a different kind of impressionism

i have slowly come to the realization that i am not as strong-willed and independent-minded as i once ventured to believe. i will read a persuasive article and find myself persuaded; then i will read a refutation and also find myself persuaded. i will feel strongly that a political stance is morally wrong, then talk to my politically opposite friend for three hours and feel less certain about my stance.

i believe at least part of the above comes from my ability to empathize in ways that are deeper than what would likely be considered normal.

i am an empath, which means i’m highly sensitive to the emotions of others and can almost feel them myself—i’m very likely to cry if someone i love starts crying, i feel the pain of others onscreen and in books to an unnatural degree, my heart breaks when i read the news.

this trait makes me both impressionable and desperate to impress. i can feel the weight of others’ convictions so deeply that i sometimes adopt them as my own; i can sense others’ approval so acutely that i sometimes fall into the trap of organizing my life around earning it. i have a strong desire to impress people and live up to the standards of those i admire most.

the most egregious way in which this has manifested is through my abrupt transition from projects focused on [environmentalism, depolarization, and fundraising for ukraine] to ai safety.

i genuinely do believe ai safety is the most pressing problem facing humanity. it’s a uniquely urgent existential risk—one with the power to cause human extinction within the next decade, according to some timelines—and i am fairly confident that AIS is the most impactful field i could be involved with at this moment in time.

but if i’m being honest with myself, i think my motivation for getting involved in the first place was nontrivially associated with my desire to impress the people i know who are in the field, as well as impress others outside the field by signaling involvement with the field.

i admire my friends in ai safety endlessly, and i saw that because of its novelty and concentrated nature, AIS was a field in which young people were actually doing work with impact—some of the most impactful, high-level roles in AIS are held by exceptionally young people, which is not something that exists in most spaces; i wanted to join them. i wanted to do something that actually felt important for once. i was lacking a sense of direction, and i saw all of my friends were getting involved with the shiny new field of ai safety, and i wanted to do something that mattered—but i also wanted to do something that i thought would help me earn the respect of the friends who i suspected saw me as “lesser than” because i come from a much less technical background than the vast majority of them.

so i jumped in. i started applying to fellowships, reading papers, working at organizations, meeting people in the field. and the work is genuinely fascinating, and invigorating, and exciting and fast-moving and dynamic and everything i’ve ever wanted in a career field—i’m not pretending to care about something i don’t.

but what worries me is that in making my initial decision, i didn’t stop to consider whether or not this was the right path for me; instead, i cared way too much about what others thought of me and was too impressionable because i wanted to impress.

the problem with wanting to impress people is that you start optimizing for their metrics instead of your own. you start measuring success by whether you got invited to the right slack channels, whether you’re working on problems that sound important when you describe them at parties, whether the people you admire would be proud of you.

and maybe that would be fine if those external metrics aligned perfectly with what actually matters to you. but they rarely do.

to be clear, i’m not saying my work in ai safety is meaningless or that i should abandon it entirely. i’m saying that i need to disentangle my genuine interest in the field from my desire to be seen as impressive by people i respect, and then conclusively determine whether or not my genuine interest alone is enough to merit switching careers—or if i should give myself a little more room to explore different paths rather than diving all in to this one.

being impressionable isn’t inherently bad. empathy, openness to new ideas, willingness to update your beliefs—these are valuable traits. the problem only arises when you become so focused on impressing others that you lose sight of what you actually think and care about.

honestly, i think i may have gotten very lucky with ai safety in that it actually does align with what i want to do—the work genuinely matters to me, independent of whether it impresses anyone. but recognizing that luck doesn’t erase the fact that my initial decision-making process was flawed. i stumbled into the right answer for some of the wrong reasons; i tried on the right hat while trying to impress the wrong guy.

and that’s what i’m trying to change.

training myself to ask “how much does this truly matter to me?” instead of just “does this make me look good to people whose opinions i value?” is still an ongoing journey—a work in progress. some days i catch myself optimizing for external validation before i even realize i’m doing it. other days i’m able to pause and actually consider what i want, separate from what would make me look impressive.

it’s uncomfortable work, this untangling of genuine interest from the desire for approval. because sometimes they’re so wrapped up together that i can’t tell where one ends and the other begins.

but i’m doing my best, and learning, slowly, to care less about being impressive and more about being honest with myself—which means living in alignment with my own values and beliefs and desires.

because in wanting to be impressive, you often unintentionally become impressionable.

and i don’t want to live my one wild and precious life defining success by the metrics of others.