🌻 statement of purpose
why & what I’ll write in 2025
Dear reader,
I took two months off and it drove me up the wall. After my last day of work, I booked trips to Los Angeles, Seattle, and New Zealand. I tried to “relax” and “touch grass.” Surprise to no one: I constitutionally cannot. Every day it was painful to not be writing. I made to-do lists for projects I couldn’t start and took reporting calls on the side of rural roads. I tweeted too much for my own good. Back in San Francisco, I keep staying up late because I’m in The Zone™; sentences rearrange themselves while I try to sleep. Fortunately I think this divine discontent means I’m pointed in the right direction.1
I feel increasingly certain that there is no better time to be a writer. The world’s weirdness demands explaining. Republicans are waging a civil war over H1-Bs, while OpenAI’s o3 can do harder math than most PhDs. The kids have AI girlfriends, and indicted Turkophile Eric Adams is doing perp walks with America’s favorite bad boy. Elon Musk is pseudo-president and the Democrats have lost the working class. Basic coding and copywriting has been commodified; meanwhile, the higher-order operations of sense-making, trust-building, and truth-finding are rarer than ever. Decades in weeks, years in seconds! We live in strange and, yes, unprecedented times.
People sometimes ask if I’m worried about generative AI replacing writers. Already LLMs can ingest infinitely more information than a human can. Algorithmic feeds provide seamless content personalization, adapting to our preferences before we know them ourselves. Soon, it’ll be possible to instantly create a compelling bite-sized, long-form, and book-length version of any take; to turn a physics textbook into a podcast or a TikTok or a video game. Though I used to be a skeptic, ChatGPT has become a valuable part of my revision process: an always-on editor, a thesaurus on steroids.2 Will computers do objectively better work than us? Am I catapulting myself into a doomed career?
Storytelling, however, is about more than syntactic skill. It’s also a relational act, as
says, “creating a discourse rather than a text.” Assembling mythologies, not products. Surprising readers more than they satisfice. I even treasure the imperfection of human memory; one of my favorite short stories reflects on the narrator’s attempt to remember his departed wife. He writes:I can’t remember the dress she wore or what she had bought; I can’t remember what we did for the rest of that afternoon; I can’t re-create the exact timbre of her voice or the precise shapes of her features, the lines at the corners of her mouth or the name of her perfume. I only remember the way sunlight through the kitchen window glinted from her forearm, an arc as lovely as her smile.
A lit forearm, laughter, food of the gods. Thus are our memories compressed, integrated into sparkling jewels to be embedded in the limited space of our minds. A scene is turned into a mnemonic, a conversation reduced to a single phrase, a day distilled to a fleeting feeling of joy.
Time’s arrow is the loss of fidelity in compression. A sketch, not a photograph. A memory is a re-creation, precious because it is both more and less than the original.
That is: we don’t—we can’t—remember everything. Humans are at least as lossy as AI is, but the details we keep matter. Two people can attend a party and leave with entirely different impressions; 300 million can witness the same election and debate whether a fraud or landslide has occurred. A photograph says more about its creator than its subject. It’s tempting to seek an oracle to bust through all this postmodern bullshit, but I’m sorry to inform that truth has always been a prism and not a looking-glass.
Train on what data? Clip which quote? These decisions—call them taste, discernment, subjectivity—these are the reasons we still turn to human work. After all, a wink is rarely just a wink, a sheep run rarely just a sheep run, as anthropologist Clifford Geertz notes. Behaviors have contexts, meanings, imports beyond the twitches of nerves and vibrations of atoms. Often this strange knowledge is metis, graspable only through experience. The job of the researcher, then, is to live in the world—to uncover the subtexts of small things—to convey them through “thick description”—to “enlarge the universe of human discourse”—and to make it possible to “converse with strangers.” (We’ll inch ever closer to understanding but never quite reach it. Knowing each other better remains the most elusive quest of all.)
My 2025 project is best conceived of as an anthropology of disruption. I’m interested in exploring the places where the past and future collide, telling in-depth stories about what happens when frontier technologies meet established communities and institutions. I want to do work grounded in place and culture, to put human faces on abstract secular trends. Translate between disciplines, fill trust gaps. Let the Hill staffers bring their memos and the economists their charts. Writers serve the public as historians of vibe: they tell us what it feels like to be here, now.
The specifics are fuzzy, I’ll admit. It’s a work in progress, but these are the questions currently top of mind. If you have recommended reads, tips, or thoughts, please do reach out.3
Neo-cities: When do people start new cities vs. reforming existing ones? Is Singapore an urbanist paradise or technofascist dystopia? What can the US learn from Special Economic Zones? Can you do urban renewal right? What’s the difference between Esmeralda and California Forever? Why do most charter cities fail? How might they succeed?
21st century privacy: Why do teens have so many randos on Find My Friends? Do CCTVs work? Who ought we most fear surveillance from: corporations, the government, or each other? Should I tell Claude my secrets? Should I use a period tracking app? How much will people trade privacy for safety and convenience? What is being lost, and how might we recover it?
The politics of progress: Where did IFP come from? Where do the left and right factions disagree? Is Trump e/acc? What’s up with DOGE? What differs from prior eras of tech-government partnership? How much is driven by China fear (and envy)? Will tech-driven abundance trickle down to normal people? What compromises will be made? Can San Francisco be saved?
AI adaptation: Who is adopting AI the fastest? Outside Silicon Valley? Outside the US? Where are the bottlenecks? How do I “feel the AGI”? Can “safety” outcompete arms race incentives? (Must someone die first?) Seriously—what do we do about jobs? Will I know when I’ve been automated? How much power will we delegate? What will we hold tight?
Easy moral-political spectrums can’t answer these questions; tech media ought to outgrow its beloved hero/villain stories. I’m practicing radical openness instead: looking strange ideas directly in the eye, taking them to coffee, inviting them to dance. Living with my Notes app always open, listening well and asking good questions. Indulging the senses. Cultivating epistemic openness: independence, nonalignment, the ability to tear down and reconstruct inherited beliefs.4 My all-time favorite aphorism: Strong opinions, weakly held. Writing not as a job but a way of life.
And what better place to start than San Francisco? This is the happeningest place in the world, a city of fewer than one million people with cultural and economic impact a thousand times that amount. Computers and startups, of course, but also the history of hippies and gay rights and the Sierra Club and the UN. A real literary city, the kind that inspired Kerouac’s idealism and Didion’s critical eye; whose foggy coast starred in Vertigo and inspired Otis Redding to write “Sittin’ On The Dock of the Bay.” The hall which birthed a century of Democratic politics—its radical freedoms and brutal contradictions. The longer I stay, the more I find life and lore to dig into. I plan to live abroad in the back half of 2025, but for now, I’m more than happy to have a home base I love so much.
I keep returning to that William Gibson line: The future is already here, it’s just not evenly distributed. When I spent a summer in Taipei, what stood out most was how residents viewed climate change as a present reality to manage versus a far-off future threat. You couldn’t dispute your environment—extreme weather events were common; a supermonsoon shut down the city on my third day of work. The garbage trucks played Fur Elise as they rolled through the streets, and my neighbors were all exceptionally diligent recyclers. Edge cases are everywhere. The future is at my front door; it’s in Taipei, San Francisco, and Port Fourchon. We just have to go outside.
Over the next few months, I plan to split my outputs among first-person essays, freelance reporting, and a more discourse-y podcast.5 At minimum I’ll share an update every Friday as a kind of “standup.” To quote our era’s most consequential oligarch, What did you get done this week?
If you believe in this project and want to support it, I’d be enormously appreciative. I have not figured out how to pay my bills yet, and only have so much in savings. But I’m impatient and impulsive and just want to start. So far I’m fortunate to have received a grant through Tyler Cowen’s Emergent Ventures, which will cover at least my first few months.
You can donate by “subscribing” on Substack for $8/month or $50/year. For those interested in gifting more substantial support, please email me directly. I heard it’s very chic to fund indie media these days!
Donations will enable me to focus on creative work rather than returning to the corporate grind. But no pressure if not—I hope to keep everything free. Send weird party invites, story ideas, or share posts with people you know. Words of affirmation, introductions, rare PDFs, etc. Not all gifts are material!
I’ve been surprised and grateful to see how supportive folks have already been. Friends, ex-coworkers, my Chinese parents. People joke that San Francisco is the only city where quitting your job is greeted with “congratulations” instead of “I’m sorry”; I can confirm that it’s both true and amazing (even when you’re doing something much less lucrative than a startup).
I feel indescribably lucky. 2024 was a pivotal year for me. I did hard things and built conviction in myself, surrounded by people I dearly love and admire. It’s only when making leaps like this that you realize how deep your roots go.
🌱 little life updates
I promise the next update will have Actual Content and not just writing about writing!
My main new year’s resolution is to find out if full-time writing can work, or if I should get a tech job again. Three questions I’ll ask myself at the end of 2025:
Am I having fun?
Am I on track to financial viability?
Am I doing work I’m proud of?
(+ extra credit) Is the work making a dent in the universe?
I did a Q&A with the WITI newsletter, though I never felt as undeserving as when I racked my brain for my “print reading strategy” and had to admit that I have none.
Favorite recent reads (+ a list of every book I read last year):
Season of the Witch by David Talbot: I’ve been recommending this book to everyone in SF. It’s a fascinating, turbulent history of the city’s political culture in the 60s and 70s. Not only do I feel like I better understand our quirks (e.g. why supervisors are so powerful), but it was also a truly thrilling read.
“Shoggoths among us” by
: AI, markets, democracy, bureaucracy—“Large scale, impersonal social technologies for processing knowledge are the hallmark of modernity. Our lives are impossible without them; still, they are terrifying.” (Recommendation via .)“2024 letter” by Zhengdong Wang: Fun semi-speculative-fiction about model training, evals, and AI progress. It delivers!
“Mother” by
: A touching eulogy for a complicated person.“how to have a good time in a phd” by
: I hope to love my research as much as Jess does.
Thanks for being here. Here’s to 2025!
Jasmine Sun
Some might say this is a sad way to live, but poster’s disease is real and you should be grateful it exists. The content economy runs on other people’s mental illness.
The best way to write with LLMs is “collaborative”: treating them like an editor (Which word works better? Which parts are least clear?) rather than a ghost writer (Generate a paragraph explaining X). Their prose style is still pretty annoyingly college-essay-core, and you should do the heavy creative lifting. Still, I’m pretty certain that LLMs will be “better at writing” than me soon, along certain dimensions—but I enjoy doing my own work anyway.
To be clear, I don’t believe in the view from nowhere; I have my priors (which I’ll openly share). Rather, a maximally open orientation will be the best way to learn, research, and refine what conclusions I’ll eventually draw.
I’ve been warned that fuzzy nonalignment will cost money and growth, but it’s a necessity of craft more than anything else. Audience capture and shallow feedback are especially pernicious in our digital age; I adore Substack and Twitter but they make these problems worse. Well-coffered idea machines are eager to fund writers to do their evangelizing—and it’s a good gig when the ring fits! But this year I want to be a researcher, not an advocate. My aim is to develop an audience of one.
Draft concept: A friend and I riff on a different trending article / blog post each week, like a live book club or reading group. Less buttoned-up than an interview, but still substantive—I like the idea of having a “third thing” to talk about. (Tell me if this is interesting to you!)
hereby appointing myself president of the jasmine sun stan network
All the best blogs should be hard to explain. Feels like you’ve got the beginnings of something great here, looking forward to following along.