea has been percolating for years. I’ve been following the save & invest & grow income mantra of the financial independence community since the start of my career, but its promise to retire early to a life of leisure never had any appeal. What I want instead is the financial freedom to work on things that I think are interesting and important, with the autonomy and creative control to drive them as I see fit. Projects and pursuits as personal expression.

There’s no fear of squandering the time. When I’ve switched jobs in the past I’ve always taken a few weeks off in between, and those stretches of time always end up being my most productive. Right now I’m positively buzzing, and 12 months feels downright endless.

Friends

I think about the quality I most admire in people, and I think it’s the courage to take a leap away from traditional paths to pour themselves into passions. One of my best friends from high school left her Silicon Valley job to move to New York and dance. Another high school classmate left investment banking to become a cyclist – and just now represented the US in the Olympics. Another close friend in Denver is a year into freelance consulting. And my girlfriend has recently taken the leap to jumpstart her own business. These people I admire and respect all the more for charging headfirst into uncertainty, whether they ultimately succeed or not.

Framing

Palladium Magazine’s “Quit Your Job” has been enormously influential on my thinking here. The framing of flourishing and finding one’s niche outside of predefined tracks as noblesse oblige – a virtue, even an obligation for those with the ability – I find very powerful.

Echoes here of “Our deepest fear“, and of Nietzsche:

Anyone who manages to experience the history of humanity as a whole as his own history will feel in an enormously generalized way all the grief of an invalid who thinks of health, of an old man who thinks of the dream of his youth, of a lover deprived of his beloved, of the martyr whose ideal is perishing, of the hero on the evening after a battle that has decided nothing but brought him wounds and the loss of his friend. But if one endured, if one could endure this immense sum of grief of all kinds while yet being the hero who, as the second day of battle breaks, welcomes the dawn and his fortune, being a person whose horizon encompasses thousands of years, past and future, being the heir of all the nobility of all past spirit – an heir with a sense of obligation, the most aristocratic of old nobles and at the same time the first of a new nobility – the like of which no age has yet seen or dreamed of; if one could burden one’s soul with all of this – the oldest, the newest, losses, hopes, conquests, and the victories of humanity; if one could finally contain all this in one soul and crowd it into a single feeling – this would surely have to result in a happiness that humanity has not known so far: the happiness of a god full of power and love, full of tears and laughter, a happiness that, like the sun in the evening, continually bestows its inexhaustible riches, pouring them into the sea, feeling richest, as the sun does, only when even the poorest fishermen is still rowing with golden oars! This godlike feeling would then be called – humaneness.

Timing

Why now?

I’ve finally found someone who I could see marrying. I’m going to skip over the sappy bits in order to keep this post focused, but in terms of going at it solo there are two implications. First, there is comfort in a stable relationship. Gone is the anxiety of trying to find a partner on a timeline that fits my plans for a future family, replaced with the assurance of someone who you know will be there to grow together with. Second, kids are still in the future but it’s a future that’s less and less nebulous by the day. The sense that right now is a window of opportunity to take higher risks in my career has never been more salient.

This is also a slice of a broader moment in time. There’s never been a better time to be a solo researcher or entrepreneur – AI tools have made learning and building new things with new technologies orders of magnitude faster & easier. I think there is tremendous opportunity for a motivated individual to tackle projects in months that would have before needed teams of people and years of work. If anything, I’m a bit late to this wave.

And the rest of my life is in a really good place. I’ve reached a point at my current company where I’ve done or roadmapped & delegated all the major improvements that I clocked when I joined up three and half years ago, and we just completed my largest project of the year. My new hire has exceeded my expectations and I think is capable of picking up the gaps when I go. I own my home, with no big financial plans on the horizon I would need to prove income for. Beyond the dog park money, I have decent savings and no non-mortgage debt – FIRE might be accelerated by sticking with my current income, but I’d just be accelerating towards the kind of life that I could simply start living right now.

Another way I’ve thought about this, is that my present self trusted my past selves to have made good decisions in our mutual self interest, and my future self will say the same of the actions I take here. If not with approval, then he will at least be forgiving.

Why not now?

It’s going to cost money. Not the $60k – that would be spent anyways – but the year of income which would be shoveled onto the pile of compounding growth. But I think it’s a price worth paying. And even then, the skills I’m hoping to pick up are valuable in the job market so from that lens this is a period of upskilling.

It’s also scary in the sense of facing down unknowns, untethered. Writing this post has been incredibly helpful in terms of making those fears concrete. It has also made the opportunities overwhelming. I’ve heard other people say that the decision to cut loose and strike out on your own is never a rational decision, but something that they couldn’t imagine not doing. I’ve hit that point.

What’s the plan?

Back to School to study AI

The first thing I want to do is take myself back to school and learn all I can about modern AI methods. This is an absolutely fascinating topic that I’ve been following and would love to dive into. Anthropic’s Toy Models of Superposition white paper is perhaps the most interesting and engaging technical document I’ve ever read. My impression is that even as the frontier research into LLM capabilities is siloed into the large labs with staggering compute resources that are basically required now to make progress, basic interpretability is still in infancy and there may be room for solo researchers to contribute meaningful advancements in understanding what is happening inside their giant inscrutable matrices.

Outside of personal interest, there is a strong sense of mission and gravitas that draws me towards the area. There are AI true believers that think the rapid advancement and widespread adoption of artificial intelligence will bring transformative societal changes on an unprecedented scale, and that the world 10 years from now will look fundamentally different in a period of unprecedented change & opportunity. I think they are almost certainly right. There are also AI doomers that think this will bring about the end of humanity. Even if the chance is small, I think they are right enough that aligning AI to human values is almost certainly the most important thing in the world to be working on. And that requires understanding it.

I’m not sure I have the chops to make a meaningful impact here. My hard computer science background is a big weak point – ask me to whiteboard a program in C++ and I’d sit there blankly. I have absolutely zero knowledge of networking, GPUs, webdev frontend or backend, cybersecurity, really any programming languages besides python, or a hundred other topics that everyone on Hacker News seems to be familiar with. Aerospace engineering doesn’t have the same software focus as Silicon Valley. I also may simply be too late to make an impact – certainly I’m several years behind a sweet spot. But on the other hand, things such as being nominated to be a core matplotlib developer or giving a talk at the SciPy conference are reassuring external proofs that I’m not totally off the mark. My math feels sharp enough to struggle through the theoretical papers in the area. Several points in my career I’ve solved & communicated out hard, technical problems with clean approaches that were completely new. Most of these fell out of thinking hard about the right way to frame the problem, and creating visualizations that made the answers obvious (if I have a superpower in my career, I think that would be it). And I’ve seen at least one other person make the switch to the field from a similar spot. All this to say, I think there’s enough in my toolbox that I might be able to help, and 3-4 months of study is hopefully enough to dip my toe deep enough to take the temperature of the water. It’s worth a shot.

Starting a Business

If that path doesn’t look fruitful by early spring, I’d love to try out a couple business ideas, throwing some things at the wall and seeing what sticks. This can be rapid – I really like Pieter Levels’ idea of going all in on a micro-startup idea every month to validate ideas. And there are many different frameworks out there about how to choose what to work on.

Dropping the veil of polite modesty for a moment – I was extremely good at my job. I saved the company from one existential peril and greatly increased satellite imaging capacity. Before I hired an employee two years in, I was covering two major technical areas solo, which at most space companies take a few people each. I developed firm theoretical foundations and technical requirements for those areas from scratch, ran back-to-back complex multidisciplinary hardware/software projects that delivered successfully and on time, hired, onboard, and managed new employees to flourish in their own successes. I’ve been working hard and kicking ass. And the voice in my head has been saying, “if you’re able to churn out results like this, then why not capture the upside?”

So I think I have the capacity to drive a small business to success. But I don’t yet know what one would look like. I think I need the time to let ideas go fallow and see what crops up.

Projects

At the risk of following too many diversions to the main goals, I have several projects percolating that I’ll have time to brew.

• Finish a physical model of the Antikythera Mechanism, which has been lying dormant as a project for years.
• Make more wooden topo maps – I’m itching to do one of Colorado, and I want to do a giant one of Mars. There’s some interesting data processing automation for this that should be fun.
• In the open source world, I want to squash matplotlib’s 2nd oldest bug and enable log scale axes for its 3D plots. I have a very specific plot in mind that I want to make which requires this, and it looks like many other people do too.
• Also in that realm, I’d like to add to scipy’s Rotation module to allow for translations and general rigid transformations in 3D, allowing for representation of generalized coordinate frames.
• It would be cool to make a website that predicts satellite lifetime given ballistic freefall and statistical ranges of space weather. This could perhaps spin up into a business.
• I’m more worried about having too many ideas here than too few!

Personal

I’m in my early 30s, and want to use this time to get into the best shape of my life – if not now then when? I want to hit intermediate level weights on all my main weightlifting exercises, and I signed up for a marathon in May where I want to finally run it in under 4 hours. Ski season is also going to be big this year, and I’m going in on a shared condo with some friends to have a place to stay up in the mountains and avoid the seasonal traffic.

Comms

I have never been a prolific online poster. But for my work either in AI research or in a personal business to succeed, it’ll need an audience. The mantra of “build in public” greatly appeals to me (though I’m wary of the selection effect that you never see the people that build in private). Twitter seems to be the platform most aligned with these niches. So my thought is to shoot for at least 1 tweet per day, and have a weekly recap of what I did each week as a blog post. I think this will also benefit from a rebrand away from my real name to a pseudonym. Still a little unsure about the comms approach, and will have to flesh it out.

The Road Ahead

I gave quite a bit of notice, and won’t actually be leaving until mid-November. But the wheels have been set into motion.

These 12 months won’t be an escape – they’ll be my chance to lean into work I find meaningful, and prove out the possibility of a life that is purposefully and personally crafted. I’m excited to see what comes.