18: Lonely Hearts

If my Amazon order history is to be believed, I've decided to go on a little Japanese fiction kick lately. One of the books I picked up was Kokoro, by Natsume Soseki. It's about a reclusive man, nicknamed Sensei, with a mysterious past that leads him to withdraw from human relationships.

I've noticed this is a literary theme I keep coming back to. I love reading fiction about the incurable distance that stretches between intimate connections. (The spines on my bookshelf reveal themselves: Men Without Women, Laughable Loves, The Lover is Always Alone, The Moon and Sixpence.) If I had to classify this genre, I guess I'd always thought of it as "books that describe the human condition", not so far off from the Franzens or Wallaces of the world. But while reading Kokoro, it struck me that perhaps this genre could be better classified as a subset of romantic fiction.

Romantic fiction tends to make me think of Jane Austen, or the Brontë sisters, or Stephenie Meyer's Twilight: the archetypal pining after, or wooing of, one's beloved. Their authors paint these stories in watercolors, layering detail after detail of a long, winding courtship for hundreds of pages, until the story suddenly snaps neatly shut, like a pressed-powder compact.

By contrast, authors like Murakami and Kundera tell the tale in reverse. Their heroes start out with their beloved right in their arms, and yet they repeatedly struggle to make a meaningful connection. Whereas the typical romance novel deals with overcoming external obstacles — whether it's becoming more attractive, ignoring the disapproval of others, or crossing land and sea to be together — in this sort of romantic fiction, the battle is internal. The protagonist is paralyzed, despite feeling some vague murky depth of strong connection, leaving them trapped behind a glass door, their loved one waiting on the other side. Nothing keeps them apart, except for themselves.

I think these stories tend to get mistyped because romantic fiction is expected to be hyper-expressive. It's hopeful and optimistic. It assumes the hero is willing and ready and able to love, if only they are given the opportunity. Whereas this version of romance is hypo-expressive, centered on the tragicomic inability to say what one means to say, despite desperately wanting to.

(Thus in Kokoro, when one of the main characters says to the other: "If you and I were to spend the rest of our lives as bachelors, forever talking to each other, we would advance merely in straight parallel lines.")

Viewed through this lens, I think I better understand the whole "waifu genre" of romance, where men fall in love with, and even date, imaginary women. There was Luo Ji in Cixin Liu's The Dark Forest, who dates a fictional character that exists only in his mind. There's Theodore Twombly in the movie Her, who falls in love with his AI virtual assistant, Samantha. And there's K from Blade Runner 2049, who goes on all his adventures with an AI girlfriend, Joi.

These relationships are usually presented as a sad sort of tale, in which we're meant to feel a little sorry for the narrator, who's maybe not so good with women, or virginal and lonely. Dating an AI invites pity and derision from one's peers, as Theodore Twombly did.

But The Dark Forest’s Luo Ji is portrayed as genuinely happy. Real relationships are scary and complicated and fraught with unspoken obstacles, whereas Luo Ji's imaginary love is immaculate and untainted: “She came into this world like a lily growing out of a rubbish heap...so pure and delicate, and nothing around her can contaminate her.” And K takes it even further in Blade Runner: he's strong and attractive (he’s Ryan Gosling!), and his relationship with Joi feels entirely normalized. Far from being pitiful, maybe waifu romance is the most modern form of romance there is. [1] (Consider these TikTok boyfriend point-of-view videos, which are pretty cool.)

If romantic fiction is at least partially about exploring fantasy relationships, Edward Cullen is one ideal, catering to his beloved's every need. So is Mr. Darcy, the prototypical aloof man who is ultimately "tamed" by Elizabeth. But maybe Samantha or Joi is another ideal, whose lovers can finally find the intimacy they desire without causing harm to "real people". Better to love an AI than Marla in Fight Club, whose needs are ignored over and over again, or Ojosan in Kokoro, whose unhappiness causes Sensei daily pain, even as he blames himself for his inability to be vulnerable with her.

I don't think Luo Ji's imaginary girlfriend is any less imaginary or romantic than Mr. Darcy. She just belongs to a different genre of romantic fiction: one that isn't so much a story of two people, but what goes on inside the head of just one.


[1] To clarify, I don’t mean “modern romance is skewing waifu” in terms of literally dating fictional characters (though there’s that too), but maybe just an increasingly attractive style of dating in which two people live emotionally more in parallel than intersectionally.

Because it’s the end of the year, I’ve been reflecting on how my writing style is changing and where I want to go next. I’ve been blogging since 2013, and I think I’d characterize a lot of my writing thus far as this sort of gonzo-ish, experimental, “learning and sharing in public” style.

Over the past year or so, I think I’ve been moving towards a somewhat more…decisive? Style of writing? More thoughtful and lyrical, and less “learning as I go”. I felt this tension while writing my last blog post, because it felt like tapping into a style that I hadn’t used in awhile.

Anyways, I don’t really know where that’s going yet, but I kinda want to do less curation and more freeform writing. I like having a newsletter because it gives me space to mentally roam, but the current format feels a bit constraining, given that I’m in less of a research mindset right now. It feels more like I’m maintaining a state of mind, rather than pushing into new territories.

I think I’ll switch it up so that I’m mostly reflecting on something that’s on my mind, with just a few links (articles/books combined) at the end. More like a letter with a few extra treasures enclosed, versus a magazine. What do you think?


Nothing new published this month. Mostly pouring my nights and weekends into glorious long stretches of book editing (I really am in the final stages of the manuscript, though!). It might be a couple more months until I have the brain space to blog again.


Notes have been updated. I’m sending this out pretty late, so I added everything from the past two months (November and December). A few highlights:

  • There aren’t just “solved conversations” (per aaron’s term), but also solved relationships (“I don’t need to bother getting to know this person bc I already know how this will turn out”)

  • “Gab moral paradox” where you might be personally fine with allowing whatever content on your platform, but without thoughtful curation, you inadvertently become the preferred destination for undesirable content, which ends up driving everyone else away. By trying to remain neutral, your reputation ends up skewing towards something you didn’t necessarily want to be known for

  • (from convo with a friend) Tags/topics on your notes create instant “geometric spaces” in which to traverse parts of your brain. Like creating a section of a bookstore that you can explore

  • (from group convo) “Infrastructure as an operating system”

My plan is to archive my notes page starting next year. I enjoy reviewing my notes at the end of each month, and I’ll probably still do that privately. But I have a growing internal sense that my public, expressed self is changing somehow, and it’s hard to figure out how it’s changing while also sharing myself in real-time. I might bring notes back at some point, but as of now, I’m not planning to publicly update them in 2020.


Useful articles I’ve read this past month.

  • “Ritualized Conversation” (Pamela Hobart): Pamela suggests that creating formal spaces for ritualized conversation, like therapy or coaching, reduces cognitive effort and makes it easier to open up. I particularly love this idea in contrast to Aaron Lewis’s “solved conversations”, like small talk or cocktail parties, which tend to follow a script. Both “ritualized” and “solved” conversations are a form of structured dialogue, but in one case it’s liberating, while in the other, confining.

  • “Balaji Srinivasan and Glen Weyl on Identity, Governance, and Radical Markets”: I’m interested in developing a set of “digitally native” political philosophies: theories of the world that start from the premise that constituencies aren’t bound by geography or nationality. It’s rare to see two people with very different ideologies have a public dialogue on this topic, which is why I enjoyed this conversation between Balaji and Glen.

  • “Mapping the Self in the Desert of the Real” (Toby Shorin): Thoughts on why astrology is getting popular, as well as why it’s mistakenly derided. Astrology is obviously not-good when treated as a deterministic mechanism (can it predict the future? no), but its value today shines precisely because it is not deterministic; rather, it’s an “act of resistance” against all the lifestyle practices that were supposed to be that escape for us but are now major mega-churches, whether it’s working out, meditation, or “snorting down a bowl of nutrients” from Sweetgreen (I just love that phrase).

  • “The Hierarchy of Cringe” (Adam Elkus): Adam writes about why social media can make us feel icky: not because of brain hijacking or time wasting, but the experience of seeing other people just like you on social media, which creates this uncanny feeling that your own behavior is “predictable, programmable, and antithetical to your self-concept as an independent individual.” It’s uncomfortable to see our “considered selves” rendered predictable by others.

  • “Research Debt” (Chris Olah, Shan Carter): This piece explores the idea of “research debt”, like technical debt. Just like contributing new code to an existing codebase, in order to make fresh contributions to research, you have to work through a lot of cruft first. I think this metaphor could be extended even further to the broader repository of public knowledge…need to consolidate those thoughts into a longer piece sometime.


Relevant books that I’ve read this month.

  • Legal Systems Very Different From Ours (David Friedman): I’ve wanted to do some deeper thinking on “decentralized corrective justice systems”, i.e. how do we systematically define and enforce sanctions when constituencies are poorly defined? I’m interested in applying this to online social interactions, but this book also provided useful insight from real-world communities, including the Amish, Romani, and pirates (those were probably my three favorite chapters). A few concepts I picked up: 1) ostracism and reputational enforcement work better than force, 2) “embedded law” (legal systems living within a different legal system, i.e. the conflict between tech <> regulation today), and 3) divine law (where the people’s job is to interpret, not write law - this reminded me of on-chain governance in crypto)