There are two ways to feel safe: you can be protected by someone else, or you can protect yourself. With the former, it is okay to walk around with blurry eyes. It is okay, maybe even better for your peace of mind, to flinch away from the threat in your periphery. When you are protected, you can afford to avoid reality, because someone else will deal with it for you.

With the latter, you must see the world clearly. You can try to keep your vision blurred, but you will not feel safe if you do so. Though you may not see the knife flashing in the alley, you know that it could exist and will catch you unaware. If you live in a world like ours, where there are no gods and no adults, all you can do is try to face reality — or at least glimpse its contours — as best you can.

* * *

With AI, it seems like we might be hurtling towards a precarious future, and I don’t know what I can do about it. The experts disagree on important things. There is no one who will hold me and tell me it will be okay. There is no one I feel safe deferring to.

* * *

Plan A, a detailed recommendation for making AI go well, was released this past week. As I was talking to a friend about it, I surfaced lots of questions that I didn’t know I had — “How do we know the US will stay in power post-deal?” I realized I was instinctively turning towards Plan A for comfort, reassuring myself, “At least Thomas Larsen has a plan.” But Thomas Larsen is still human, albeit a smart one. Although it is probably one of the most fleshed-out visions of the future we have right now, Plan A is an optimistic forecast rather than a perfect playbook. It is helpful to understand why AIFP (the team that published Plan A) thinks what it does, but it is also necessary to think about the ways it could be wrong.

* * *

I remember feeling very confused two weeks ago, when I first arrived in Berkeley. I felt like I had murky models of the world and no positive vision for the future. In the time since, that feeling of confusion has dissipated, though the source of the confusion — my lack of knowledge — remains. Where did my confusion go?

I don’t want to be soothed by false comfort. I worry that I’m lulling myself into a false sense of security, that I’m conflating “conversations during lunch about Plan A” and “nice list of well-scoped tasks” with actually thinking about and working on the problem. I want to notice my confusion, because no one else will notice it for me.

Life in Berkeley is good. It’s almost too good. I’ve spent an unusual amount of time in Constellation (the office I work out of) — something on the order of 12 hours a day for 7 days in a row — and I’ve done so enthusiastically.1 I am served three meals a day, an abundance of snacks, and an influx of interesting conversations. I feel a little suspicious, honestly. It seems to me that if I am actually working on the most important problem in the world, if I am actually trying to do the most good I can, there should be some amount of personal sacrifice. I don’t feel like I’m sacrificing much of anything to be here.

I am getting a little sick of spending so much time in one place. In Berkeley, each day repeats the last. I recently visited an ice cream store named “Endless Summer Sweets”. At first, I was confused by the name (“What does it do for the other three seasons?”), until I remembered that Berkeley really does experience an endless summer. Constellation feels similarly endless — herein is all the kombucha and fresh fruit I could ever want. My life is so self-contained that it sometimes doesn’t feel real.2

I am worried my beliefs will drift away from reality in Berkeley. This concern is not baseless; I already sense my mannerisms changing. Each day, I become more like the people around me. There is a particular way EAs speak, especially EAs in AI safety, littered with hedging and in-group references. I’ve picked up these EA-isms, especially the performative uncertainty — my sentences start with “plausibly” and end with “or something”. I find myself joking about how sunscreen doesn’t matter, because AI will either solve skin cancer or kill us first, and I wonder if this is healthy. I worry that I’m losing my voice — I think EA-speak is bad, and that there is immense value in being able to communicate in-group ideas to the outside world. That’s part of the goal with these blog posts, to remind myself what I sound like.

I came to Berkeley because I wanted to interface with the real world, but it feels a bit like I’m in high school again. All my needs are met. I hope the sunniness of being here doesn’t distract me from my real goal of doing good. I worry that there will come a time when the most valuable thing for me to do will be to do unpleasant grunt work in the outside world, and I will find some way to justify staying in the pleasant confines of Constellation.

Though I am cautious about losing myself, I am nonetheless excited for the coming weeks. For one, the people I’m working with seem wise and capable, and I expect to learn a lot from them. For another, I love being in this community — of truth-seekers, of hard workers, of people who are earnestly trying to make the world better. I love that I can debate moral philosophy and make Hamilton puns in the same conversation, that I can reference Schelling points or fuzzies and people just know what I mean. At Princeton, I feel like I have not found “my people”. Being here might be the closest I’ve gotten to that feeling.

* * *

Why do you care about AI safety?

I’m not sure I do, actually. Or at least, care feels like the wrong word to use here.

What do you mean?

I don’t think I’ve gotten to the point of caring, I guess? It’s like having a crush — caring implies a fondness, a familiarity. I think what I feel towards AI safety is better described as interest, intrigue.

Okay, why are you interested then?

I guess it seems important to have opinions about the future of AI, or at least to have thought about it. I remember going to Action Potential and feeling like I’d woken up — like I’d not been awake before. People spoke with this depth of thought that I’d not accessed before, this whole new consistent way of seeing the world. It is a weird thing to hang out with a guy and to be asked if I want to have kids before or after the Singularity. I’m convinced that the future of AI is a thing that I should think seriously about, but I’m not sure where my thoughts will end up.

And have you thought seriously about it? It doesn’t feel like you have.

Maybe? I don’t know, I think I still have really murky models of the world. I feel like there is nothing stopping us from building AI that can do everything humans can and having those AIs improve themselves, in approximately that order. But I think I still have a pretty poor understanding of what is actually important to work on. And I don’t have a good sense of how we can coherently get to a safe future. For all the LessWrong posts and 80k podcasts, I still feel really confused a lot of the time in a way that is difficult to dissolve.

I guess this is one thing I’m hoping for from the summer. I want to have better world models; I want to understand what a coherent good future looks like; I want to concretely know my cruxes. Sometimes, I talk to AI safety people and I have this feeling of, Wow, that just makes sense. I want to have more of that feeling; I want to cut through the mud with clearer eyes. I cannot eliminate uncertainty, but I can reduce it.

I don’t get it — why do you feel like it’s important to have opinions about AI? Or at least, why can’t you just defer your opinions to experts?

Don’t you see?? The experts all disagree — they’re also confused, just in more specific ways than I am. And I am scared because the thing they seem to agree on is that AI will be transformative, whether it kills us all or makes us immortal. Obviously, there is a self-selection thing going on here — I should think about this another time.

It feels like everything I love is downstream of powerful AI. If I wanted to conquer global poverty, it seems like the best way to do this is plausibly to build powerful aligned AI. And even if I wanted simpler things — to fall in love, to have friends and family, to have improved the lives of the people around me — it seems likely that powerful misaligned AI will deprive me of this pastoral ideal.

At the same time, I don’t think I “feel the AGI”. I don’t think I’ve internalized how quickly models are improving. When I think about powerful AI, there is white noise in my head. I follow a pointer to null space; I get an error back. One of these days, before I go to Generator, I will sit down for 10 minutes and allow myself to actually imagine reality as it could be.

* * *

The biggest thing I want out of Generator is to have a better model of the world. The second biggest thing is to have made something I am proud of and to have learned in a relevant way. The third biggest thing is to have made good friends.

The biggest pitfall I want to avoid over the summer is getting distracted. I think AI safety people are really, really cool. I don’t want to become so enamoured by the social aspect that I never actually interrogate my beliefs, and I fall into a life that is not my own. I think this is what happened to me last summer, where I didn’t actually grow intellectually because I enjoyed the social aspects so much.

Anyway, I am so excited to be in the Bay. See you there :).

1. Kiss someone

2. Go skydiving

3. Learn to do a pull-up

4. Go to a haunted house

5. Make a short film and submit it to a contest

6. Go on a week-long silent meditation retreat

7. Write one work of fiction I’m really proud of

8. Learn how to spin a pen

9. Design a cool t-shirt for a club

10. Go vegan for a week

11. Sleep over in my sister’s dorm

12. Publish 1 paper

13. Run a half marathon

14. Develop one physically challenging hobby (e.g. weightlifting, rock climbing, dancing)

15. Develop an artistic hobby (e.g. figure drawing, painting, photography)

16. Quit doomscrolling on any platform

17. Write something that changes someone’s life

18. Fall in love

19. Play Outer Wilds

20. Read a book that makes me cry

I will update you all on my progress this time next year!

At the time, you seemed so large in my head, learned in subjects I’d never heard of. You could talk endlessly about multivariable optimization, or John Coltrane, or Augustine’s pietas. And it wasn’t just academic subjects — you knew so much about Princeton, about the world. We stayed up until 5 a.m. talking about the startup culture on campus; we categorized our crushes over scoops of Bent Spoon; you talked about why you chose your pronouns over free cups of matcha. I admired the easy way you walked, like you were certain of yourself and your knowledge. Every other sentence of yours felt like it was something to be cherished. There was a world in your head which I wanted to absorb.

At first, I was hesitant to reach out to you. I felt like you had everything to offer me, and I had nothing to offer you. It surprised me when you invited me to the reading group, or to get boba together. And it surprised me when you told me you liked talking to me, that you were curious about what I had to say. Now that I’m closer to being in your shoes though, I am no longer surprised.

I’m not sure you know how formative you were to my Princeton experience. When I think of Fine Hall, I think of you and the other math seniors in the common room talking over each other in that intense way you all do. It is difficult to imagine what it will be like to step into Fine and see a new set of faces. When I think of Spelman, I think of oatmeal breakfasts in your dorm and debates about God. When I think of Frist, I think of Adam Sandler and freestyle rap battles. You taught me so much about the sort of person I want to be.

I am not sad because I think I will never see you again. I am sad only because we will see each other less frequently. But what I feel most strongly is a sense of anticipation. I feel like I am watching myself grow up.

In the fall, I will be the same age you were when we first met. I imagine that I will have my own freshmen who look up to me in the same way I looked up to you, with shining eyes. It’s weird — I don’t feel ready to be that kind of mentor yet. Is this how you felt back then?

I’m excited for the summer, because I feel like I’ll grow a lot then. I feel like I am becoming at a rapid rate, like I am watching myself morph before my own eyes. You’ve left me with big shoes to fill. I hope I will have big enough feet.

Love,

Eliana

1. Overview (feelings beforehand, logistics)

2. Reflections (feelings afterwards, things I would do differently)

Overview

This past weekend, I attended my first EAG event, the 2026 EAGxDC, which is just a conference where you can meet other cool EA-minded people. Here’s a more detailed event guide if you want to get a sense of the logistics. I arrived in DC on Saturday at 11 am and left on Sunday at 5 pm. My goal for the conference was to “learn as much about AI policy as I can,” since I felt like I knew almost nothing about that side of things and wanted more context.

The common advice they give is to book lots of one-on-ones, so that’s what I did. I was actually being pretty indecisive about whether I wanted to go or not, since I had finals soon. Even though I only started to book one-on-ones the Wednesday before the event, there were lots of interesting people with availability in their schedules, contrary to my expectations. Maybe this was because it was an EAGx event, not the (maybe?) more popular EAG, and because I have a wider range of people who I would find interesting and valuable to talk to, since I am a university student and have less experience than most people there.

Ultimately, I didn’t go to any talks at all and ended up having 17 half-hour one-on-ones: 9 on Sunday and 8 on Saturday. I also had some other random unscheduled good conversations. One might be afraid that this event would have an icky, networking sort of feeling, each conversation cursed with Aboutness. But it wasn’t like that. Though the point of the event was to network, it was networking in the sense of, “Let’s talk to each other because this conversation could be interesting!” My conversations did not at all feel superficial. It felt like we both cared about doing good and helping each other figure out what that meant for ourselves.

I have an inherently high social battery, but also think the quality of my conversations just made me endlessly excited to talk to more people. It felt like everybody in the room knew something I didn’t and was eager to share it with me. Though be warned: I was definitely losing my voice towards the end of both days.

Reflections

I’m really glad I decided to go. I had a lot of fun, had many great conversations, and feel happy about the people I met there. I think one of the biggest differences between me and my friends who are not into EA is that I feel a warm fuzziness when I think of EA people, and they feel a sense of distance or repulsion. I also used to feel that distance — I remember being vaguely embarrassed when I told friends I was going to the Northeast EA retreat. But going to these events makes me more inclined to associate myself with the eager, thoughtful, kind people I’ve met. I’m excited to apply for more EAGs!

Things I would do differently:

• Be more thoughtful about my one-on-ones.

  • I kind of just booked one-on-ones with either people who I knew of and wanted to talk to, or people who seemed policy-related and put effort into their profile. I didn’t have anything specific I wanted to talk about for most of the meetings; I just asked about why they worked on what they did, and what they wish more people were working on.

    • Just like the purpose of self-help books is not specific pieces of advice, but rather to get a sense of how this unique person thinks about the world, I was trying to get a sense of how these people thought about what was important.

    • The conversations might have been more productive if I came with some more specific questions in advance, but I’m pretty happy with how they went.

• Book more one-on-ones with people my age.

  • I originally booked one-on-ones just with people more senior than me, since I felt like their experience would be more valuable, but I’m really happy a couple other university organizers reached out to talk to me. I forgot that the conference was also a place I could make friends. I ended up getting dinner on Saturday with a group of organizers from a bunch of schools in the Virginia/Maryland area, which was fun. I tried paneer pizza, a goated dish.

  • I bumped into another university student in the hallway randomly and we ended up bonding over the blogs we read, which was such a lovely experience.

• Leave some unstructured time on the last day to mingle.

  • There were some friends I really wish I could’ve talked to more, and some other people I randomly met that I would have loved to talk to more.

• Not book anything else that weekend

  • I had an interview Saturday night, which was bad because I was tired of talking after talking for the entire day and because I had to leave dinner earlier than I wanted to.

  • I would also have loved to stay to hang out with people on Sunday evening, but had to get back to Jersey.

Takeaways + next steps:

• I have a bunch of notes on my conversations in a physical notebook that I’m going to type up sometime in the near future.

• I got lots and lots of recommendations for things to read! :D

• I’m excited to keep writing about things in my own voice. Talking to someone made me realize that EA-specific lingo (e.g. “counterfactual impact”) can push people away, and that there’s great value in talking about EA concepts in my own words (e.g. “what else could you do instead of this?”)

• There are some people I want to follow-up with, or didn’t get a chance to talk to over the weekend.

• And there are lots of things I want to think more about. Here are some unstructured thoughts off the top of my head:

  • Still feel like comms is seriously neglected — I want to think more about what is important to communicate and how.

  • Still don’t really get the case for caring about concentration of power + it seems like concentration of power includes many cases, and I don’t really understand the difference between these cases. I want to learn more about this.

  • Fuzzy on the intersection of technical and policy work — are we just cooked if technical alignment isn’t tractable? What policy measures can we take to mitigate this?

  • One conversation made me feel like democracy is much more endangered than I thought it was; I want to learn more about that.

  • Wondering to what extent it’s possible to accept the AI as Normal Technology thesis that AI will take a long time to integrate into the economy and still be concerned about x-risk.

• What you write about will change. In particular, your writing might turn personal. I originally started this challenge as a way for me to develop my thoughts about EA and AI safety ideas, but my posts quickly deviated. I think that because I am the subject I know best, it is easiest to write about myself.

• How you write is hard to anticipate. Since this was a daily challenge, I gave myself advance permission to work on longer content in short chunks. I never did this. I think the goal was misguided, because long-form content is not made sequentially, and I felt uncomfortable posting incomplete parts of a whole.

• Know what you will give up to write. Writing takes so long. Like, so long. Wow. Unexpectedly, I ended up giving up a lot of the time I spent in classes. I was a truant for the past month. I spent at least a couple hours every day on blogging-related activities — even the day I just posted seven pictures of my cat. Some posts took the majority of the day, like “The practicality of political violence”. When I did this challenge with poetry last year, it was so much easier, because with poetry, the burden of explanation is not on the writer.

• Don’t forget to take care of your emotions. Much of the time I would have spent journaling went to writing blog posts, because it sometimes felt like I was wasting time writing just for myself. The hours in day felt short, and I became reluctant to follow complex thoughts to their end. Continuing this behavior over a long period of time might hinder my emotional processing.

• Don’t forget to live in the real world. I became kind of a hermit the past month. I spent a lot of time sitting in my room, reading books and blog posts. But inevitably, the best things I wrote were born of interaction with the real world. There is a lot of blogging advice out there, and much of it recommends paying attention to the sort of things you repeat to your friends. To do this, you have to hang out with your friends.

• Set a deadline for yourself. My sleep schedule got fucked up. I think something like the Inkhaven 11:59 pm deadline would have served me well, instead of just “before I go to sleep.”

• It’s surprisingly hard to forget your audience. I began this exercise purportedly for myself, with public posting as a social accountability thing. But it’s so exciting when people like your posts and follow you and leave comments! Also, when I was writing more personal things, I couldn’t help but think about who would be reading what I wrote. Especially when people you think are cool follow you, it’s hard to be okay with presenting yourself as you are, in all your inevitable mistakes.

• The personal and public will blur, and you will learn to be honest. The conversations and experiences you have will influence what you write, especially when you’re trying to write something every day. Inevitably, I wrote things about people who could see this blog. On occasion, those things were critical. I don’t think I hurt people’s feelings, but I was sometimes worried I would. The only way to not hurt people’s feelings is to be straight-up with them, instead of anonymizing them and hoping they don’t figure out who the person you’re writing about is.

• It’s okay to shitpost. Doing anything daily feels pretty challenging to me. Writing a blog post daily is especially so. Two days ago, I couldn’t think of anything I wanted to write about, hence the aforementioned cat pictures, inspired by Katja Grace’s preeminent Elephant seal 2. Many of the things I wrote felt half-baked, most recently like the post against face-saving lies. But it’s okay, I tell myself. I’m glad I wrote something at all.

Reading the above makes me feel dissatisfied. I am not some machine that spits out well-formatted, neat advice. I am a real human who felt many different things over the course of the month. So, here is the other part of the post.

There were times when I felt really burnt out. When I look back at my journal, I see that on April 14th, I wrote, “Some days, I’m so tired. Some days, I have experienced something heartbreaking, or I’m bone-tired, and I’m not sure that the things I write are even coherent.” On April 24th, I wrote, “getting anxious, feels like i have nothing to say, just dumping soul.” Yesterday, I was sick and miserable and really didn’t want to write anything. I sat at my desk, staring at the wall. I wrote, “tired tired burnt out.”

I was often frustrated with myself and my writing. The feeling of producing eliana slop was persistent. Towards the end, I didn’t feel like I had anything worth saying most days. The feeling reminds me of talking to my high-school ex — our conversations mostly felt like me yapping to fill the silence. It feels like I am not an expert in anything besides myself, and even then, there is only so much I can say about myself.

Also, writing is hard. It takes me a while. I admire Katja Grace for how consistently she writes. I’m also not very proud of most of the stuff I write. I admire (and envy) some of the folks at Inkhaven (which I learned was a thing in early April) because they’re writing some amazing things on a daily basis. Writing is hard, and I have so much room to get better at it, and so much more life to live to have more to say.

Am I glad I did this? Yes! There were times when I was proud of what I wrote, when I felt like I’d reached some clarity I lacked before. Especially in the beginning, I felt like I was overflowing with ideas, and I felt like this exercise reawakened a love for argumentative writing. It was like being in AP Lang again (in the best way possible).

Would I do this again? Maybe. If I did it again, I would wait until I had a backlog of good post ideas I was excited to write about. I think some people find the pressure of having to write something every day helpful for being a better blogger, but I personally just feel tired.

My favorite posts to write were the lowercase ones, inspired by Celeste and vincent huang. They felt closer to the sort of thing I write in my journal, more like my true voice. I love personal writing in general — maybe you could guess that from all the posts about being known. I haven’t ended the month with a better sense of my niche, or what I’m good at writing, or with any abnormally popular posts. Most of the stats for my posts were similar, so I’m not really sure what people like reading. In the future, I hope to write some longer work and some fiction.

When I started off the month, I’d just returned from the AISST/MAIA workshop and Action Potential the week before that. I have to thank Carolanne for unknowingly inspiring me to write blog posts. I don’t know that so much has changed since then. The month has passed by very quickly. I know much more than I did 30 days ago, and still almost nothing at all. I got somewhat addicted to Substack, and ended up on some new corners of the internet. I don’t know when I’ll post next, but see you sometime in the near future! And thanks for sticking along.

In complex environments, it makes sense to maintain a non-trivial probability of exploration. By complex environments, I mean environments in which it’s difficult to predict the downstream effects of actions. I think we live in a complex environment. My closest friends are people I met pretty randomly, without initial intention; many heuristics that make my life better are ones I stumbled upon randomly; the topics I am most interested in are ones I learned about randomly.2 Given that we live in a complex environment, it is good for us to explore regularly.

However, I know a handful of people who seem to only exploit. I’m talking about the productivity-maxxers — the friends who (think they) optimize every moment of their life, who structure every minute of their calendar, who reject social invitations because they just know their time would be better spent on something else. There is a kind of hubris to this attitude. These people might be optimizing on the knowns, but there is a vast space of unknowns beyond imagination.3 True optimization includes unstructured time — time spent on exploration, time spent being bored.

I find that exploration disproportionately leads to really really great outcomes. This is because unlike RL, humans are not randomly choosing among all possible directions to explore. Human exploration is led by one’s curiosity and instincts. When you solely exploit, you deprive yourself of the opportunity to follow your heart, to listen to the iron string inside. In this way, excessive exploitation suppresses the self.

Usually when people talk about someone who optimizes their life, they mean someone who is purely exploiting, who has become a vessel for the hacks, heuristics, and frameworks they use to structure their life. This person is too certain of their likes and dislikes to make first-impression-defying friendships; too sure they want to study CS to take any classes in any unrelated subject; too busy arbitraging the dating market to fall in love.

• Woke up late and speed-walked to a lunch with Prof. David Walker and Prof. Arvind Narayanan, the latter of whom is half of AI as a Normal Technology. Arvind talked about four reasons he thinks the argument that AI will suddenly replace jobs is flawed.1 I think Arvind’s ideas are interesting and I emailed him to ask for some readings he mentioned during the talk. Arvind is also an um-less man — it takes a lot to be an um-less man. Some questions and thoughts I had after the talk:

  • It seems like one does not have to believe that AI will be transformatively incorporated into the economy in order to believe it could be catastrophic. In other words, is it possible to agree with Arvind’s claims in the talk — that AI will not quickly replace lots of jobs — and still be very concerned about x-risk?

  • I’m concerned about centralization of power. Prof. Walker talked about how transformation is not politically incentivized, because it’s hugely unpopular to make 10% of people suddenly lose their jobs overnight. But if the people’s opinion doesn’t matter, then the space of possible actions suddenly includes ones that are good for people at the top and bad for everyone else. This reinforces my belief that even outside of x-risk reasons, AI safety comms is important and I want people to be more informed.

  • I’m also concerned about inequality. AI is being applied more in sectors that can afford to keep up with new technologies. It seems the people most ahead are accelerating fastest. This is not a very coherent thought, just a bad vibe at the moment.

  • Every time the ATM/bank teller story is told, I am reminded of this blog post: “Why ATMs didn’t kill bank teller jobs, but the iPhone did”. I will read it in the ambiguous “eventually.” I want to understand when automation becomes faster than augmentation.

  • Later in the day, someone made a group chat with a bunch of people who attended the talk. Very cool. Turns out you can just do things. I’m always scared to make group chats.

• I got boba with some people who lingered after the talk. Many things were discussed.

  • I really really think that some big AI safety folks should try harder to appear like “normal people.” I can believe the argument that the productivity losses can outweigh the PR benefits of some quality-of-life changes, but would it really hurt so much to lose the fedora? Relatedly, it seems like for some people losing polyamory would be a terrible quality-of-life change, but I am a little concerned that polyamory is not a good look, especially when it comes to things like name-drops in this post by Aella (which to be fair is from 2021).

  • I was reminded that I want to sign up for a silent meditation retreat. I don’t think I’ve been silent for a day in my life.

• Some boba people and I headed to the PAIA reading group. The meeting was purportedly to discuss “Current AIs seem pretty misaligned to me”, but was really just a yap session. My brain was getting tired by this point, so the meeting is a little hazy in memory.

  • Will maybe give X another chance, since I want to be in the loop on tech/AI news. I was originally deterred because my signal-to-noise ratio is rather low, but was convinced that I should interact a bit and give the algorithm time to adapt.

  • Realized I know almost nothing about the state of computer vision.

  • I also know nothing about JEPA.

• After leaving the reading group slightly early because I was getting tired, I ate dinner in my room. While eating, I read some materials to prepare for Week 1 of BlueDot’s AGI Strategy Course. Might I say that, with remarkable turnaround time, they assigned me a cohort a mere 12 hours after I submitted my application.

  • Preparing for Launch: I was a little surprised by the principle of, “We should neglect the costs of stalled progress.” I have not thought carefully about the costs of something like a global 10-year pause on AI R&D, because it seems obvious that the potential stakes of development outweigh the costs of a pause. This may not be true though, and is something I should be more critical about.

  • Chapter 1 of Utopia For Realists by Rutger Bregman (the Moral Ambition guy): I really liked this. I feel a little dissatisfied all the time, and some of the things Bregman says resonate with me. I want to believe in utopias again, to have hope for the future, to appreciate how far we’ve come and how far we can go. Would highly recommend reading the first chapter!

  • I’m excited for this mini-course, because it seems like it’ll be a great way to build AI safety context and will be useful for whatever I do over the summer.

• After this, I did a Zoom to talk about summer stuff, went on a walk with my roommate, got free cupcakes, and rotted away a little bit. Now, I’m here writing this post.

  • I really don’t like the feeling of churning out slop. Recently, when I sit down to write blog posts, I get this feeling strongly. :/

    • Somehow, I am both consuming too much and producing too much. There is some third thing I can do, which is just exist. I don’t feel like I’m doing enough of that these days.

  • Here is a haiku I wrote last spring about redbuds (see photo below). It is the first entry in my Google Docs journal.

    Text within this block will maintain its original spacing when published

    Baubles of blossoms
    The redbud holds her neighbors
    Small spheres of embrace

“Just say you had a family emergency,” says one friend.

“Or say you were sick,” says the other.

“No! Why would I do that?” I say, and press send on an email that contains no excuses. But I know why I would want to do that — why I would want to tell small lies to save face. In this blog post, I want to explain why I don’t tell face-saving lies (which I will call FaSLs1, because I like shortening things).

I think the main reason people FaSL2 is because they imagine the other person’s opinion of them will be preserved. This is not applicable in settings with repeated interactions. You can only tell your professor that you were sick a few times before it’s obvious that you’re lying, and being caught lying is a worse look than telling the unflattering truth. This is also not applicable in settings where people can share information. After all, if my dog died when I missed my chemistry midterm and my dog died when I bombed my biology presentation, I better hope my chemistry and biology profs never get to talking about me. In general, in situations where one person has multiple data points about your flaky behavior, FaSLs are unwise, even if all you care about is your reputation, because they don’t work in those types of situations.

But what about in one-off settings, where the other person will almost certainly not catch you in a lie? Will their opinion of you be successfully preserved? I’m inclined to think that yes, their opinion of you will usually be preserved if your lie is convincing enough, since I think most people try to interpret others’ actions in the best light. If you told me you missed our call because you had a family emergency, I would trust your word, even though I might be a little skeptical. So the lie works for its intended purpose.

Then, does that mean it’s okay to FaSL? The answer is still no, but for more debatable reasons. Staying in the framework of “I want their opinion of me to be preserved,” it seems to me that their opinion of you will be preserved if you tell the truth. I frankly don’t care that much about why you missed the meeting or didn’t do the work you promised you would. Regardless of whether you tell me it was because you were sick or because you just forgot, I’ll be a little irritated and a little forgiving either way, because the result is the same. To me, both reasons — illness and small irresponsibility — are understandable, relatable reasons for being unreliable.

Thus far, it seems like I’m arguing that truth-telling is about as good as FaSLing. I want to convince you that it’s better, and that the betterness of truth often outweighs the potential reputational harm of truth. I see FaSLs as a specific instance of a broader problem, in which we focus on the apparent result rather than the process. Many things I did in high school, I did for the sake of a line on my college application, not out of any great passion for the pursuit. I had the appearance of passion without actual passion. In the same way, FaSLs separate the apparent you from the real you, the appearance of blamelessness without actual blamelessness. I’ve found focusing on appearances over substance to be a tiring way of life.3 If you feel the same way, if you want to present yourself authentically, in all your mistakes and triumphs, I hope you don’t FaSL.4

The other subtle reason people FaSL, besides to preserve their image to others, is to preserve their image to themselves. I still get tempted to FaSL, especially when I know I’ve messed up big time and I can’t bear to speak my own mistake. But flinching away from the truth doesn’t make it less true; painting over the truth doesn’t mean it’s gone. In order to address the root of our unreliabilities, we have to be able to look at ourselves clearly, critically, and generously.

There are caveats to this post. For one, it’s not helpful to actually flake on responsibility, acknowledge your flakiness, and do nothing to change. Self-awareness isn’t a solution. For another, I don’t think you should (or can) tell the complete truth all the time. However, I think omission is more permissible and often appropriate. Depending on how chill your professor is, it’s probably not smart to tell them, “I’m skipping class because I think there are better ways I can spend my time.” But a simple, “I won’t be able to make it to class today” suffices.

One hates itself. One loves itself. They have not met yet.

The self-hating wolf (who we will call Wolf H) has good reason to hate himself. He cannot seem to do anything right. In his social life, he doesn’t feel like he belongs to the pack. He thinks he’s kind of awkward, and he never knows what to say. Oftentimes, he lies awake in bed replaying all the embarrassing things he said that day. He wants to scream. In his academics, Wolf H is never at the top of the pack. He’s not even sure he wants to be. He goes to wolf school and procrastinates all of his assignments, then hates himself for procrastinating, then hates himself for hating himself. He doomscrolls for hours, all the while knowing he should be working on his pawlitics homework. And forget about his romantic life. The last relationship he was in didn’t work out, but his ex seems like she’s thriving. Wolf H wonders if it’s a him problem, if he’s fundamentally unlovable.

The self-loving wolf (who we will call Wolf L), like Wolf H, also makes mistakes. No wolf is perfect. But Wolf L knows that mistakes are a part of life, and all his favorite wolves made mistakes at some point. He forgives himself for the social blunders and the wasted time, because he is proud of the wolf he is becoming, and he knows that he is trying his best. Wolf L feels an intrinsic fondness towards the other wolves he meets, because he knows that it is hard being a wolf, and that everyone is trying their best. He does not know why he loves himself, but he knows that he does and he tries to protect that impulse.

One day, Wolf H and Wolf L meet. Slowly, they become best friends.

Now, when Wolf H says things like, “I hate myself” or “Why did I do that?”, Wolf L is there to support him. “Everyone makes mistakes,” says Wolf L. “I still love you,” says Wolf L, “and I want you to love yourself like I love you.” Sometimes Wolf L isn’t there, and Wolf H still finds himself spiraling. Eventually though, Wolf L always comes back around to comfort Wolf H, and to remind Wolf H that he is still valued, and that the things he hates about himself are normal wolf qualities.

Inside you are two wolves. One hates itself. One loves itself. Have they met yet?

After The Artist Never Dies by Harry Moss

* * *

In a world where all the most beautiful art is AI-generated, I would still write. And I would still paint and I would still sing, even though I would not be close to the worst of the best. What I value in art doesn’t exist in the final canvas. It exists in my knowledge of what came before.

Last semester, I read Klara and the Sun by Kazuo Ishiguro. In the novel, Klara, a humanoid robot, is purchased to be a companion for Josie, a sickly human girl. As the story unfolds, we realize that Josie will die soon, and that Josie’s mom intends for Klara to take on Josie’s appearance and mannerisms and replace Josie after death. Klara has exceptional imitative abilities, so the difference would be visually imperceptible.

But Klara knew she could not succeed in this imitation. While talking to Josie’s dad about Mr. Capaldi, the man designing Klara’s new Josie-like body, she says,

Mr. Capaldi believed there was nothing special inside Josie that couldn’t be continued. He told the Mother he’d searched and searched and found nothing like that. But I believe now he was searching in the wrong place. There was something very special, but it wasn’t inside Josie. It was inside those who loved her.

The thing that makes Josie irreplaceable is not anything about Josie, but rather that the people around her will know that Klara is not Josie. Josie’s mom loves Josie not because of any particular quality about Josie, or even any combination of qualities, but simply for the fact that Josie is Josie.

Perhaps AI art can replicate the art that I love for aesthetic purposes alone, the art I glimpse at in museums. But for the art I love most, I love it not because it is beautiful or moving or intricate, but because it is the reflection of a soul. Even if there comes a time when my favorite novels are written by AI and when my favorite songs are composed by AI and I can only tell the difference between human and AI because the latter is so much better, I will value human art. I will never tire of writing poems to my friends, or if I do then I’ll start sending paintings, or writing songs. And I will never tire of receiving the novels and sketches and dances they impart upon me.

Even in a world which blurs the difference between what is human-made and what is AI-made, even if I could not tell the difference by looking at the two, I would value the one with a lived experience behind it.

This is not to say that people shouldn’t be concerned about AI art. But I want to explain why I am not concerned about AI art, because AI art doesn’t yet replace the particular invisible thing I care about. Maybe there will even come a day when we recognize the other art — the AI art — as the image of another kind of soul. On that day, I will not feel threatened. I will feel glad to understand another.

It is always a feeling I want. I want to have done the thing; I care little for actually doing the thing. I want to have found my people; I have no patience for the process of finding my people. I want to have lived a fulfilling life; I am impatient when it comes to actually living such a life.

And I have wanted this feeling before. I think I have never felt true belonging, not in the pure way that I feel it in my imagination. I have never felt comfortable, not really. But I have hoped for spaces that could make me feel this way. And they have never met my expectations. I don’t think they can. Reality can only hope to converge upon the impossibility in my head.

Someone recently described the Bay Area EA vibe to me as “manic,” and I agreed. I get that sense from this side of Substack alone. But I am not off-put. I am almost morbidly curious; I am the patient in a room, choosing shock over boredom. It is like wanting to live among evangelists — maybe I hope their conviction will infect me. I don’t know what I want. I only know that I am curious about this life. I feel like I have never been alive, and I want to feel the sting of breath.

I am letting the idea balloon in my head. When I think about what I want to do over the summer, the prospect of being in California disproportionately outweighs other considerations. This seems unwise. Probably, the people in California are mostly like the people I meet everywhere else, and the everyday does not shine like it does in my head. I am setting myself up for disappointment either way — whether I end up there or not.

I don’t know how to end this post. I feel like I have not exhausted the desperation with which I want to be in California, with the people who have a shelf of productivity supplements and talk in Yudkowsky-speak and write about dating arbitrage, people whose beliefs I often find abhorrent and compelling at the same time. But this is a people in my imagination. I am looking for salvation, but there is no Promised Land waiting for me.

Inspired by Annabel Lee

On colder nights, I oft recall
    that town I left behind,
with empty streets and limestone walls
    and the boy with the burning mind;
the boy I loved, who loved me too,
    yet thought me not his kind.

I was a child and he was a child,
    in that town I left behind,
but we saw the world with clearer eyes,
    him and I, our minds—
at least, he did, with his piercing gaze,
    though I was not his kind.

Then it was that autumn night
    in that town I left behind,
when I told his boy I loved him so,
    the boy with the lovely mind,
but his love for me burned not like mine,
    though the words were not unkind;
The moon that night was thin and white
    in the town I left behind.

I thought my soul could have been seen,
    and yet I was so blind—
    what did I want to find?—
and neither the daylight shining lean
    not the nights I left behind
warmed my heart like the fiery words
    of the boy with the burning mind;

for there is no fall I do not recall
    the autumn he wasn’t mine,
and my heart’s grey embers will always remember
    falling for his mind;
and so, all these nights cold, the thoughts I unfold
    are of the boy—how I thought we’d grow old
    in the town I left behind
    with the love that isn’t mine.

The sentiment above has helped me have better relationships with people I interact with on a regular basis, like my friends. Only repeated interactions provide the necessary feedback loops to actually improve my models of specific people. However, the idea of mental models has also improved the way I relate to strangers.

In particular, changing the way I see myself has changed the way I see others. Because my default model of strangers is myself (as yours probably also is), my model of myself affects the default assumptions I make about others. Having a more useful model of myself improved my interactions in two ways, (1) by changing how I perceived other people’s actions and (2) by changing how I thought other people perceived my actions.

Let me give a concrete example. When I was quite young, I latched onto the idea that the best conversationalists are actually the best listeners, probably after reading Dale Carnegie’s How to Win Friends and Influence People. I’m not disputing the validity of the idea, but I think I took it too far. However, in my desire to be the best conversationalist, I learned to minimize my own opinions. In conversations, I almost exclusively asked questions of the other person, even with close friends, even if I had something I desperately wanted to share. When I was asked about myself, I felt a little guilty, like I was wasting people’s time. Because I judged myself for talking, I assumed others judged me as well; because I viewed asking questions about others as an obvious act of consideration, I assumed people who did not ask me about myself were simply inconsiderate. Ultimately, my model of myself colored my model of other people, leading me to resent other people for faux pas they were unaware of.

When I tried to talk about myself more freely, I started to enjoy conversations a lot more. Because I stopping judging myself for talking, I stopped judging other people for talking about themselves. Likewise, because I stopped judging other people for talking about themselves, I stopped assuming they would judge me. Changing my perception of my own actions changed my assumptions about how other people would perceive my actions, which allowed me to act more freely and test whether my assumptions were correct. If I had talked about myself and the world reacted poorly, I could at least have more information about my assumption. Not talking about myself, though, caught me in a pernicious loop, where my assumptions could not be disproven.

Here’s another example which also has to do with judgement. Suppose you have very low self-confidence. In particular, let’s say you’re self-conscious about your weight — you think you’re too skinny. When you go out into the world, your default assumption will be that other people are noticing and judging your body. You will also probably disproportionately notice and judge other people’s bodies, simply because this is a thing that you think about. However, if you become more confident in your body, then you will probably stop assuming (or assume less frequently) that other people are judging you for your body, and you will probably also judge others less. Being more confident — even fooling oneself into confidence — is like allowing ourselves to test an alternative hypothesis, another way to view the world, and seeing if it’s better.

The way we relate to ourselves is intimately tied to the way we relate to others. The good thing is we can change our perceptions of ourselves, or at least act our way into a new self-perception. In large part, we treat other people like a mirror of ourselves. I want to interact with the most positive reflection of myself.

Maybe this sounds kind of obvious. I’m going to elaborate anyway. The Golden Rule is a good heuristic for people who don’t have a complex ability to be empathetic, like most elementary school kids. When I was in third grade, I did not think hard about — much less embody — the motivations of the people around me. But there comes a point in your life when adhering to the Golden Rule just doesn’t cut it anymore. There comes a point in your life when you realize that treating others the way you want to be treated often makes you inconsiderate.

Our interactions with people depend a good amount on the quality of the mental model we have of them. When I’m talking to my friends, it’s important that I have a good idea of what they’ll find funny and what they’ll find boring; what they’ll judge me for and what they’ll totally relate to. Being able to predict their reactions makes for better connection, so I do this subconsciously, as most of us do. My mental model of them is tuned with further interaction — when they react poorly to something I say or when they react positively to something a little unexpected. I presume that almost everyone has a better mental model of themselves than they have of anyone else in their life. This makes sense — you have intimate access to every thought in your head, and you probably think more about yourself than about anyone else.

I think empathy can be divided into three rough levels, depending on how well you model other people:

1. No empathy: you have no mental models of other people. You don’t try to predict their needs and wants.

2. Some empathy: your mental model of other people is your mental model of yourself. You can somewhat predict their needs and wants, especially the common ones, and especially for people who are like you.

3. More empathy: your mental model of other people is close to how they actually are. You can accurately predict their needs and wants.

In this framework, the Golden Rule advocates for being at (2). When you treat others the way you want to be treated, you assume that they are like you, and that their motivations are like yours. Though (2) is certainly a step up from (1), it’s far from (3). Being at (2) — assuming that people are like you — can lead you to harm others and remain ignorant of the fact. For example, maybe you love being told, “You look better now that you’ve lost some weight!” If you said the same thing to someone else though, they might be insulted. Even though you can hurt people at (3), the danger of (2) is that you will not understand their hurt. You will not acknowledge their feelings as valid, but rather think them irrational. After all, you would be delighted if someone said this to you! Likewise, if you’re the other party in this exchange, you will assume the speaker intended malice. After all, you would never say such a thing to somebody if you were trying to be kind. If you have flexible mental models of other people, it’s easier to interpret other people’s actions generously, because you recognize that they are not like you.

There are obvious qualms one might have with this framework. For one, the levels are not actually discrete. Most people sit in between (2) and (3), where they have some accurate mental models and inaccurate ones. This makes sense — it’s easier to have accurate mental models of people who are similar to you. Nonetheless, I find mental models useful in explaining a lot of misunderstanding, like when parents force their kids to do things their kids obviously hate, or when someone’s efforts to flirt are interpreted as offensive. In either scenario, it’s probably that the parent is doing what they wish their parents did for them, and the flirter is doing something they wouldn’t mind being done to them. Again, obvious caveats — the creep at the bar might be at (1) and just not thinking about other people at all.

Nonetheless, good intentions don’t excuse harm done. I wish more people would actively try to have better mental models of others, or at least try to recognize the limits of their mental models and apologize for well-intentioned harm. Such self-awareness might reduce a lot of fruitless other-optimizing. To my surprise, I have more to say on the subject of mental models. In particular, I want to write a follow-up to this post about how strategically changing our model of ourself can improve our outlook on life, because we are our default model. I will link that follow-up here if I ever write it.

In physics, inertia is the property of matter that resists changes in its state of motion. In this spirit, I will use inertia to mean the property of people that resist changes to their current direction.1 When my professor advised me to do something related to the experiences I already had, it was presumably because he felt my math background would be more helpful in CS research than aging research. However, I don’t think my very limited math background would have been particularly helpful in either context, so I don’t think it made sense to use that as a primary reason for choosing one opportunity over the other. In this case, inertia was not a good signal for which decision was better.

I’m not arguing that obeying inertia is always the worst decision. In some cases and for some people, obeying one’s inertia might be among the best decisions. If you have genuine career interests, it probably makes sense to do an internship related to those interests. If you have niche, in-demand skills, it makes sense to apply more to jobs which demand that skill. My argument is only that inertia in and of itself, unexamined, is not a good reason to do things.

In my case, I am a college student with very ill-defined interests. My math background wasn’t because of a particular love of math, but simply because that was the thing I’d done and kept doing throughout high school. It was something I felt like I should do, not something I intrinsically wanted to do. So, making a decision based on my math-y past was probably not a good decision for my future, if I wanted my future to look different from my past.

In general, I notice that a lot of my peers are not really interested in the things they do — their major, the careers they’re pursuing. I notice this the most in peers who come from competitive, high-pressure environments. One example of this phenomenon is a friend of mine who was IPhO Gold in high school (read: very good at physics) and, after two years of feeling like he should do something in STEM, lest he waste his technical talents, decided to study East Asian Studies. In this case, inertia didn’t serve him well. Even though continuing down the path seemed like the “natural” next step, it would have been a continuation of an unfulfilling past. I’m really happy he decided to do something that he actually liked.

I think inertia becomes harmful when people start optimizing at bad points. Imagine your fulfillment is a manifold over all the lives you could have, and making decisions allows you to traverse the manifold. The local minimums of the manifold are lives which take many related decisions to get there. So, listening to inertia is like gradient descent into a minimum.

The point of all the exploration people encourage in college is so you can pick a good starting point — one that will lead to a minimum you like. You can get a sense of whether you like the eventual minimum by walking down the hill a little bit, such as by reading a quantum physics paper to see if you might like quantum physics research, then talking to a quantum physics expert, then doing a quantum physics-related internship, then getting a quantum physics-related job.23 I’m a big fan of testing a wide spread of starting points, if you don’t like where you’re headed right now.

I hypothesize that the reason why students from high-pressure backgrounds seem to like what they do less is because they’re anchored to a bad starting point. The thing they do, the thing they identify as being good at, is a pursuit chosen for them by their parents or their sixth grade self. They feel like they’re in too deep in the current valley, and it’s tiresome to start all over again. But to this I say: we’re young, and we’re really not in as deep as we think we are.

I think people my age are too hesitant about doing things that don’t align with their current trajectory, even if their current trajectory isn’t one they really like. They look at what they’ve done, and they choose their next step among the small set of “sensible next steps” (I did math stuff in high school, so I guess I’ll keep doing math stuff here.) A better way to do this, I think, would be make the decision that seems most likely to lead to an outcome you want, whether that be money, personal happiness, impact, or something else. (Math stuff seems like a dead end — I can’t see my work doing a lot of real-world good. But this other thing seems impactful?) Sometimes, this decision will seem foolish, too far away from your current position. Do it anyway.

I have met amazing people who pivoted from quant to technical AI safety work, from Wall Street engineer to high school physics teacher. I have read books by people who pivoted from writing TV scripts to becoming a therapist, from being a waiter to a freelance journalist.4 I get the sense that it matters little how much you know. It matters how much you can learn.

I’m writing all this from the perspective of someone who is very uncertain about her future. I don’t know what I’m going to be doing over the summer — whether it will be technical research, policy research, or something management-flavored. All I know is that I am not skilled or experienced enough for my inertia to help me whatsoever. I am trying to find an area at which I could be great — not just good. At some point, my exploration will become sub-optimal, and I may commit to path I am still slightly dissatisfied with. For now, though, I am young. Even if I were five years older, I would still be young. I will keep exploring, and with earnestness.

The bulk of your impact is likely in your future. If you’re unhappy, consider randomizing your position a little bit. The life you have lived need not be the life you will continue to live, if you don’t want it to be so.

a quote from a substack post i read recently comes to mind: “…a creative work is real if it has a good reason to exist. If its creator can look themselves in the eye and say with full honesty that they know why they’re making it, and that they like why they’re making it.” why am i making this, and do i like my reasons?

one reason i am making this is because i have made a commitment to the 30/30 challenge. is this a commitment worth sticking to? i have a friend who is against the idea of the 30/30 challenge, who is against any forced creative work in general. but so many authors — including octavia butler, if i recall correctly — emphasize the importance of writing every day. persistence beats motivation. will sticking to this commitment make me a more persistent person though? probably not. but i’ve written some things that i like on this blog and that i’m certain i’d never have written otherwise. i have to continue with the blog because i would be disappointed in myself if i didn’t, and because i hold out hope for the few gold nuggets hiding in the pan.

another reason i am making this is to have clearer beliefs. i was going to say that i don’t think my blog posts are significantly helping in this regard, but when i scrolled through my posts, i changed my mind. a lot of my posts actually made me realize something new and helped me partially resolve some uncertainties. still, i’m a little frustrated by the format, because i feel like some of my uncertainties feel too large to chew off. there is too much i would need to read and think about. and though i allowed myself to break a post up over multiple days, i think the whole is greater than the sum of its parts; the meaning of the post is unclear until all the ideas are sorted out and a tentative first conclusion is reached.

those were the two reasons in my mind when i revived the blog. but two more reasons have come up. the first of those two is that i now want to express myself publicly. while scrolling on substack, i read something somewhere that was like, “you should write as your true weird self, so you can find the people who like the kind of weird you are.” i want to find my people; i want my people to find me. also, having records of your weird self is useful for screening purposes. for example, when you’re considering a new relationship, now you have a thing you can point to and be like, “if you don’t like that, you probably won’t like this.” (this is the point in the conversation where you gesture towards yourself dramatically.)

relatedly, the other motivation i’ve found since starting my blog is that i enjoy having something i can call my own. this blog helped me remember my love for writing. i write blog posts almost obsessively these days.1 i love the process of noticing an elusive idea and pinning it down in writing. this is the labor i have missed in all my compsci classes and conversations with claude. i feel all warm and fuzzy when i share this blog with people, because yeah, i’m really proud of myself for the effort i put into this. i have so much fun taking photos for each post’s cover image, customizing my site, and adding miscellaneous touches of whimsy (like the gif in my about me). look, i made a logo!

writing this post actually made me feel better. a lot of what i write is slop. this is okay. people can choose to not read it, and i don’t need to feel bad about their choice, whether they choose to read a post that i think is a waste of their time, or because they choose not to read a post i wish they would. people can decide what is best for themselves. i will almost certainly not continue posting every day after this challenge, but i’m really glad that i started this, and i’m determined to see it to its end. though i may be serving slop, it is eliana slop. and that makes all the difference.

The first essay we wrote in Writing from Life was an “About Me”, and the first book we read in the HUM sequence was the Odyssey. Over the course of writing the “About Me”, I realized how deeply I want to be known. I framed the “About Me” around a poem by Sean Singer, in which the speaker takes off articles of clothing line-by-line. Each line in the poem preceded a paragraph in the essay, in which I related the piece of clothing to a person in my life, someone I was close to but who I did not think knew me. I’m proud of this snippet:

then I took off my bones.

I yearn to be known, known to the bone, known to the marrow. I dream of flying sometimes, of the lightness that would come from my bones in another body’s hands. But when I poke around my flesh, my fingers slip through my ribcage, through my knees, failing to find solid self. If you find my bones, can you let me know?

At the same time, reading the Odyssey made me realize that not everyone feels this need to be completely known by another. Odysseus constantly shows different faces to the people around him. With Nestor, he is deferential and shrewd; with the Phaeacians, he is theatrical and calculating; even with Penelope — his wife, the one he is desperately trying to return to — he is not honest about the events of the past 10 years.1 I cannot pin down who he is. If I was as inconstant as Odysseus, I would feel so far from myself. But this, I suppose, is why I’m not Odysseus. I wrote this in my journal last September:

I have this idea that Odysseus is scared of being vulnerable… his facades are because he can’t stand to discover the real Odysseus within. But I’m romanticizing this. I think he is happy with his facades. The goal of glory matters more.

Some people in my life are Odysseis.2 My mom, for example, is an Odysseus. Around the same time I was writing these essays, I asked her if she felt that anyone in her life completely knew her. She said, “There are some things I never tell anyone.”

“Isn’t that sad?” I asked her. “Isn’t that frustrating?”

“No,” she said. “I don’t feel a need to.”

Maybe the dichotomy I’m drawing between Odysseis and non-Odysseis is a false one. There are also things I’ve never told anyone because I’ve never felt a need to do so. I cannot say everything I am. After all, life is more information-dense than language. But I guess what I mean is that I am willing to share many things that others are not, and it is unbearable to me when I feel unable to share something. Perhaps I just have a higher bar for what it means to be known.

In the creative writing class, I once said something like, “I don’t really believe in boundaries.” I half-meant it. I respect the boundaries people put up, but I think I ask more personal questions and push a little harder than most would. This is sometimes not good — sometimes I push a little too far — but I find that people are surprisingly willing to share when you just ask the question. I think I ask these questions because I wish someone would do this to me. Indeed, I wish someone would earnestly try to know my soul.

The relationships I treasure most are ones in which I feel uninhibited, in which I feel like I can say what is on my mind and be understood, or at least unjudged. Maybe the longing for a romantic relationship is just the longing for commitment, for permission to be freely oneself. In most conversations I have, a small part of me feels like I am watching my shadow puppet from above. I’m not sure how to be free of this feeling, or whether it’s only real because I keep thinking about it. Can the shoggoth be known?

She is trying to be. She does not want to be like Odysseus, floating at sea, with no one who knows his soul. I have no visions for this blog; I have nothing in particular I want to talk about. This blog is more for my sake than yours. I just want to put myself out there, so people who want to know me can know me a little better. I am happy to publish both self-pitying poetry and long polemics, as long as they are both honest. As the Goo Goo Dolls say, I just want you to know who I am.