Introduction

I received an email from the head of software at Neuralink. He’s also a CTF player, and he asked if I’d like to help found a team at Neuralink. The email became interviews, which became an offer. Now, I need to decide if I want to take time off of school to work there. There are some pushes and some pulls, and by putting my ideas on paper, I hope to more carefully decide the strength of each one of these forces. I have a couple straightforward metrics in mind that I can use to mentally appraise any position (Is it interesting? Is it impactful? Is it financially rewarding?), but, in reality, this specific decision process requires weighing a mishmash of complicated factors.

The Opportunity Cost

What will I be doing if I don’t take my Neuralink offer? Well, I can stay in school or work in vulnerability research. The other options I have besides these three don’t match my risk profile.

My internship this summer was in the vulnerability research industry, and honestly, it’s fun! I like VR’s day-to-day work, comparatively immediate money, and familiar social environment of CTF players. Unfortunately, VR doesn’t offer you a huge potential upside like a startup does, and also doesn’t seem like a high impact positive force for world. VR initially feels like a good option if I didn’t want to take on much risk, but given that I can only consider my VR offers by taking time off of school, it’s not a super sensible choice to me right now. Maybe if I was optimizing for minimal risk, I’d work in VR after graduation.

So, Neuralink’s next best alternative is staying in school.

Life enjoyment against life enjoyment per probability. Although scale is important, my feeble human fingers struggle to capture the fuzzy probability distributions approximated by my feeble human mind.

Thoughts on School

By far, Caltech’s biggest pull is its people. I love the people in Dabney (and my mole friends in Blacker!), and I think there are very few other places with this density of kind and accepting yet passionate and ambitious people. It’d really hurt to move away from this community, but I’ll have to do this in two or three years anyways. Still, I’d much rather later than sooner.

Also, I think holding a Caltech degree unlocks a lot of opportunities, (e.g. securing an interview, raising money for a company). In some tech circles, dropping out of a school with “prestige” fetches the same price as holding an equivalent degree, but this is definitely not the norm and I feel it’s risky to bet this’ll be the case years from now. I don’t weigh this too highly since Caltech is just one of many ways to get your foot in the door. For some hiring or funding processes, Neuralink is a more meaningful signal than Caltech.

Caltech’s biggest push is its classes. I don’t think that’s a Caltech problem; it’s more of a me problem. I’m self-taught and I’ve never enjoyed or learned well in classroom environments, so I end up skipping almost all of my classes. Caltech’s humanities requirements eat up my time, and I’d rather spend those hours learning by myself about the things I enjoy. I don’t like the thought of writing another essay about a topic I don’t feel strongly about.

Thoughts on Timing

I don’t have to choose between Neuralink or Caltech. I can get my degree first, then ask Neuralink for another offer. I can also work for Neuralink for a year and then complete my degree.

I think Neuralink is more time sensitive than school, but not by much. I don’t think I’ll have the same opportunity to shape the infrastructure team at Neuralink when I come back, if at all. However, because Neuralink promotes quickly and gives individual engineers lots of freedom, this is less relevant than equity maturity and FOMO.

Thoughts on Big Ticket Items

Okay, that’s cool, but will people actually want to use BCIs?

• BCIs have already found a lot of use for those with motor disabilities. I think there’s a very continuous path to mainstream adoption if its potential uses begin to stack, starting from medical uses, to industrial (teleoperation of robots for dangerous situations) and gaming (probably not SAO, sorry) uses, to widespread adoption. However, going mainstream requires offering a serious QoL improvement with essentially no downside, which seems really hard to do.

• I think there’s a very real possibility that BCIs stall in the same way virtual reality headsets did. VR definitely failed to provide the many-upsides-with-no-downsides expected of a mainstream technology, which is why people don’t really live in the “Metaverse”. One big barrier of entry is the, yknow, intrusive brain surgery, but sentiment is that with a precise enough surgery robot this could be relatively unimposing. Some promotional Kool-aid I was served at Neuralink featured a future where you stop by a Neuralink clinic, pick your implant color, and had a painless 15-minute surgery, but honestly, I’m pretty skeptical this will happen in the next 5-10 years. Then again, I have no idea.

• I don’t think people are eager to get chips in their head, but I think that pacemakers and LASIK felt that way once too. If a BCI implant can demonstrate real benefits, I think it’s plausible that the predisposition against BCI will soften. I think Neuralink being an Elon company certainly doesn’t help…

• Stimulation is not a far-future idea. The Blindsight project has already shown it can inject arbitrary black and white dots into the vision of a monkey. This to me is one of the most exciting projects of Neuralink. From my CTFer perspective, I thought we had read access to the brain, but apparently we already have a practical write primitive. I’m so curious if we’ll have a real life HUD in the near future.

Thoughts from the Monkey Brain

I think all the arguments above don’t really change how I feel.

I am not an extreme risk-taker, and this seems risky. I’m scared Neuralink will sink, taking my time and effort with it. I’m scared of closing the doors I would’ve kept open, of never meeting the people I would’ve met. Most of all, I am scared of drifting—really, of losing the incubation time that would’ve solidified my relationships post-graduation—from my friends, my girlfriend, and the community that I cherish like a second family. I fear that when I return to visit, no one will recognize me as their friend or as a Darb.

But for every ounce of fear I’m grappling with, I feel twice as much nervous excitement. I want to work at the frontier of something, tread new ground, and sate my craving for novelty and challenge. I don’t want to solve more contrived competition or homework problems—I want to see what I can do when I give my all to solving real issues plaguing real people.

As a kid, I found novelty and challenge in pwnable.tw and OSDev knockoffs, and that was enough to excite me. Gradually, my projects also needed to feel important for me to be equally enthused. Maybe my brain learned that, when interesting projects are plentiful, it can cram in a secondary metric to keep its dopamine bar raised. Now, I sense the once-waning high bubbling up again.

So, I’m doing it. While I don’t think I can reasonably justify the decision to work now over later, I’m too hyped to say no. Given the fear I’m feeling, this may not be a rational decision. But, in keeping with my younger self, I’ll put off school to work on what I find exciting right now. Maybe I’ll get back to it in a quarter, or maybe a year. Maybe I won’t.

My new team