choosing by Vincent H. '23

in which i document some of my thoughts regarding my job search and how i decided what to do after graduation. this is a condensed version of a more comprehensive post that i wrote for myself but didn’t think was appropriate to put on the mit admissions blog; feel free to contact me privately if you want to read the longer version though!

1. options

at the end of my job search i had offers from the following three places, and i was too tired to look for more options:

• a large quant hedge fund, let’s call them D. i had an offer for a quant research (QR) position, which i wasn’t super excited about because i don’t think working in finance is particularly good for the world (i don’t think it’s bad either; i think it’s very neutral) and i also don’t think QR work is that interesting (i previously did a QR internship and it was mostly just applying linear regression in various forms). however, it did mean i’d be living in new york, which i was interested in because most of my friends will be there next year, and obviously it was extremely high-paying. so even though my initial impression was negative, i thought if i committed to doing QR as my full-time job i’d likely end up discovering parts of the role that i enjoyed
• a small trading startup, let’s call them A. while i’m not very interested in finance itself, i do think it’s impressive that trading companies have extremely performant and robust software. the problem is that, at larger firms, these systems have basically already been developed and are only being improved incrementally, so it seemed like if i joined a smaller company still in its early stages there would be a lot of interesting systems to build. in general i also just prefer working in smaller teams, so somewhere like A would be significantly more enjoyable for me than traditional finance places. other relevant considerations were that they’re in-person in chicago, and pay similarly to D
• an ai research startup, let’s call them G. one of my big regrets in college was not attempting ai research seriously, and G seemed like a good way to try that out. the part of their work which i’m excited about is developing ai tools that can help humans with more creative tasks (more broadly than the current paradigm of generating images and autocompleting text). another thing that stood out to me was that they have probably the highest-eq team i’ve ever encountered – throughout the process i felt like they were trying very hard to understand who i was as a person, the ceo read my mind several times and gave me good career-agnostic advice, etc. however, the main downsides were a) they’re in-person in sf and almost all my friends are staying on the east coast b) it wasn’t clear to me if they would be successful or not, since their research direction is somewhat unorthodox and some of my ai research friends did not think it was promising c) i felt a lot of insecurity related to committing to ai and maybe not being competent enough to make meaningful progress in the field

initially i wasn’t that enthusiastic about any of these options because they all presented different tradeoffs and i didn’t want to accept any tradeoff whatsoever. after all, i’ve basically been in school for my entire life and haven’t seen that much of the world yet. how am i supposed to know how much i actually care about factors like changing the world or having money or doing good research, and how am i supposed to understand the value of each of those factors well enough to choose between them? sure, i think i know what i care about, but only from the perspective of a student who commits to projects for a semester or a summer at a time. so how do i choose something that will last for multiple years, possibly indefinitely? i’m reminded of this line from the commencement speech this is water:

“The plain fact is that you graduating seniors do not yet have any clue what ‘day in day out’ really means. There happen to be whole, large parts of adult American life that nobody talks about in commencement speeches. One such part involves boredom, routine, and petty frustration. The parents and older folks here will know all too well what I’m talking about.”

it was so tempting to just say no to everyone and try working on projects on my own for a year, and maybe support myself through tutoring and couchsurfing or something. if i was a little more impulsive and reckless maybe i would’ve actually gone through with that. somewhere in my head i knew that wasn’t actually the right choice though – i’m the kind of person who’s attracted to uncertainty because it allows me to project whatever i want into the void, but realistically i think if i went off for a year starting now i’d end up being less productive and more confused than if i stayed somewhere with reasonable mentorship

2. choosing

i was very lucky and had offers from D, A, and G at the same time, so i spent some time trying to compare them to each other. for a while i tried to get the best of all worlds – i was looking for a way to do interesting research and engineering work while staying on or near the east coast and making as much money as possible. this would probably involve either convincing an employer to let me work remotely, or switching jobs after one or two years

i also tried to answer the question from a different angle – what problems do i care enough about to spend a large portion of my life tackling? i brainstormed a bit and came up with three general areas – ai, climate, and education (i am being deliberately vague here because the details would take another post to explain. also, this list will probably change in the future – for instance, i could definitely see brain tech or certain kinds of drug discovery being included). from that perspective it seemed like the obvious choice was G

but there’s a fundamental issue with this kind of reasoning – it assumes that you should pick a job based on what kind of problems you want to solve. you could just as easily argue that you should pick a job based on where you want to live or based on which job maximizes wealth (and people follow both of these guidelines all the time), in which case D or A would be the right answer. brainstorming which problems i cared enough about to dedicate my life to was a helpful exercise that clarified my intellectual interests, but that on its own couldn’t be the basis for deciding

we could spend all day arguing about a specific choice, or we could go up one level and argue about decision-making criteria and metrics, or we could go up another level and argue about the process of picking a decision-making process, or so on. i think all supposedly rational decision-making processes fall to this problem of infinite descent, of being turtles all the way down

in the end, none of the analysis i did in the previous section was particularly helpful. instead, it was a series of emotional heuristics pointing in the same direction that finalized my decision

2.1. prisoner’s dilemma

one extremely, extremely reductive way to model people’s career choices is as a prisoner’s dilemma: a multiplayer game where each player can choose to cooperate (contribute to social good by working on impactful problems, while gaining less personal utility) or defect (maximize personal utility while working on something that doesn’t necessarily provide much value to other people). this isn’t actually a good representation of the world because choices are rarely binary and impact and utility are complicated, but just bear with me for a few minutes for the sake of analogy

anyway, in the classical version of the prisoner’s dilemma, people defect because it’s the rational thing to do

but one of the main reasons everyone always defects in the classical prisoner’s dilemma is because they’re all making choices independently. the career decision problem is different, because all of our choices are correlated and entangled in ways that are difficult to quantify – one person exploring a new field can cause their whole social circle to begin exploring it, people who start companies often hire many of their friends, and so on. or, to explain this more generally:

i thought about all the people older than me who have inspired me and shaped my life over the years, and then i thought about all the people younger than me who look up to me for advice. just as i use the stories of people who have come before me as benchmarks and references for what i should do, i know that people who come after me will use my story as guidance for what they can do

and while there are many parts of this decision that i’m very confused about, here is one fact i am absolutely certain of: i want to live in a world where young people choose to work on interesting, impactful, and difficult problems

it sounds silly and irrational when i say it out loud, but i believe that we are unstoppable, that if we put all our heads together and try really hard we can solve almost every problem currently plaguing this world. and when i think about the fact that me cooperating will likely play a small role in many other people deciding to cooperate, it begins to feel like the obvious right choice

this is the direction i want to push the world in, and once i thought about it in those terms i simply couldn’t bring myself to defect anymore

2.2. freedom

the thing i like the most about growing up is that you become more and more free over time. this process comes in many forms: you gain freedom from people like your parents telling you what to do, you gain freedom in social interactions as you learn how to have better conversations and be a better friend, you gain freedom in what you can create as you learn skills like programming or drawing or electrical engineering, and so on

that’s a big part of why this quote from one of my favorite blog posts resonates with me so strongly:

“Don’t surround yourself with ‘smarter’ people. The trick is to surround yourself with people who are free in ways you’re not.”

i’m extremely familiar with the kind of person who goes to work in trading, because i did math contests with many of them for most of my childhood. a lot of them are smarter than me, or better at problem-solving, or can process information faster than me. but are they free in ways that i am not?

i know that i can learn engineering and modeling and statistics wherever i go. but where will i learn how to be a better research collaborator, or how to inspire other people, or how to have a less anxious relationship with work?

it’s too early to be sure, but i really like what i’ve seen from the people at G so far and i think it’s the right choice for all the nontechnical directions i want to improve in

2.3. ghosts

there are two ghosts i’ve been haunted by lately. haunted in the sense that they keep coming back to me, without me trying to remember them, at least once every few days. here’s the first one:

“How we spend our days is, of course, how we spend our lives.” – Annie Dillard

so many times in the past year i’ve been frustrated with what i was been doing, with feeling like i was taking too long to do homework or write code or take calls that felt meaningless and not having time to do things i wanted to as a result, and sometimes i would tell myself after all this is over i’ll have time for things that matter. and then it would dawn on me that how we spend our days is how we spend our lives, that this feeling of spending all my time on unimportant things will never end unless i end it myself, that i could easily keep this up for weeks and months and years and i would never make much progress on anything in particular

and here’s the second one:

“The part of life we really live is small. All the rest is not life, but merely time.” – Seneca

i don’t want the hours i spend working to be merely time passed and money accumulated and tasks completed

life is too short to spend so much time not really living, and when i think about what kind of work makes me feel alive, one common theme is that of creation, of making something meaningfully different from anything that has come before it, of unlocking something new in the world. that’s one of the reasons i enjoy writing and arranging song mashups and teaching web development. it’s also one of the common themes in the three areas i mentioned earlier of ai, climate, and education – each one, if handled correctly, will give us a lot of capabilities we didn’t have before. that’s a big part of why these subjects feel fun for me to explore, even if objectively speaking i’m unqualified to do so

2.4. conclusion

to be honest, i still have some doubts about committing to working on ai research. part of it is because i have a hard time committing to things in general, and part of it is because i think the problems in the space are much more complex than anything i’ve worked on before. but i’ve realized that even if i don’t come up with anything, even if i spend all my time flailing around and making no progress whatsoever, even if everything work-related goes wrong, i still won’t regret this decision because i believe these problems are worth spending a few years trying. at the very least i’ll get a better sense of what it’s like to work in the space, as well as what my own limitations are, and if nothing goes well then afterwards i can find a different field to work in

i’m also still pretty sad about leaving my east coast friends and moving to sf. i’ll try to keep in touch and visit often, but i’m not really sure what else there is to be done

at the same time, there’s a lot to look forward to. for instance, some of my current interests are pipe dreams i’ve had ever since i was a kid – can we build systems that turn memories into music? can we make interpersonal communication less ambiguous by augmenting it with context and subtext? and so on – and i think we’ll get closer to answering these questions by exploring ai

for that reason and all the other ones i’ve written, i’m very excited to spend the next few years at generally intelligent :)