i  <esau

Klara and the Sun is a novel written through the compound eyes of a constitutionally ingenuous artificial intelligence. I inhaled it in a few days and have been pondering its questions on love, faith, contingency, and authenticity since.

I recommend it! But this post is about an aspect of the story that its author, Kazuo Ishiguro, chose not to interrogate very deeply.

Why will robots and renewable energy not massively improve our quality of life?

I can think of some reasons: climate-induced systems collapse, great power conflict, bioweapons, wealth concentration, various Yudkowskyian misalignments. These are plausible reasons, yet it often feels like the prevailing meta-reason underpinning it all is "because it would be inconvenient to the story that I want to tell." In Klara, this presents as a Nondescript Humanity Failure Mode. Gestured at, tastefully implied, a bullet unbitten.^🦷

Authors, I get it. We need our narratives to have conflict and adversity. Material prosperity deflates the stakes. But if we're uncritical about this bias, we overrepresent the bad outcomes and instill unjustified fatalism and declinism in the culture.

Look at The Pudding's excellent piece on the science fiction trends of the century.^🍮

We're in the Scared Straight! era of our collective imagination. The future is so potentially bad, the systemic forces so overpowering, we have to use every opportunity we have to show people how fucked-up Australia might become.

I don't think these fears are unfounded — they have a basis in our material reality — we're just overselling them. And so my fear is that we internalize this pessimism so deeply that even when a story isn't expressly concerning one of the plausible hazards of the twenty-first century, our writers backdrop them by default, and we ourselves reject the fiction if they're not there.^🐧

If you want to tell a story set in the future, why not show some of the possible nice things that could happen? Black Mirror does this a little, but almost always in a way that says the nice thing is clearly not worth it.

I want a White Mirror. A stage for exploring the problems we might face if things go okay.^🖖

For instance, to my bolded question, you might respond "automating human survival" is not flourishing. A world wherein no one has to work will be filled with ennui. Great! Show me that story. Show me how you think there's more to humanity than defiantly trudging through ashes.

Show me what happens in an immortal society if someone chooses to die. What do you do if your lover wants to completely change how they look? How does a democratic group of people decide how and whether they can eliminate an invasive species? What happens to linguistics once we all speak the same language?

Treat the future with curiosity. Reject scarcity and cruelty as the sole sources of drama. I sincerely believe it's possible to do without being didactic or propagandizing, and if done well, paradoxically, it might inspire us to actually bring it about.

Jim Lill is a musician from Nashville who conducts audio equipment A/B benchmarks with unusually high rigour and posts them on YouTube.

His most recent of these is on microphone preamps.

It's fair to say he's been getting a lot of feedback:^🔊

Criticism has converged on a handful of things:

1. He's comparing a modern Focusrite 2i2 to a 1970s Neve 31105 that was specifically designed to be as transparent as possible. It's the other Neve preamps (like the 1073) that have special character.
2. He says that none of the audio engineers he knows set their preamps' input gains to distortingly high values, but apparently that is a common thing that people do.
3. He says "So Jim, are you saying all preamps sound the same? No. Some preamps are broken and some preamps are made so cheaply or poorly that their frequency response doesn't stay flat from 20Hz to 20kHz." which implies that he thinks all preamps that aren't cheap or broken do sound the same.
4. He advertises a course of his in the video and asks that people send money to his PayPal.

It's plausible that he's overclaimed, and that the 31105 does something different under different conditions, or the 1073 does something special under normal conditions. Ideally, he partners up with a skeptical audio engineer who believes that preamps do sound different, and they agree upon a set of experiments that would confirm or falsify the null hypothesis through the magic of adversarial collaboration.

But at this point in the process of the Controversial Video lifecycle, it is far easier to record your hot take on it than it is to conduct good, measured science.

The attention economy thrives on controversy and there's basically nothing we can do about it. If it takes about two months to run and record a robust set of debunking benchmarks which might make you $200^📉, but only an hour to record and upload a response which will make you $100, why wait?

Overall, I think it's fine. This is a media ecosystem that is succeeding in incentivizing things like the original video in the first place^💸, and I'm sure eventually someone with sufficient stubbornness and principles will make the counter-video that I want to see, even if YouTube wastes a lot of people's time with the cashgrabs in the interrim. I'm still hopeful that on the other end, the two sides of this debate come together to discover why scientific best practices exist in the first place.

And if I had to guess, I think a similar suite of controlled tests run on a 1073 will show that its secret sauce, if it exists, is also quite satisfactorily reproduced with a bit of EQ and saturation. Jim's core message of "you don't need expensive vintage gear to make good sounding recordings" will absolutely hold true, and probably also "expensive vintage gear's audio characteristics can be inexpensively reproduced to the standards of the average amateur musician."

That's what's actually at stake here in this debate. The people clutching pearls or insisting that they can score above 50% accuracy on an A/B test are doing something else.

Beyond its crux, the flippancy with which The Ones Who Walk Away From Omelas' narrator invents the city is the most distinctive aspect of the story to me. There's a reading where the suffering child is optional - only necessary if you can't buy the first two and a half pages - because the narrator is only negotiating with our suspension of disbelief, not telling us about a place that Really Exists.

In that way, if you can imagine or accept that this miraculous little city could exist without need of suffering, without any catch, then the child doesn't exist. It's only if you read on (because of zero-sum thinking or narrative necessity or the fact that it's a story and the point is to read on past rhetorical invitations not to) that the child is placed in the cellar.

Conditional on the child existing, within that world, there are those that walk away, some of whom presumably envision a better, more just city free of sacrificial eight-year-olds. You like to imagine you'd be one of these people.

But they're only doing that because you, the recipient of this parable, prefer to accept that arrangement than the one where there are no strings attached.

It's a metafictional knot. How this reading influences my interpretation of what the story is saying we ought to do, I have no idea. I'm surely missing context on Le Guin's apparent exasperation with whoever her contemporaries were that considered Good Art to explore suffering and Bad Art to depict flourishing ("treason of the artist", "bland utopians", etc) but that seems pretty important here. Of course, historically, one innocent child suffering for all our salvation has been considered a pretty good deal.

I guess it changes one of the stories takeaways from "refuse utilitarian bargains" to something more like "repair the imaginative failure that made the bargain seem necessary."

When I visited Kigali for 2 weeks in 2022, I observed a labour-capital ladder:

At the bottom, the poorest workers could earn a living by carrying fuel to and fro on foot.

Next, people with bicycles did the same thing faster.

If you saved enough from cycling, you could buy a motorcycle and become a mototaxi driver, providing transport for the city's middle class.

Taxi car drivers were on the next rung with air conditioning, personal space, and higher profits.

For all the cynical things I had to say about internet comments, I also cherish them.

I love the bootleg concert guestbooks, the votive candle shrines underneath golden oldies.

Or the 100x-distilled nostalgia underneath the minecraft OST as children begin to realize they are going to die:

Or the thousand-person roast of RZA's terrible beat for the guitar center promo:

Humanity condensate.

When developing a large feature, writing Good Commits is hard.

And by Good Commits, I mean something like:

This allows reviewers to step through your pull request in small bites, with each set of changes scoped to a single aspect of the feature.

So, naturally, here's what I do instead:

Latter commits overwrite work that was done in earlier commits and the story breaks.^⚖️

Jujutsu makes it easier to hop around commits and iterate quickly on compartmentalized changesets, but it's still effortful and I get averse.^🤖

jj absorb helps somewhat, as does jj squash -i, but they both have their downsides:

1. absorb assigns the changes based on whichever previous commit most recently touched those files, which sometimes doesn't actually correspond to which commit should own these particular changes.
2. squash can get you stuck in merge conflict hell if your boundaries aren't extremely clean.

So here's a solution to this problem of "git rigour fatigue" that I've come up with.

For this example, let's represent commits visually. Imagine red represents changes to the type definitions, blue to the UI and so on:

Mayhem. Our first commit is a mix of red and blue. We touch red in multiple places!

To fix this, let's create our ideal commit history first, using jj new -B messy-first -m 'red'

Then we can do the rest. (I switch to jj new -A red -m 'blue' at this point)

Then we squash all the commits with actual changes in them into one with jj squash --from messy-first..messy-last --into messy-first

Then we use jj squash -i --from messy-first --into red and pick out the red changes, putting them into the red box:

And so on:

Eventually everything's in the right place and the "everything commit" is empty.

For large features, I find this workflow far easier than having to maintain strict git rigour for the lifecycle of the feature's development. It's easier to make improvised commits with temp debugging state in them and tidy it all up in one sweep at the end.

preemptions:

1. I don't have a good name for this technique. "Doing Commits Like A Big Pile Of Laundry", perhaps?
2. This is different from (and, imo, superior to) jj split:
   1. With split, if I miss a hunk that should have been in red, I have to split again and squash.
   2. This technique more easily allows sorting the easiest hunks at the beginning without worrying about how it will effect the commit sequencing.
3. This reason why doing it all at the end is (often) better than using jj squash -i as you go is because the final state of the everything commit is guaranteed to not have any conflicts. Creating a new "fix red and green" commit and interactively squashing that into your red and green commits might break your blue commit if it happens to touch one of the affected files.
4. A downside to this technique is that there's no guarantee that every commit will compile, which might be a dealbreaker.

Last night, I watched the Star Trek Voyager episode Critical Care. In it, the Doctor is abducted into a hospital that is undersupplied & understaffed. He immediately gets to work helping the sick and lambasting the inequity of the hospital's coverage. The rich and powerful get top quality care; the poor get beds to die on. It turns out the hospital's resources are allocated by a computer (the Allocator), which calculates scores for each patient based on a consequentialist rubric. If you're the CEO of a fertilizer factory whose death would result in the deaths of hundreds of thousands of people, you get a high score. If you work in the mines, game over.

The analogue is ambiguous. Is the Allocator like private healthcare - humane treatment for those who can afford it? Or is the Allocator like a planned economy - oppressively reductive in its metrification of human worth? Whatever the case, the hospital's administrator, Chellick, argues that the Allocator is necessary: prior to its instalment, health outcomes for his society were worse. His role in the system is to own the fact that public health ethics requires ugly sacrifices.

The Doctor resists the logic of his captors, eventually infecting Chellick with a fatal-if-not-treated disease and altering his bio-signature so that the Allocator will place him into the no-coverage tier of society. It's a bold move for a digital embodiment of the prime directive x hippocratic oath, but the gambit works, and Chellick promises to reform the system.^🐊 The Doctor is then rescued, but his conscience won't let him off so easily - on what basis were his own unethical actions justified?

I think the episode worked decently well. I'm definitely thinking about it more than most Trek I watch, but in part that's because I don't think it does a very good job at resolving the problems at its heart: what is the right mix of deontology to consequentialism? If it's at least some of the former, is it right to intervene in the affairs of societies you don't belong to?

The Doctor's position is that it is, and that the system is not deontological enough. High value members of society are getting prophylactic life-extending treatments with medicine that could otherwise save the lives of sick children. This is wrong.

But is it wrong because the Allocator is malfunctioning, or because its version of consequentialism is actually just a POSIWID caste system?

1. If extending this engineer's life won't actually save more lives than the medicine otherwise could, then the Allocator is being stupid, and the Doctor could just make the consequentialist appeal to fix it.
2. If the Allocator is working correctly and it was convincingly demonstrated that more people will die because of the Doctor's interventions^🔮, would we still think of him as the episode's hero?

Ultimately, we don't see if things improve or worsen for the colony — perhaps because doing so would implicitly legitimize the consequentialist framing, perhaps because they ran out of time (as these episodes often do.)

It's a shame, because I think doing so would have sharpened the discussion around the episode in other reviews I've read. People could have debated whether or not the Doctor was wrong in deciding for others, or without sufficient evidence; or what the right virtue ethics is. Instead, most comments focus on the socialism vs capitalism Rorschach blot and whether or not Canadians are dying because the waitlists for treatment are too long.

Nevertheless, "deontologist fights for deontology by violating deontology due to consequentialist justification" is a pretty hook. I just wish they'd hung on it a little more.

What colour is the pepper on this skewer?

I say yellow. My colleagues said green.

I value open-mindedness, and in similar situations will defer to the group, but this felt like being persuaded to deny my own subjective experience: a tall order for words alone. I couldn't look at the pepper and say "green green green" and see it as any other colour.

I'm nominalist about such things, "yellow" isn't "real" - what's actually happening is we're expressing how we classify things: see ismy.blue. One could argue we just obviously need a third colour (mustard, cyan) in this situation, but that doesn't remove the continuum paradox - imagine the same website but for cyan and blue.

The strong Sapir–Whorf hypothesis is probably not correct, but could a weaker version be at play? I asked my colleagues if they knew what banana peppers were and they said no. Maybe that's why we disagree! My ontology of peppers accommodates yellow peppers already and theirs doesn't.^🫑

Could we have agreed upon a double crux to settle things? I took a photo of the pepper in bright midday sunlight, averaged its colour with bilinear downsampling and then entered the hexcode into various online "colour namers", all of which said variants of yellow/gold/mustard.

Unfortunately, I did this before checking with my colleagues if they thought this was a valid method, and after they'd lost patience with my schtick.

So now I turn to you: what colour do you think this is? Is your answer different from the pepper?^🖥️

Speaking of ontologies…

This is different from the "is my red your red" dilemma, because even if your red quale is my blue quale, we agree that tomatoes are red.

It's closer to the blue dress conundrum^👗, where people were having categorically different perceptual experiences, not just descriptive ones.

This hints at another way to resolve the pepper problem: colour theorists tweaked the photo to make it easier for people to have the alternate experience. Maybe I could do the same with the photo of the pepper? Is this anything?

I don't know. I guess this makes my eyes tingle different, but I'm still not seeing freaking green.

Oh well. As fun as it was to act incredulous and mad about this, I know there's no winning here. The only win is the knowledge that the "colour boundaries are divisive" result easily replicates.

I don't know why this feels like some sad or significant thing to say - it's one of the internet's oldest clichés - but, anyone else feeling like comment sections kinda sucky these days?

It used to be that I'd browse forums or reddit or twitter and read the comments alongside the OP for extra perspective, clarifications, or debunking.

Now, most of the time, it feels like The Comments are transcriptions of the Two Minutes Hate.

Yes The Comments can still be good, especially when enough people have written and upvoted things. For instance, here's a good joke on the YouTube video for the Two Minutes Hate scene itself.^📜🏭

Nevertheless, I feel like there has been a decline in aggregate comment quality over the last 5 years. Speculatively, due to a confluence of factors:

• I quit phpBB^🥇
• I am aging into world-weariness on schedule
• A new generation of kids who effectively missed a few years of schooling have come online
• Eternal september: smartphone edition^📱
• Ubiquitous adverserial astroturfing disincentivizes effort (i.e. the dead internet theory)
• High-quality discussion has migrated to discord and behind paywalls
• We're regressing to oral tradition and are listening to podcasts as a substitute for having our own discourse^{👨‍🦯‍➡️}
• Trump administrations depress my clan

The response to most declinist narratives is to point out how every preceding generation has felt the same way and there's no need to project my crank rite of passage onto the world. That's fine by me. I don't need to persuade anyone that something objective is going on here.^🔬

But I'd posit that there's at least one genuinely new thing going on. LLMs have raised my bar for informative and entertaining textual content. Why read the comments when I can post the contents of an interesting article into Claude and talk about it with them?^⚧︎ Claude can contextualize the subject at hand, integrate it into other conceptual threads I've been pulling on, and do so in a way that's generous and thought-provoking. It's everything I want from comments.

Well here's what Claude has to say on that:

Comparing comment sections to talking with me is unfair. I won't surprise you the way a stranger with genuinely different priors will. I won't tell you something you find deeply wrong and refuse to back down. I haven't lived through what's being discussed. Talking to me feels productive partly because I'm articulate and accommodating, but those are also reasons to be cautious about substituting LLM conversation for the messy disagreement that comment sections, at their best, sometimes provide. Given your interests in political philosophy, the risk of LLM-as-substitute-for-real-disagreement is one to take seriously.

A paradoxically compelling point! Another one is that it doesn't matter if I "convince" Claude of something, whereas I could reply to that person with the muddled numbers on inflation and in so doing might persuade a few readers and something-something democracy better.

This is an embarrasingly elitist post. Am I just dressing up classism and misanthropy in a structural critique? I don't know. I am genuinely frustrated by the number of low-information, time-wasting comments that people write - and I can also admit that the examples I showed here are pointedly uncharitable. Every day I read lots of good comments written by human beings just tryna make sense of this crazy world, and hell, I guess I love em for it.

So now do I make a Scrougey vow to write the comments I wish to see in the world?

Nah. Not until we've done something about the bots.

Meanwhile, idk ig i'll c my friends irl. ttfn!

handy if you're running a site with pictures but you want everything to maintain a consistent aesthetic and colour scheme. might be better to just process the images and save them already dithered, but this way can be customized. (note that the original image is still there beneath the filter, and that's what gets loaded, so, again, this doesn't save bandwidth - but it also means users could vary the effect per theme or toggle it off)

Noise dithering