i  <esau

Chorus

Shorty a lil' baddie (She a lil' baddie)

Shorty my lil' boo thing (Boo thing)

And shorty got the fatty

Shorty be catching mood swings

Every time I fuck without a rubber

I nutted on the covers

And I kept it undercover

'Cause I don't kiss and tell (Kiss and tell)

Verse 1

Every time I fuck, she call me daddy

My lil' mama nasty

I see the pussy through the panties (Whatever you want)

She taste like candy

She a queen, like Nefertiti

Uh oh, my lil' mama sittin' pretty

And we be shoppin' through the city (Oh, oh)

I gave her keys to the Bentley (Woah, woah)

Get off your knees, you don't gotta cry to me

I'm your best friend, baby, you don't gotta lie

I'll get you everything that you want and you need from Chanel to Celine

It's on you to decide

Valentino, yeah, I put you in the best

Start liftin' up your dress, start kissin' on your neck (Oh no, no)

Start rubbin' on your butt, start massagin' your breasts

I ain't wanna give you a baby just yet, so I backed out and nutted on your breasts

I put you in a Uber (Uber) and sent you to your bed (Oh)

The very next day (Oh), you sent me a text (Oh)

You pulled up to the crib and we did it again (Oh, oh)

Verse 2

You know, I never found love until I looked into your eyes (Until I looked into your eyes)

First time blushin', I ain't even gon' lie (Ain't even gon' lie)

Four rounds in, leavin' bruises on your thighs

You the only bitch could make me spin off on the guys

And the only bitch that make me wanna nut when I'm inside

It's your vibe, leave with me, you gettin' high

Girl, ain't bitches fuckin' wit' you even if they tried

Pussy a tsunami, I ain't never fucked it dry

Close my eyes 'fore I sleep, hope you be there when I wake

Honestly, I feel the vibe we have is great

And if there's something in your mind, you can tell me

And if I told you what's on mine, would you help me?

You gon' see the bigger picture, it's just gon' take time

Just because you mine, I'ma keep you lookin' like a dime

Ain't no imperfections, girl, you fine

Baby, what's your sign?

I hope you ain't against the slow grind

(Work, work, it's me and you) Me and you

Keep it silent, there ain't nothin' we can't do

Through the storm, baby, we can make it through

Why you actin' brand new? You know you my lil' boo thang

Got up on my Birkin, it's workin'

She said she a virgin (It's hurtin')

She my biggest fan, she always lurkin'

And she know I'm a man, I'ma put that work in

I know she can't stand me, I'm fancy

So I'ma bring her out when I get my Grammy

Late night sex, she can't find her panties

Couldn't hold it in, now she need a plan B

Chorus

Shorty a lil' baddie

She my lil' boo thing

And shorty got the fatty (Shorty got the fatty)

Shorty be catching mood swings

Every time I fuck without a rubber

I nutted on her covers

And I kept it undercover

'Cause I don't kiss and tell

No matter how hard they try, brain scientists and cognitive psychologists will never find a copy of Beethoven’s 5th Symphony in the brain – or copies of words, pictures, grammatical rules or any other kinds of environmental stimuli. The human brain isn’t really empty, of course. But it does not contain most of the things people think it does – not even simple things such as ‘memories’.

- Robert Epstein, The Empty Brain (2016)

Okay.

• What would it even mean to "find" Beethoven's 5th in the brain?^🔍 At best, you could take a perfectly precise measurement of someone's brain state and use that data to reconstruct their memory of the music. (In the way that scientists are reconstructing visual imagery via fMRI and machine learning. They haven't "found" a cat inside the brain, but there's clearly some information about the cat in the brain that is statistically useful)
• metaphors are inescapable. some are better than others.
• the evolution of metaphors for the brain (clay → hydraulics → machines → electrochemistry) traces a series of practical but incomplete models: other paradigms, with their own entities and rules, which allowed us to iteratively complexify our conceptual model of the brain^🧠. once the limits of the metaphor were reached, or new technology with more complex or expressive dynamics was discovered^⚙️, the paradigm shifts.
• it's not a bad thing to have an incomplete paradigm. it's a bad thing once it supresses a better, more explanatory one from coming along. Robert suggests that this is what the humours metaphor did, though I think there were probably many more significant factors at play if you want to talk about why it took so long for the renaissance to happen.
• there's a type of computer process that people colloquially refer to as "baking." the computer isn't literally putting bits in an oven at 180°C, but the metaphor is helpful because it quickly conveys that some ingredients were assembled, and now a process is about to occur that will take some time, and there will be some output on the other side that we care about. language is full of these and we have to make subjective calls about which metaphors are best at capturing the dynamics of a system that are relevant to us based on how well we think they strike the right balance between clarity, correctness, intuitiveness, etc.
• it's not clear to me that the information processing metaphor is stunting us. on the contrary, artificial neural networks are allowing us to interpret the brain more than ever before. materialist theories of cognitive science are giving us actually falsifiable theories of consciousness.
• his example of a "non-algorithmic" explanation of how to catch a ball is literally an algorithm. I think we have fundamentally different understandings of what "information", "processing", and "computation" means.
• his "metaphor-free" everything-is-operant-conditioning proposal is a theory that collapsed in the 70s?

Misleading headlines notwithstanding, no one really has the slightest idea how the brain changes after we have learned to sing a song or recite a poem. But neither the song nor the poem has been ‘stored’ in it. The brain has simply changed in an orderly way that now allows us to sing the song or recite the poem under certain conditions. When called on to perform, neither the song nor the poem is in any sense ‘retrieved’ from anywhere in the brain, any more than my finger movements are ‘retrieved’ when I tap my finger on my desk. We simply sing or recite – no retrieval necessary.

Why don't we just say that that's retrieval, which is maybe kind of similar to other sorts of things that we call retrieval, and see if this helps us explain or make predictions about a system any better than before? "The brain doesn't retrieve. We simply sing." is not a predictive theory that could help us understand things, rather it suggests that we shouldn't even try.

I don't think this was a good representation of anti-representationalism, and I don't claim to have any answers as to how the brain works (I'm just a guy who really liked Being You) but I think if you're going to point out why the dominant paradigm doesn't work, you need to make a better case than this.

There's a good version on Standard Ebooks. Greg Wagland's rendition is also great.^👂

For thirty years, I've sailed the seas and seen good and bad, better and worse, fair weather and foul, provisions running out, knives going, and what not. Well, now I tell you, I never seen good come o' goodness yet. Him as strikes first is my fancy; dead men don't bite; them's my views—amen, so be it.

• It's wild to think how many of today's pirate tropes this popularized or influenced^☠️ Yo ho ho and a bottle of rum!
• Pew was a really menacing minor villain! I was genuinely scared of him.
• Despite the mutiny, everyone has a very strong code of honour. Jim makes promises to the pirates and desperately hopes to keep them, to his great personal risk.
• They didn't know what caused malaria back then. The doctor is going on about the quality of the air: low, boggy air was bad. dry, mountainous air was good. Cool to see some proto-science.
• The novel's awash with nautical terminology. How many of these words and terms did the average British kid know back then?

I think I might be entering my audiobook era. Going for a run getting lost in a story is a delightful double-dip, though I wish it were possible to highlight noteworthy excerpts while listening.

A friend just asked me this. Might as well write it down.

I stopped eating meat when I was 17 because my girlfriend at the time was vegetarian, and I knew that it was one of the remaining ways I could significantly lower my carbon footprint.

A year or so later I made the switch to veganism because of the same logic.

After a few years of being "outside" of meat consumption, it became possible for me to be able to think more critically about the moral aspects of breeding/farming/eating animals.^🧱

So now I stay vegan both to reduce emissions and habitat loss, and to try and absolve myself somewhat from the suffering.

The cruelty and pain is greater than anyone can comprehend. It is one of the most evil atrocities humanity continues to commit. I desperately hope for its end. I want children to grow up in a world where they're truly horrified to learn what we did to trillions of animals for hundreds of years.

Veganism is not a movement that has flattened the curve of animal suffering - I believe that will come from in vitro meat (which is decades away from being economically viable) and medium-term animal welfare reforms - but I cannot in good conscience participate in the system in the meantime.

I first saw this on Hakushi Hasegawa's website¹ and after a little digging around in the source, saw that it was implemented with a shader and three.js

But I had a hunch that it might be possible with CSS blending modes. Not because I fully understood them, but just because it seemed like the sort of thing they're made for.

And as you can see, my hunch was correct, but as I write this sentence I still don't completely understand why.

They've been explained a million times by other people, but it's an open secret that blog posts like this are for me, not you, so here we go again:

What actually are blending modes?

In a black and white image, every pixel can be represented as a value between 0 and 1.

0 is fully black. 1 is fully white.

Blending modes are rules for what should happen when you combine two of these numbers. For instance, the darken mode takes the lower of the two.

Or the multiply blending mode, which can be written as:

where

• $A$ is the base pixel value
• $B$ is the blend pixel value (the one on top with the blending mode set)
• $C$ is the output

If you have one pixel that is 0.5 (medium gray) and another that is 0.7 (light gray) and you blend them together with the multiply mode, you'll get a pixel that is 0.35.

And to be clear, this isn't just 3 divs with 3 different hexcodes. The third square you see is two divs stacked on top of one another with mix-blend-mode: multiply set on the top one.

need proof? Toggle animation Toggle blend mode

One thing you might take away from multiply is that it's not possible to get a lighter value than the values going into it. A fraction of a fraction is always smaller than the fraction itself.

Using blending modes as masks

So how do we use these tools to make the spotlight? The key is understanding that the abrupt edges of the color-dodge mode can turn a gradient into a mask.

With color-dodge, as the blend pixel approaches 1, the denominator approaches zero and the output value increases abruptly.

Here's a medium gray base layer and a black-to-white gradient blended over it using color-dodge

Think about what's happening here.

A is 0.5

B ranges from 0 to 1.

On the left, B is 0, so we're dividing 0.5 by 1, which equals 0.5 ²

As B approaches 1, we start to divide A by a smaller and smaller number. When B is 0.5, we divide 0.5 by 0.5, which equals 1: pure white. Any higher value of B makes no difference because 1 is the lightest value possible.

As soon as the blend pixel becomes lighter than the base pixel, the output flips to white.

What if we blended a gradient over a noise texture?

Some of the noise pixels have slightly lightened, and many of them have turned fully white.

Effectively, what's happening is the computer is sampling each pixel and comparing it to the gradient. If the gradient is lighter, it hides the pixel. The further along the gradient you go, the more pixels get hidden.

Now all we need to do is make the remaining pixels black. Fortunately, there's another blend mode for that: color-burn

color-burn takes the inverse of your base layer, divides it by the blend layer, and then takes the inverse of that.

Let's color-burn a black layer over the color-dodge layer.

We start by picking one of the remaining visible pixels. Say we pick one that's dark gray (0.3)

1. The inverse of dark gray is a light gray (0.7)
2. The blend layer is black (0)
3. 0.7 / 0 = Infinity
4. 1 - Infinity = -Infinity

CSS can't display negative infinity: it clamps the value to 0. This is why all the gray pixels turn black: we're always going to be dividing them by zero.

But there's a special case! What if the pixel we sample is white? (1)

1. The inverse of white is black (0)
2. The blend layer is black (0)
3. 0 / 0 = NaN
4. 1 - NaN = NaN

NaN means "Not a Number", which is a special value used to represent undefined or unrepresentable numerical results. Fortunately for us, CSS renders these specific errors as white.³

And so there you have it. There's some artifacting going on due to tolerance for near-white values, but to recap:

1. color-dodge "hides" pixels that are darker than the gradient.
2. color-burn blackens the remaining pixels

This gives us a pixelated falloff based on the lightness of a gradient that we can overlay atop a noise layer.

Building the effect

Now that we understand the principle, let's build it step by step.

First, we create a div and center some text in it:

Then we place a new div above the text, with a noisy background image. Statically positioned elements' backgrounds paint first, which is why the absolutely positioned text is shown above it.

Then we place a div inside the noise layer and give it a radial-gradient from white to black.

Because the div is absolutely positioned, it's now painted above the text. We can set a darken blend on the darkness layer so that the darker black of the text shows through.

Now we can add mix-blend-mode: color-dodge to the spotlight div.

But wait, why doesn't the spotlight hide the text? Isn't white lighter than black?

Here's our layer structure:

<div class="container">
  <div class="text"> <3 </div>
  <div class="darkness">
    <div class="spotlight"></div>
  </div>
</div>

Nesting spotlight inside the noise layer puts it in a new stacking context.

When you set a blend mode on an element, it only blends with elements that are behind it in the same stacking context. To use a metaphor, the spotlight is inside a box (the noise layer). It can blend with the walls of its box, but it can't reach outside the box to blend with things sitting next to the box.

This changes how the box looks, but the box (the noise layer) is still blending with the text in its stacking context using darken.⁴

Finally, we add a new div over everything, with mix-blend-mode: color-burn and background: #414141 to turn all the remaining non-white pixels black.⁵

A little JavaScript to make the circle follow the mouse and you've got a spotlight!

If you'd asked me how blending modes worked before writing this, I would have said "by iterating through them until you find one that looks right."

I think that's still kind of the case. Experienced photographers probably have a stronger intuition about when to use a particular mode for adjusting one of their photos, but I'd be surprised if most of them don't also just tap the down arrow on the selector to preview what each of the modes would do.

And I could have implemented all of this with canvas and it would likely perform better, but I'd need to have learned the theory regardless, and it's cool to learn what's possible with regular CSS nowadays 🙂

Two years ago, a colleague asked me "concretely, what's a way in which you think AI is going to change the world?"

I said: the way we relate to and use software. We will stop writing or installing it.¹ We'll instead have an AIs that dynamically write our programs for us.

Two years later, I'm feeling even more confident about this. Claude Code and Codex are the worst they'll ever be, and they're already one-shotting simple webapps.

Check out this ~special~ metronome I asked Codex to make.²

I wanted it for months, but never got around to making it myself, because to code it by hand would take more time than the idea was worth. But now I can get it made in 30 seconds, provided I have my laptop on me.

The next step could be making it possible to create these apps by typing a prompt on my phone: "Write a program that does X, Y, and Z."

The model writes it and stages it somewhere. I'd maybe ask for a few iterations, and once satisfied, deploy it on my site.³

I think this kind of capability would be really useful, and I'm sure the ergonomics will improve rapidly (alongside their capabilities and possible complexity, incorporating our phones' cameras and gyroscopes and what have you.)

For now, it remains a niche power among those who are somewhat technical or tool-making. But it seems kind of easy to commodify and market: what could be more useful than making it easy to make useful things?

Interoperability will be important, and there will still likely need to be some organizational effort to design and spread standards so that my bespoke apps can interact with other people's bespoke apps.

But, basically, we're all going to become Geordi Laforge, Chief Prompt Engineer, and everything will go great. There will be no problems.

this isn't merriam-webster, but i reckon it's when you're motivated by something other than what you should be.

say you come out of the cinema with your friends and you start raving about how many of the shots were inspired by some french new wave guy.

maybe you're authentically stoked about the cinematography (the pretense), but maybe you mostly just want to sound smart about movies because of how it affects your status in the group.

someone who loves film for its own sake should be happy to talk about the film at any level with their friends. people aren't familiar with godard? that's okay, you could also talk about the acting or the story. if you're incapable of meeting people where they're at, i think people will be rightfully suspicious that you've got an ulterior motive.

sometimes accusations of pretentiousness are just anti-intellectual. i've been called pretentious for referencing the history of a genre of music as part of my appreciation for a song. it felt genuine to me at the time, and i wish the person who said that had just said "i don't know enough about that to engage in the same way and i don't want to hear a lecture." or "yeah... i liked a different aspect of it." instead it was basically like "shut up, nerd."

it takes two. one to notice that their passion is alienating their interlocutor and one to engage supportively if someone seems excited to talk about a subject they love.

this stuff all gets worse if you're talking about disliking something. if someone's perfomatively liking something, it's easier to overlook that it's pretend. if they're being a hater, it's a bad deal: being persuaded to enjoy something less in order to feel insecure about my inferior taste.¹

pretentiousness doesn't have to be about taste or aesthetics. i was pretentious the other day when i got in a little back-and-forth with my driving theory instructor about whether or not he had explained something correctly. ostensibly i was speaking up to clarify for myself and my fellow students, but once he began to double-down on his confusing explanation and i started to try to explain to him why it was confusing, i could feel my motivation shift to something else: i wanted to outdo him and show that i was smart to the group.²

it's hard to know what the exact mix of any of our motivations is. i like to flex my vocabulary in a way that is a little pretentious, but i also genuinely enjoy finding the word that feels like it perfectly conveys what i'm trying to say. having multiple motivations isn't necessarily pretentious, but you should try to reduce the amount you decieve yourself or others as to what they are.

More Perfect Union released a video on automated truck drivers (which are now on the roads in Texas.)

They argue:

1. The "driver shortage" framing is industry spin on the fact that employers are deliberately underpaying workers to increase churn and keep the average pay grade low.
2. Automation will lead to further degradation of the trucking career that has allowed many men to earn a decent wage and quality of life. The consequences of this will be dire.

Is automation necessarily anti-labour? Is it possible to automate jobs while protecting the people who do them?

My supermarket installed 12 self-checkouts several years ago. In the pen, there's now one worker who harriedly flits between the stations, swiping their lanyard to authorize alcohol purchases. Repetitive and thankless. Pretty bad, as far as jobs go.¹

I assume most of people doing this job don't love it, but that doing it is still one of the better options available to them. Maybe they don't have any skills, or the skills from a previous career aren't valued anymore. Maybe they like the flexibility of part-time work.²

All I know is that I wouldn't want that job, and that I'm grateful for the privileges of my life that mean it's not yet been an option I've had to consider - privileges that I wish were extended to everyone.

It's all well and good for me to say that I wish for a world where everyone is so taken-care-of that we'd find it unconscionable to require labour that alienates³, but steering our shambling society towards such a utopia is really hard to do. How can we identify and agree on why we have "bad jobs", align everyone's interests to the collective cause of reducing suffering, and correctly triage the opportunities we have?⁴

Maybe the answer is to read more Marx, but I don't think it's possible to get the Good Automation that reduces toil, without also fundamentally lowering the value of the equivalent human labour and therefore the power of those workers' strikes.

i.e. the truckers are right to be concerned.

But I want jobs to be automated. I want humans to do things other than scan barcodes and drive along the interstate for the majority of their lives. If we still have warehouses filled with humans in 15 years⁵, I think that's because we've shambled wrong, even if it's because we've fought tooth and nail for labour.

Maybe there's a more middle-ground Scandanavian model. Companies that automate must invest some additional amount in continual retraining such that automation remains complimentary rather than rivalrous. I'm not actually sure if that approach will be able to keep up with the rate at which human skill is outcompeted.

In any case, I think it's going to be really hard to form a coherent story and set of policies to sell automation. Workers are proud of their worth, and are going to be skeptical that any self-sacrifice will contribute to a greater good, because the collectivist spirit is so dead in the West. Conversely, corporations will argue against paying redistributive taxes, because it's the profit incentive that drives the transition to full automation.

Yet, ultimately, I think automation will win. It's going to be hard for labour to globally impede technological progress, whereas tech companies will be able to easily relocate and improve, until they're so cost-effective that even the most labour-sympathetic regions will be unable to resist interfacing with the automated economy.

Whether or not that irreparably tears the social contract, idk. Hope not.

When people say "the problem is education" they tend to mean "I want everyone to have the same beliefs as me, but persuading adults is too hard."

I think it's naive to assume that agreeing on a national curriculum is any easier, but assuming that you could, it's even more naive to assume that students would internalize the lessons you want them to.

For example: in high school, I didn't understand our English curriculum's emphasis on literary analysis essays. They seemed mysteriously tautological: fill in the essay template to demonstrate to the person marking you that you can fill in an essay template.^😱

I don't mean to insist that the subject of english "justify itself" in that stembro way. I just don't remember anyone ever telling me why it was worthwhile to learn how to analyze literature. I have since read some greek plays and can say with confidence that this was a tragedy.

Here's an explanation that might have worked for me:

Stories are one of the main ways we make sense of the world. Writing is another. We write about stories to gain an ability to think about how they're written and what they mean. This allows to be more persuasive and less susceptable to bullshit. It also allows us to share our perspectives with one another, which is essential for democracy.

I smashed that out in 2 minutes. It's arbitrary and a thousand alternate justifications exist, but the idea is that, for each student, there is a justification that could appeal to them.

Why didn't we do this? Some ideas

• We did
  • but not often enough to remember 10 years later (wisdom has a luridness problem)
  • but I didn't like the justifications ("it'll be on the test", "it's needed for jobs that pay a lot of money")
  • but it doesn't work. Existentialism doesn't vibe with kids before they know their spelling, or, it doesn't make a measurable difference. (also, if every teacher was making the best case for their subject, wouldn't it all come out in the wash and students would drift towards their natural competencies as we currently do?)
  • but in the wrong way. Teachers mistakenly believe they are teaching purpose, simply because the purpose is so obvious to them that they don't realize how much more explicit articulation it needs.
• We can't
  • it's too politically contentious to agree on
  • it's too expensive to personalize per student, teachers are too busy, etc
  • teachers are lacking in real-life experience to know all the potential applications of what the kids are learning.
• We shouldn't
  • this is something you can only figure out yourself, and it can only ever be an ongoing process.

I suspect the answer is some combination of the above and more. ^🏫 It's clearly not "no one's ever thought of it before", but pretending that it is, what outcomes would I hope for?^🌏

If I was motivated to write literary analysis because I cared about democracy, maybe my teacher or I would realize that it would then be good to actually try to engage with the substance of other students' arguments. Let's actually test if it's working! It's wild to me how little time we spent as a class, or in groups, debating what V for Vendetta actually meant.

We never tried to persuade one another to adopt different beliefs about what a story was about, or whether it was good or bad. Our essays were written in a near vaccuum, graded more on their internal coherence than in relation to other people's arguments. It's astounding, really, how little dialogue there was. No opportunities to say "I don't think that makes sense" in response to someone's poorly constructed argument. No way to learn better ways of seeing things.^🧠

And now we're sleepwalking into education's purpose being entirely lost with language models. I see students at cafés copy-pasting to and from their moodles and can't help but think "Damn, if only you were a little better at questioning what any of this is for."

It's so parallel to the alignment problem that I'd argue it's a comedy.

Laura C by Big Supermarket is a song about Tomb Raider. I love it.

Part of the reason why, is that I cannot reliably hear the downbeat in the right place. There is a clear "right" way to hear it that my mind can snap into if I try, but most of the time I hear it spasmodic style.

In the "wrong" way of hearing it, the bass drum hits on the one, which is a ubiquitous convention of rock music, and something I am very biased towards hearing.

But by that counting, the rest of the beat makes no sense: there's a snare hit on the first "and" and then on the "four" - a non-ubiquitous unconvention.

By comparison, hearing it the "right" way (with the bass drum on the first "and") gives the song a natural, head-bumping groove. When my mind lurches over to hearing it this way, it immediately makes sense and I feel almost, like, relief?

This doesn't seem to happen to other people as much. Some get a little confused in trying to count it, though I think it often has more to do with the song's timing being kind of imprecise and wobbly, not due to an ambiguous downbeat. In my surveying, most people only ever hear it the "right" way, and struggle to hear it my way at all - if they do, it's usually with effort and incredulity.

It reminds me of the spinning dancer illusion. I used to be scared that I would lose my ability to perceive it one way or the other, but so far, it seems like I'm always able to oscillate between the two, unlike another kind of illusion that is famously one way.

It's been highly enriching to have a song that highlights instability of my subjectivity so reliably... and it bumps, too!