2026-04-26 00:04:00
I paid for a lifetime subscription to Bear after discovering that this is a better place to spend my time scrolling. Thanks to Kagi Small Web for the introduction.
I do not have much of a plan here yet. I just want a simple place to write things down, collect and work through ideas, and keep a public trail of what I am learning.
Some posts will be technical. Some will be half-formed. Some will probably just be notes to myself. That sounds fine.
So: hello, Bear.
2026-04-25 22:55:00
Yesterday, I found the post i think we're obsessed with aesthetic on the Discovery Feed. It featured quite a few nice quotes. My favourite was this one:
It's a real "people die when they are killed" kind of quote. Your first reaction is probably going to be some faint amusement, accompanied by a thought like Well, duh. Thanks, Captain Obvious.1 But aside from the factual layer, it also has a personal, contemplative element. This scene shows one of those moments where you see something that you've seen a thousand times and suddenly feel a deep sense of appreciation. We come from non-existence, and go back into it. The time where every one of us gets to experience sunsets or soothing spring rain or beautiful autumn leaves is just a tiny fraction of existence as a whole. From that perspective, even the most commonplace, mundane or tacky things can seem truly beautiful.
I think this scene is from the documentary "The Kingdom of Dreams and Madness", which has circulated pretty widely online. I've never actually seen the whole thing, just the clips of it that get uploaded on YouTube. Another one I like is this one. The interviewer asks: "Aren't you worried about the studio's future?"
Miyazaki's response:
The future is clear. It's going to fall apart. [laughs] I can already see it. What's the use worrying? It's inevitable. [...] "Ghibli" is just a random name that I got from an airplane. It's only a name.
He then walks away, and the interviewer quietly says "how beautiful", to herself. I find that last part ruins the scene a bit; it feels like the interviewer is telling the audience how to interpret what was just said. It would've been much more effective to just end with the shot of Miyazaki walking across the garden.2 However, a short scroll through the comments under the video will tell you that people still have different opinions on Miyazaki's stance. Some criticise him for being a stubborn old fart who'd rather let his studio fall apart than properly train a successor and make sure it continues to live on. Others just see him as a bitter nihilist.
To me, these quotes embody a kind of optimistic nihilism. Yes, nothing is forever. Everything will return to nothing one day; the world and our lives don't have inherent meaning. But despite of that, or rather precisely because of that, it becomes meaningful what you do with your life, and how you choose to live it.3
Miyazaki's oeuvre makes it obvious that he's not a bitter defeatist who declared life to be meaningless and simply gave up. Quite the contrary, he's dedicated his life to art and creativity, with all the self-discipline, self-criticism, doubt and despair that comes with it. He's a "suffering artist" type, but not one that wallows in their suffering or sees it as the source of their art. He simply accepts it as a consequence of life.
When he laughs after saying that the studio is going to fall apart, it's not a laugh of self-pity, or a mask for disillusionment. It's a laugh born out of the awareness that all things will eventually fade away, and that, from a "cosmic" perspective, all human efforts are but small scratches on the ground. This serves as a way to balance out the deep attachment he has to his work. It's an Absurdist take, really. The same is true for the comment about the name of the studio. It's just that, a random name, an empty container, like a drinking glass. The glass is only a vehicle for the water, tea, or juice you fill it with. At the same time, it's what allows you to drink the water, tea, or juice - meaningless and meaningful at the same time.
Some of the comments under the YouTube video point out that the end of the movie The Boy and the Heron represents Miyazaki truly coming to terms with the things that he says in this interview, or maybe truly internalising their meaning. Instead of desperately trying to hold onto a form (the Studio) and trying to find a successor, he realises (and posits to us, via the movie) that it's not the form that matters, but the spirit.4 It doesn't matter whether or not people use your drinking glass - treating it as some kind of heirloom, or letting it shatter on the ground. As long as they too realise the value of drinking a cold lemonade on a summer evening or hot tea during a cold winter, any glass, cup or mug will do.
Unless you believe in some kind of afterlife where you get to enjoy sunsets for the rest of eternity.↩
This is a documentary, so it's obviously "staged" to some degree. I don't believe that Miyazaki is lying, or the interviewer is simply reading off a script with her response. Still, the context that allowed these images to get made ends up guiding what they say in a certain direction. It's another kind of inevitability.↩
Fittingly, the Japanese title of the movie The Boy and the Heron (and the book it was based on) is "How Do You Live?"↩
When watching the documentary clips, I get the impression that Miyazaki already came to terms with this a long time ago; The Boy and the Heron is simply the first time he expresses this thought in one of his movies in a very clear and straightforward fashion.↩
2026-04-25 22:31:00
How do you know the output is good without redoing the work yourself?
You've received a report, a market analysis for the new product you're planning to launch. Reading through it you notice problems: the date on the report doesn't match the date you requested it on, it's from 6 months prior. Several paragraphs have obvious spelling errors. Some graphs are mislabeled and duplicated.
The report is disregarded. The existence of typos and copy-paste errors which may not change the main conclusion of the report is enough to discard it. Someone who didn't put in enough care to make the report presentable on the surface level also didn't care enough to produce good research.
You have judged the quality using a proxy measure: the superficial quality of the writing itself. It's not what you ultimately care about — what you care about is whether the report reflects reality and points you toward good decisions. But that's expensive to check. Surface quality is cheap, and it correlates well enough with the thing you can't easily measure.
All of knowledge work has this problem. It's hard to objectively judge the quality of someone's work without spending a lot of effort on it. Therefore everyone relies heavily on proxy measures.
Proxy measures kept misaligned incentives in check. LLMs broke them.
Large language models are great at simulating a style of writing without necessarily reproducing the quality of the work. You can ask ChatGPT to write you a market analysis report and it will look and read like a deliverable from a top-tier consulting firm written by Serious Professionals.
A software engineer can write thousands of lines of code which looks like high-quality code, at least if you have just a couple of seconds to skim through it. Their colleagues will ask AI to do a code review for them, the code review will uncover a lot of issues and potential problems, and these will be addressed. The ritual of working will be upheld with none of the underlying quality.
We have built a working simulacrum of knowledge work.
The incentives almost guarantee we are in big trouble. Many workers, quite rationally, want to do well on whatever dimension they are being measured on. If they are judged by the surface-level quality of their work, then it's no surprise most of "their" output will be written by LLMs.
The LLMs have the same problem.
The training doesn't evaluate "is the answer true" or "is the answer useful." It's either "is the answer likely to appear in the training corpus" or "is the RLHF judge happy with the answer." We are optimising LLMs to produce output which looks like high quality output. And we have very good optimisers.
So here we are. We spent billions to create systems used to perform a simulacrum of work. Companies are racing to be the first on the tokens-spent leaderboard. The more LLM output workers produce, the less time anyone spends on looking deeply at the output. All we have time for is to skim it, slap "LGTM" on it and open their 17th Claude Code session.
We've automated ourselves into Goodhart's law.
2026-04-25 17:30:40
You see them in the Bear discovery feed regularly.
Titles like “My first blog post” or “My new Bear blog”. The posts may only contain a couple of sentences, yet they quickly get upvoted. Some even end up on the trending page.
I love it! I can’t think of any other place where this loving support is so strong.
It’s not about how long or unique the post is. It’s all about the joy of seeing others getting into blogging. That shared beauty always deserves an upvote.
Keep on blogging and upvoting.
2026-04-25 12:59:00
It's an interesting feeling to realize you've been propagandized to for your entire life. I'm simultaneously angry at those who've lied to me, although they're often victims of the same propaganda. I also feel free, and able to reanalyze all these things I thought I knew with an entirely new context for understanding the world.
I stumbled across an article from the Global Times titled “Phrase ‘US kill line’ sparks debate on American ordinary people’s economic fragility and social safety nets on Chinese social media." The Global Times is explicitly Chinese state media, so take it for what it is, and be aware of it's bias.
The article explains that "kill line" is zoomer slang referring to when you're playing a game, and your health is low enough that you can be killed with one shot (apparently, idk, I've never heard it, but I'm not a gamer). The gist of the article is that Chinese kids are starting to learn about American poverty, and about how close so many Americans are to being homeless, and they don't have a way to relate to it. In 1979, the Chinese government made a goal of creating a "moderately prosperous society." That is, a society made up almost entirely of the middle class, with no extremes at either end. It was an ambitious goal, as the World Bank estimated that 88% of the population of China lived in extreme poverty. Over the last forty years, they've managed to bring 850,000,000 people out of poverty by devoting their economy to centrally planned jobs programs similar to The New Deal in the United States. It's an absolutely staggering accomplishment, and they're not done yet, but the progress they've made is nothing short of miraculous. It's remarkable what can happen when government focuses it's spending internally, on it's own citizens.
The idea that all Americans know that we're just a few bad dice rolls away from complete financial ruin, and that we all more-or-less accept it, is absolutely wild. It struck me recently that last year was without a doubt the most devastating year of my life, and if I didn't have family that could support me financially, and didn't get extremely lucky at work, there's a good chance I'd have lost everything as a result.
To set the stage, I was unemployed for about 10 months in 2024. I lost my job in March, and didn't start a new one until December. So I came into the year with considerable debt just from trying to make it through 2024. Then, of course, my wife died in early February. It was impossible to give a shit about literally anything else for months (which I still struggle with now, 15 months out). I'm extraordinarily lucky that my brand new job gave me basically as much time as I needed. I had only been there for two months, after all, they didn't owe me anything. Even so, I don't like that people need to get lucky to make it through situations like that, as opposed to having a publicly funded safety net that we could count on.
Then, in May my gall bladder decided to get infected, and once again, had I not gotten lucky with how supportive my work was and if I didn't have health insurance at the time, it could've single handedly put me underwater. The hospital bill was something like $60,000 without insurance. If all that hadn't managed to wipe me out, a microburst of wind in my back yard sending an entire whole-ass tree falling on top of my house might've. There's a possible world in which I'm homeless right now.
I've realized that people don't hate homeless people for any particular reason (and no doubt about it, housed people do hate the homeless, although they rarely come out and say it). People hate homeless people because they're a reminder of exactly how close 95% of Americans are to the same fate. Instead of blaming the systems that allow homelessness to happen in the first place, we've been propagandized into blaming the homeless people themselves.
But it doesn't have to be this way. There are better ways to organize a country and economy, we know that for an absolute fact. China's done it. Do you know how much it would cost to permanently house all of the homeless people in the entire country? Roughly $9.6 billion. Put another way, it would cost about five days of this war we started in Iran for no particular reason. That's just the military spend. The devastation that this war will continue to wreck on the economy makes the two costs almost non-comparable.
Is China a utopia where everyone has everything they need? Of course not. Do they have their issues with human rights and censorship? Of course. And naturally, that's all the New York Times boiled down the whole "American Kill Line" concept to - Chinese propaganda, nothing more. But it seems obvious to me that they're doing something right.
Today, Meta and Nike announced new rounds of layoffs, and Microsoft announced they'd be offering early retirement to people (and presumably laying people off if the offers aren't taken). Amazon laid off 30,000 people already this year (but don't worry, the CEO got a 30% raise and stock is up 14.67% since the start of the year). Block and Salesforce have also announced massive layoffs, and have explicitly blamed it on AI reducing their need for human employees. And of course, because of last year's Big Beautiful Bill, all of these tech industry workers being laid off won't be able to get on Medicaid, and because of the war and tarriffs, inflation is skyrocketing and everything is more expensive.
Our economy is broken. We need to try something new, and we need to try it soon, because the amount of people suffering has been rising drastically over the past 10 years, and it's going to continue to rise as wealth inequality reaches new extremes.
We have a rare chance these midterm elections to elect representatives who are about as far left as Bernie Sanders, Alexandria Ocasio-Cortez, Rashida Tlaib, and Ilhan Omar. People absolutely hate Trump and the Republican run congress. Yet somehow, people hate the Democrats almost as much! It's remarkable. But it leaves the door open for Democratic challengers from the left who want to implement real, drastic change. In Denver, Melat Kiros is launching a strong primary campaign against Diana Degette, our 14-term congresswoman. In Michigan, Dr. Abdul Sayed is running a strong primary for the Democratic nomination for the Senate. There are examples of this all over the country, of socialists and leftists running for office. I don't know about you, but I've seen enough of where imperialist neoliberal policies get us, and I'm ready to try something new.
2026-04-25 09:30:00
i have google alerts set up for my high school and its associated creeps. this guy’s name was a jump scare in my email this morning.
when i thought i knew him, he was pursuing an arts-related phd and was the smartest, most interesting person i’d ever met. he traveled internationally to present at conferences, which sounded important, and lectured as if building up to an o captain! my captain! ovation from students.
for over a decade, our relationship lived in my head as a largely impotent schoolgirl crush – i was enamored and he indulged, but only a little bit, and only for a moment.
a few years after graduation, a friend visited me at college and yelled in my ear over the bass at an outdoor club: “you know what he did was fucked up though, right?”
/
between pestering me for suggestive photos and inviting me to “say something sexy, say something scandalous, say something provocative,” he found time to nurture my love of writing and budding interest in feminism.
he introduced me to artists like carolee schneemann and marina ambromovic, to bell hooks and audre lorde. he gave me access to his jstor account and printed a dozen articles for me to read about the body as art or medium or whatever. he would send me excerpts from his dissertation and i would vague-post poetry about him online, which he would respond to via text.
we spent hours “synthesizing” my “research” into a cohesive “narrative” “portfolio” for my application to his undergrad alma mater; his feedback was signed “your biggest fan” in big, curly letters. when i was accepted into the program, we cried and embraced dramatically in front of my classmates. and when i was in the fitting room trying on junior prom dresses, he told me he wanted to see.
/
it’s impossible for these vignettes to coexist in my mind – the image of him pouring life into me and the one of him teasing it out – and so, over time, both gave way to a new memory, where i was the one goading and coaxing and prodding and he humored me to avoid hurting my fragile, teenage feelings, which is still bad, but in a way that insists he was just a little drunk on his own koolaid and not the alternative.
the best parts of me, i knew, came from him, and my stomach burned with betrayal when i considered anything else.
then, in my late-twenties, our school was in the news for a similar story involving another teacher and i started having panic attacks when driving in its direction on my visits home.
it was a visceral repulsion, as if the big atrium of the main building had a miles-wide blast radius i couldn’t breach.
/
the relationship was secretive, but not subtle. we ate lunch together in his office daily, sometimes in coordinating outfits if he texted me that morning with instructions for what to wear.
our “work” was often in dark rooms during skipped class periods, or in the hours between school and some monthly fundraising obligation i was eager to volunteer for because i could, like, you know, just stay here instead of going home and coming back later, if that’s okay with you.
it was always okay with him.
i’m still not clear on what was my idea and what was his – whether i was the one who asked to go for coffee, whether he was the one who offered to pick me up.
/
another year and the disorientation became overwhelming.
when you are hiding something from yourself – protecting yourself from it – you feel fractured: one part of you braced against the door and another part of you slamming into it from the other side. you want to believe it’s for your own good, but you know that it isn’t, you know that opening this door is your one task of this moment.
so i eventually pulled the two peeling pleather journals off the shelf, the ones where i’d written this story out in real time, the ones i had refused to read because i already knew what happened and i didn’t need to go looking because what was there to find.
i bargained with myself that i could flip to a single random page, and if i didn’t see anything weird, i had to admit defeat and for the love of god move on with my life.
except i did see something weird.
obviously.
/
the next week was effectively spent in a k-hole of my own past trying to understand how i didn’t remember literally any of this. it’s like i was retroactively violating myself reading back through such meticulous records and transcribed texts – comments about the size of my body, a story of him spamming messages about my perfume during class and watching for my reaction, the way he punished me with silence when i wouldn’t take the bait to escalate our relationship further.
it wasn’t just him, or this, but two brutal years of my life that had been neutered in my memory. i stared at the wall for hours trying to square what i was reading with what i was sure i thought i knew.
this shattering of your perspective feels physical. your vision goes blurry and it’s as if you’re floating in some liminal space, because your through-line has snapped and suddenly your journey from there to here doesn’t make sense anymore and where is here anyway? you don’t know because you don’t know anything.
/
anyway, looks like he started a new youtube channel. curiosity got the best of me and i found myself watching an ai-generated version of him in an ai-generated boardroom, his ai-generated voice reading some ai-generated script about being hungry for hard work or something.
i scrolled more and had to stop when i heard, “what’s up, guys!” in, to his credit, the most youtube voice i’ve ever heard.
his websites, too, are full of blog posts and media analyses written by chatgpt, and not even a good model.
which is all to say:
i guess the rest of the illusion broke. whoever i thought he was would hate him.
