The dream is meat that grows like plants, in reactors.

I would pay double - nay, triple - for cultivated meat: no animal suffering, no guilt, no problem. So I’ve been wanting to try it for years. And today, I got the chance. Wildtype does indeed grow real salmon cells in conditions similar to the inside of a fish, then mixes them with “plant-based ingredients” to make the final product.

There are some red flags. One, you can only buy this cultivated salmon at four restaurants in the US. The one near me, where I went tonight with three fellow bloggers, only served it as part of an omakase, along with more conventional fish. And two, the whole “plant-based ingredients” thing, suggesting that pure meat is unviable. Which is the rub, with cultivated meat. We theoretically, kinda-sorta “have the technology” and that’s very cool. But we can only make teeny tiny amounts, and at too high a cost to sell in normal markets.

Whatever. At least I got to try the stuff. It looked like this.

And it tasted like soy.

Sorry. I’m more bummed than you. But it tasted like fake meat, seitan or tofu or something. It’s possible it has some other advantages; fake meat products tend to make me queasy, and maybe this sits a little better. But it’s still not as tasty as an Impossible or Beyond Burger, much less actual sashimi.

I am maximally biased in favor here. I really want cultivated meat to be good. I’m optimistic about technology in general, and I desperately want a cheat code to let me eat meat without subsidizing animal torture. If I could have liked it, I would have liked it. I let it sit in my mouth for a while, waiting to see if I would pick up the telltale salmon taste that I love so dearly. I dipped some in soy sauce, in case that was the missing piece. But no. It just wasn’t very good.

Do I Still Believe?

Growing meat in a vat is hard. There are political headwinds and numerous technical challenges, including a very high risk of contamination, since meat outside an organism is basically the most attractive possible medium for harmful bacteria. No immune system! Not even any skin! And you can’t feed antibiotics to a vaguely pulsing slab. So you’d better have an industrial-grade clean lab, with zero contamination, while managing a supply chain of rare ingredients for which there’s currently little demand. And unless you somehow scale this bespoke operation to billions of times its current viable scale (and solve tricky marketing problems), it will barely even put a dent in factory farm revenues.

And yet, yes, I do believe. There are few things more powerful than small groups of morally motivated nerds. The challenges seem insurmountable today, and my first taste of cultivated meat did not please me, but something is better than nothing. Layer by layer, from a tiny seed, the solution may yet grow.

Or last night, after midnight, I was in a getting-to-know-you circle of four people. One person had gone to Yale. Another, Oxford. A third, Stanford, and this person had also dated a celebrity. Remarkable!

I recall a stanza from Jamie Paige:

little envy
your chartreuse color compliments me nicely
but compliments ain't likely
does it scare you? to be
everything you can be
never surely complete
don't i know it

After all, I have put plenty of writing on the internet, but none of it has won me fame and fortune, or the attention of tens of thousands. And while I applied for Ivy League schools at the advice of the Princeton Review lady who visited my high school, I didn’t get into them.

How strong the urge now, to defend myself. To make excuses. Or failing that, to show that I don’t care, that I am rich in some other, different currency. And what a shame that urge is! Not only because it is painful and pointless, but because it hides another, better reaction, hiding in plain sight.

I should simply be impressed.

A Good Impression

Being impressed is the mirror image of envy. Envy notices that someone else has something that I don’t, and experiences that fact as pain. Being impressed notices the same reality, and experiences it as pleasure. Seeing someone juggle many balls is impressive to me because I can’t do it, but being impressed is a warm feeling, a sense that it’s wonderful, innately, that someone can do such a thing.

Being impressed mirrors envy in a second way. When I’m envious, I am quite focused on myself. My inner monologue fixates on why I don’t have some desirable property, and I barely notice the envied person other than as an effigy of that trait. When I’m impressed by someone, on the other hand, I am closely attuned to details about them; to be impressed is to be interested, focused, trying to understand a positive example rather than a negative space.

Being impressed often inspires no particular action; I’m impressed watching a skilled juggler, but am not motivated to gain that skill. But sometimes being impressed is galvanizing. I was impressed by a girl’s fashion sense at a conference, and it made me want to pay closer attention to the clothes I wore and how their colors went together. I also read a book she recommended.

Best of all, being impressed is a gateway to sincere praise. People love compliments. One of the saddest things about envy is the arms race it can kick off, where everyone fluffs up like a pufferfish, fortifying their egos against each other’s achievements. Much better to relax.

Wow! You really did that? It sounds hard. What was it like?

How often do you think the impressive person hears that? If they’re extremely impressive, maybe enough to bore them.

But I doubt it.

I know this because she says “wheel bus? wheel bus? wheel bus please?” Or, if she has recently had the pleasure of wheel bus, she will say “wheel bus again?”

Sometimes it is no wheel bus again. Why?

Enough Wheel Bus for Dada (or Mama)

Sometimes, after four renditions of wheel bus, it feels like enough wheel bus for whoever is looking after baby. The Wheels on the Bus is an annoying song. Furthermore, any specific rendition (except one’s own cute baby attempting to sing it) will have its own quirks that wear thin. In the version linked above, the singer lazily drawls “all through the town” a few times at the end, for example. Fine on the first wheel bus. Or the third. Exhausting on the fifth.

But actually, it’s pretty easy to just sort of zone out, with The Wheels on the Bus in the background. We could put it on loop, leave baby enthralled, and go about other business. Probably even more effectively if we let baby watch the video, which we generally don’t.

So a new question: why not automate wheel bus again?

Enough Wheel Bus for Baby

It feels like it would be bad for her.

We love when baby sings The Wheels on the Bus. When we sing it for her, and leave out some lyrics, she will helpfully supply them, or answer quiz questions like “what does the driver on the bus say?” But she doesn’t sing along when The Wheels on the Bus is playing from a speaker. Rather, she is quietly enthralled. Best case scenario she will see me wiggle back and forth during the intro and say “dada dancing” and dance a little herself, but once the singing starts she’s fully zoned out.

Plus, the desperation of “wheel bus again” feels a little suspicious. I am happy to repeat some wholesome activities many times in a row when baby asks for them, but the intense eagerness for wheel bus again, plus the zoned out vibe during wheel bus again, makes me think maybe wheel bus should not be again all of the time.

Though there is a worse cousin to wheel bus again, which deepens my paranoia about wheel bus again’s dark side.

We Have Wheel Bus At Home

If wheel bus again is allowed repeatedly, rather than just as a treat, baby will sometimes start asking for itsy bitsy spider midway through wheel bus. But after the spider gets up the water spout a scant two times, she will announce “all done itsy bitsy spider, wheel bus again please,” or go really off the rails and request twinkle twinkle little star.

I’m simple. I don’t do TikTok. But I hear about its long shadow, cast over ever-younger children, dooming them to instant gratification and the whims of impersonal, hostile optimization. Back in my day, you finished all the wheel bus on your plate. In fact, back in my day wheel bus was a song on a Raffi cassette tape, which only played certain songs in a certain order, and then was all done. It’s all well and good to want things, but the desire for one nursery rhyme midway through another, different nursery rhyme is too restless a craving. If wheel bus is interrupted by itsy bitsy spider, and itsy bitsy spider gets the all done treatment ten seconds in, what you need isn’t actually a return to wheel bus. It’s less stimulation.

Enough Wheel Bus (Reprise)

Is toddler iPad a moral panic? They’re making a Toy Story about it (featuring Toby Fox), so yeah, probably. Maybe the Ms. Rachel should flow freely in general, and not only when Mama or Dada is very tired, or we’re at an airport, or whatever.

But all you have are your instincts and your principles, when you are canvassed for the fifth wheel bus again. My instincts say no. Maybe it’s the id, that I would personally find wheel bus again annoying, and I dress that preference up in sanctimonious claptrap. Or maybe it’s pointless anxiety, trying to exert control when it’d be just as well to give baby what she wants. Or maybe I’m right about everything, and am withdrawing addictive stimulus at precisely the right moment for her growing mind.

But she doesn’t ask for wheel bus again when we’re playing outside, or when I’m running fast (dada fast it, as she says), or when she’s playing with a marker on a whiteboard. She asks for wheel bus again when stuck in bed with us, or when she sees a phone, probably because we’re checking it, and she’s bored.

Maybe wheel bus again is just like mindless adult entertainment, some mediocre video game I play too much until I’m somewhere interesting, and then I don’t play at all. It’s fine. It’s not even bad for you. But it’s not worth the fourth again.

However! While having a baby makes attentive parents turn inward, it does not make us passive or listless. With a 17-month-old baby, I use agency a lot.

The Coaster

A friend in our neighborhood donated a front-yard roller coaster to our family, in exchange for some limes from our lime tree. We put baby on this roller coaster once and helped her ride it, and while she didn’t cry or scream, she didn’t seem eager to repeat the experience. In fact, when asked if she wanted to ride the coaster in the future, she reliably said “no” or even “no way.” In fact, she would even say “no” to other questions concerning the coaster, such as, “should dada push it?” No. Dada should not push it.

The default response would be to say “oh well, guess she’s too little for that” or even “shucks, she hates the coaster.” But it was right by our front door, and her seeing something she didn’t like all the time didn’t sound like a great experience for her. Plus, it was sad to have a fun object sitting worse-than-uselessly around.

Thus, I began a campaign. Any time we came into the house, I’d ask “should dada push it?” I’d linger by the coaster, while carrying baby inside, asking if perhaps her opinion had changed. She enjoyed saying “no” each time. But we had a ritual, and by accepting her “no” I gave her a little bit of precious, fun agency. And if she was equivocal or seemed okay with it, I would give the coaster a little push, then put it back at the top of its track.

This went on for weeks. My mom even taught baby “scary” and “too scary” to describe her attitude toward the coaster. And yet, one day, when we were inside, baby floated the idea herself: “coaster? push it?”

We went outside. We pushed the coaster. I made a whole game of putting it back, too. Now coaster offered some serious advantages, that baby figured out herself:

• It was an excuse to go outside

• It was a way to make dada do a repetitive action that made a predictable noise

Whoa! Those things are great! So great that “dada should push it” was her very first four-word sentence. Soon after, she’d announce that “baby should push it.” That’s right. She started pushing the coaster herself.

Now that coaster is her favorite toy. She says “happy coaster” sometimes or “peek a boo coaster” when I put it somewhere silly, like under an outdoor table. She drags it around the yard all by herself, and takes special pleasure in dragging it down the stairs.

None of this was inevitable! It required sustained attention, patiently observing her reactions and figuring out how to make the coaster more fun over time.

The Last Word Game

Baby loves having books read to her. Like many toddlers, she memorizes her favorite ones. Not perfectly, but she generally remembers rhyming parts, and has an idea of what’s going on on each page.

So when I read her a book, to make it more engaging, I’ll deliberately stop speaking before the last word of each sentence or stanza, and let her say it. She really likes this! When playing this game, she’ll much more often make it through a whole book rather than getting bored and trying to repetitively return to her favorite page, or trying to switch books mid-stream.

This “leave out the last word” trick required thinking through her capabilities and interests, and coming up with a strategy to help her develop them. It also opened up the opportunity for games: there’s a character named “Lucy” in one of her books which she mispronounced as “Juicy” once, and we found that really cute and funny. So when we pause before that word, she has the opportunity to play it straight or make a joke.

Finally, if baby doesn’t bother saying the last word of each sentence, it’s a sign she’s not super engaged with that particular book. Useful information!

Subscribe now

Different Music

Baby loves music. She used to enjoy listening to our music, but now she knows Ms. Rachel exists, and so she will tearfully ask for “different music” when we put on something else, until we provide the good stuff.

When tired or overwhelmed, it’s easy to immediately cave in and give her too much of the Ms. Rachel superstimulus, or to turn the music off and distract her with an unrelated treat.

But when alert and engaged, there are many more options. The other day my wife suggested we play a specific Ms. Rachel song that was about phonics, since baby has been enjoying some phonics stuff recently (she particularly likes announcing “B goes Buh”). So I played the song, letting baby know that it was only this one song, and then it would be all done, and then warned her when it would be over soon, and then turned it off. She cried angrily, which was an opportunity to help her label her emotion: “You were liking your show, and now that it’s over you’re mad.” To which she could say “Mad! I mad!” And then we could observe her getting over it much more quickly than usual, and we could be proud of her and support her transition into a new activity non-reactively.

So Many Things Between

On a deeper level, there are just non-stop moments with a baby where more attentiveness, or more creativity, makes a positive difference. You can always try to make up a new game to play with your baby, and if she likes it a little bit, you can always repeat such a game so she knows it’s an activity she can request. You can always try teaching your baby a new word (ours is now fond of “mystery” as in “where mama go? a mystery!”). And there is an infinite treasure trove of books, studies, blog posts, etc. on parenting, for private leveling up. Not to mention the wonderful returns, for those (like me) not parenting alone, of communication and bonding with coparents. Or of taking the time to send a little video of baby to her doting grandparents overseas. Never before in my life have I had such a sense that my extra energy and attention can be directed toward something so worthwhile, with such clear and enduring rewards. Writing novels is close, and a very different sort of transcendent pleasure, but the baby era soaks up not only creative mania but also quiet observation, social stamina, and bog-standard patience. Baby wants you to lift her up the stairs all at once, announcing that she’s climbing them “in one jump” ten times in a row. Do you go until your arms are about to give out? Then do you think of some other game, to ease the transition back inside? You will use your whole brain, and then some. But no other job, for me, has ever been so sweet.

Because 10⁰ is just one. To belabor the point, big numbers are like this too. Like, when you read 32,406 you naturally interpret it as:

To use another “base”, like base 8, is just to swap out “10” for some other number. Sometimes bases can be represented with little subscripts, like so: 154₈ means “154 in base 8” and would be interpreted as:

So, translating into base 10, we’d have:

In other words: 154₈ = 108₁₀.

You can use just about any base. The main rule is in any base, you can only use numerals up to that base. If you do, you can represent any integer in any positive base in exactly one way, just like in base 10. So for example, “19” isn’t a thing in base 8, because it has a 9 (which is more than 8), but “77” is.

Any positive base, I said? Well. What about negative bases?

2 + 2 = 130

Consider base -4. Now things are getting fun. You are having fun now. Trust the process.

Specifically, let’s consider 130_-4:

Okay. So “130” represents “4”. Thus, 2 + 2 = 130. Crazy! Can this possibly work? Sure! Not only can you do ordinary arithmetic in a negative base, but you don’t need the pesky negative sign for it; you can represent any negative or positive integer without it. Though you do get some zany inequalities, like:

100 is 16, and 130 is 4, so in fact 100 is much bigger than 130. In general, numbers with odd numbers of digits are positive in negative bases, and numbers with even numbers of digits are negative.

Doing addition can be quite an adventure in negative bases. You have to work out or memorize a few simple additions, much like a multiplication table in positive bases. For example, just like 2 + 2 = 130, 2 + 3 = 131. For adding multiple-digit numbers, you just carry both extra digits as needed. Like, say:

12 + 12 = 10 + 10 + 130 = 130 + 20 = 100 + 1310 = 1000 + 10 + 13000…

(If you are enjoying this post at all, try writing it out on a piece of paper and you’ll understand the next observation much better.)

Wait a minute! This is infinite repetition! We just keep adding ever greater 130s and 10s! And if we think about it, it makes sense these perpetually cancel out, because “130” translates to 4, and 10 translates to -4.

So anyway, we’re left with just the residual 10 at the end. 12 + 12 = 10. Is that right?

Yes! 12_-4 is -2₁₀, and 10_-4 is -4₁₀. Again, trust the process. 12 + 12 = 10. You just have to notice that you’ll be carrying the 130 for all eternity, and stop.

2 + 2 = 10300

We can go deeper. Negative numbers, schmegative numbers. It is time to go back in time with me. The year is 2004. I am 12 years old, at my grandparents’ house. That once forbidden fruit, Spongebob Squarepants, is no longer enough to entertain me. But there is a treasure trove of printer paper, and pencils, and boredom.

I have entertained myself with negative bases for some time. These have tormented my poor parents, who struggled through much worse explanations than this post, delivered with excitement that boils over into frustration and condescension when not immediately understood. But perhaps I can befuddle them even more. With… imaginary bases.

Like, say, what’s base 2i like?

The good news first: if you allow yourself to use a decimal point, you can express any complex number1 in base 2i. Moreover, you can add, subtract, multiply, and divide complex numbers using the rules of good old fashioned arithmetic! The bad news: since the complex numbers are best represented on a two-dimensional plane rather than a number line, you have to allow yourself more digits. Specifically, you have to square the number of digits that you’d need in a real number base.

In other words, in base 2i, you have to let yourself use the numerals 0, 1, 2, and 3, as opposed to base 2 or -2, when only 0 and 1 would suffice.

Anyway! On to the good stuff! What’s 10300 in base 2i?

Thus, 2 + 2 = 10300. I think I’ll stop here.

That’s right. You heard it here first. Play my strange games, and ascend the economic ladder. What could go wrong?

The one rule: you have to do them in your head.

Prime Finder

The classic. Start with some number - 100 if you’re new, 1000 or 2000 if you’re a veteran - and figure out what the next prime is. If you have a friend to play with, first of all, I seethe with jealousy, but anyway you can either race or take turns. Overall, this game is pretty easy, even for high numbers. I will give you some pro tips:

• You only have to check for divisibility of all the primes up to the square root of each candidate number. The square root of 1000 is less than 37, so that means you only have to check the numbers up to 31 if you start there.

• You can check for divisibility by 3 by adding up the digits and dividing by 3. Combined with the fact that divisibility by 2 or 5 is immediately obvious, you can skip most numbers immediately.

• This one may be hard to follow, but the mechanism I find easiest for seeing if some number divides some other number is adding (or subtracting) whatever multiple will make it end in 0, then dividing by 10, and repeating until the answer is obvious. I’ll do an example.

Suppose we want to see if 1016 is divisible by 7.

7 does not go evenly into 11, so 1016 is not divisible by 7. If you didn’t follow that, these games may not be for you. Alternatively, I may be bad at explaining math. I sure was as a kid; just ask my family.

Find the Factors

This is a variant of the previous game, but this time our mission is to find the full prime factorization of each number in sequence, starting with some moderately high number. Lower than for the last game, though. I recommend about 200. Using 200 as an example:

Etc.

This one is most fun with some time pressure. I usually play it when driving on an interstate, trying to factor each mile number before the mile ends. It’s less cognitively intense than “find the primes” and is extra good when the numbers get higher over time.

It’s also fun with bigger numbers, like 2886, because you have to hold more things in your head as you go. (For example, you add the digits and notice it’s divisible by 3, then divide by 3 in your head and get 962, but now you need to figure out what 962 is divisible by while also holding in your head that your answer includes a 3. Thrilling!)

Long Division

You just do long division in your head. Pretty straightforward. The trick here is that any long division problem will eventually start repeating decimals, so even though the result may be infinite you can always come up with a definite answer. This isn’t the most fun version of this particular game, though. That would be…

Long Division in a Different Base

Now we’re talking! By “base” here I mean alternatives to “base 10.” In base 8, for example, you count 1, 2, 3, 4, 5, 6, 7, 10, and 77 + 1 = 100. To do long division in a different base, you have to do all the computations in your chosen base.

Actually, I’ve kind of skipped some steps here. To warm up for this one, first you should try doing some simpler mental math in a different base. Like, say, multiplying two-digit numbers in base 8. For example, let’s do 17 * 23.

The fun thing about this warm up game is you can check your work by converting both the question and answer back to good ol’ fashioned base 10.

17 in base 8 is 8 + 7, which is 15, and 23 in base 8 is 16 + 3, which is 19. 15 * 19 in base 10 is 300 - 15, which is 285. So to check our work:

Tada!

So anyway, you do that, but with long division. Good luck! Next time I do zany computation-posting, we’ll get to one of my favorite mathematical diversions: a mysterious world where 2 + 2 = 10300. But that shall have to wait.

PAULO: But you promise to begin restoring Man’s control over Nature. But who will govern the use of power to control natural forces? Who will use it? To what end? How will you hold him in check? Such decisions can still be made. But if you and your group don’t make them now, others will soon make them for you. Mankind will profit, you say. By whose sufferance? The sufferance of a prince who signs his letters X? Or do you really believe that your collegium can stay aloof from is ambitions when he begins to find out that you’re valuable to him?

TADDEO: What you really suggest is that we wait a little while. That we dissolve the collegium, or move it to the desert, and somehow—with no gold or silver of our own—revive an experimental and theoretical science in some slow hard way, and tell nobody. That we save it all up for the day when Man is good and pure and holy and wise.

PAULO: That is not what I meant—

TADDEO: That is not what you meant to say, but it is what your saying means. Keep science cloistered, don’t try to apply it, don’t try to do anything about it until men are holy. Well, it won’t work. You’ve been doing it here in this abbey for generations.

PAULO: We haven’t withheld anything.

[…]

TADDEO: Not intentionally, but in effect you did—and for the very motives you imply should be mine. If you try to save wisdom until the world is wise, Father, the world will never have it.

A Canticle for Leibowitz was written in the 1950s. Back then, nuclear weapons were pretty new, and it was an open question whether humanity, as a whole, would blow ourselves up. Now we’ve had the capacity to destroy our collective environment for nearly a century, and even though some somber guys with a clock assure us that the nuclear threat is worse than ever, it’s hard to really feel that.

This dialogue jumped out at me because it reminds me so much of Pause AI, and the Long Reflection, and other modern threads of thought that say “we need to slow down and develop our collective wisdom before we increase our power.” Other than a million papers gaslighting and gatekeeping frontier models in various ways, AI alignment as a field has a strong vibe of “reviv[ing] an experimental and theoretical science in some slow hard way.” That solving philosophy may sound hard or even impossible, but that it is our solemn duty to achieve it before we become too powerful.

One feeling to take away from this passage, from the 1950s, is that nothing new is under the sun. That however current discourse is, it’s all been going on forever. But actually, I think that’s not quite the right lesson. My preferred lesson is that the 1950s in particular must have been so, so crazy. AI might end the world, and in principle might be more dangerous than nukes, but nukes were actual, deliberate weapons! That were actively deployed against cities with civilians! And then we quickly figured out how to make even more powerful ones, and made thousands and thousands of them. Maybe people have been somberly saying “this new technology will kill us all if we don’t wise up about it” forever; who knows? But in the 1950s, that was simply a smart and sensible inference. Whether it will similarly all work out (for 80 years, at least) this time is left as an exercise for the reader, but we really have been here before.

Another feeling is that I wish the “let’s slow down and reflect” movement had a better answer to the Taddeo strain of thinking. Maybe because I see the outcomes of focus group-y processes, and the general public is very into safety, the messaging I see often feels so cloying. Like, “let’s pump the brakes until we figure all this stuff out!” Is that a good idea? Sure, maybe. But what if we never figure it out? “If you try to save wisdom until the world is wise, Father, the world will never have it” is a banger. On an emotional level, I’ve yet to see a good reply.

Of course, Paulo in the book isn’t trying to reply to Taddeo on his own terms; the modern debate is all from a secular and consequentialist perspective, while Paulo comes at it transcendentally. He serves God by preserving the old knowledge, and not using it for worldly or destructive ends. Maybe the best argument for pausing AI is simply that most human beings hate its aesthetic, are uncomfortable with the deep and eerie mystery it poses. “I just feel like we should turn back” is easier to respect, for me, than “we can only do it when we’re perfect.”

If you use AI to write something, people will know. Not everyone, but the people paying attention, who aren’t newcomers or distracted or intoxicated. And most of those people will judge you.

The Reasons

People may just be squicked out by AI, or lossily compress AI with crypto and assume you’re a “tech bro,” or think only uncreative idiots use AI at all. These are bad objections, and I don’t endorse them. But when I catch a whiff of LLM smell, I stop reading. I stop reading much faster than if I saw typos, or broken English, or disliked ideology. There are two reasons.

First, human writing is evidence of human thinking. If you try writing something you don’t understand well, it becomes immediately apparent; you end up writing a mess, and it stays a mess until you sort out the underlying idea. So when I read clear prose, I assume that I’m reading a refined thought. LLM prose violently breaks this correlation. If some guy tells Claude to “help put this idea he has into words” then Claude will write clear prose even if the idea is vague and stupid. If the guy asks to “help find citations” and there are no actual good ones, Claude will find random D-tier writeups and link to them authoritatively. Worst of all, if the guy asks Claude to “poke holes in my argument” when the argument is sufficiently muddy, Claude will just kind of make up random “issues” that the guy will hedge against (or, let’s be real, have Claude hedge against). So you end up with a writeup which cites sources, has plenty of caveats, and… has no actual core of considered thought. If you read enough of these, then you start alt-tabbing away real fast when you see structured lists with bold headers, or weird clipped parenthetical asides, or splashy contrastive disclaimers every 2-3 sentences, or any number of other ineffable signs subtler than an em dash.

Is it possible that a 50% AI-generated hunk of text contains a pearl of careful thinking, that the poor human author simply didn’t have the time or technical skill to express? I suppose. But it ain’t worth checking.

Second, and closely related, AI prose is a slog. There’s way too much framing, there are too many lists and each list has a few items that serve no purpose, the bold and italics feel desperate, and it’s just all so same-y. In your own conversation with an AI that you can fully steer, you can sometimes break out of this feeling for a little bit. But reading the output of someone else’s AI conversation is rarely any fun.

In short, if someone reads writing “by you” and it seems LLM-y, they will think both that:

1. You probably don’t have an actual good idea under the cruft

2. Even if you do, the cruft is going to suck to get through

If you want them there, they are not going to stick around. In fact, the more you want a reader, the more likely they are to be turned off by this stuff. Even if they’re the biggest AI fan in the world.

Luddite! Moralizer!

Fine. I admit it. Just this week, I too experienced Temptation.

You may know me as an editor. In this capacity, I was revising an academic paper’s abstract in response to reviewer comments. But I had several papers to work on in the same project, and the owner of that project actively encouraged me to use AI to move fast enough to meet deadlines.

So I gave Claude the paper and the reviewer comments, and asked it to come up with a new abstract that would satisfy the reviewers. The result looked good.

“It’s just an abstract,” I whispered to myself, face lit eerily in my laptop screen’s blue light. “Summary. Synthesis.” I rocked back and forth. “I could… just…”

But no. Claude’s abstract was a useful reminder of which paper this was, and Claude helpfully catalogued what the reviewer requests were. Still, I rewrote the abstract myself, from scratch. In so doing, I noticed a lot of things I hadn’t seen, when I was just skimming the AI output. Stuff it included that it didn’t really need to. Stuff it emphasized that wasn’t actually that important.

Did I run my abstract by Claude in turn? Yes! It had two nitpicks, one of which I agreed with, and fixed in my own words. Use these tools. You should totally ask Claude to find you sources for a claim, but then you should check those sources like you would check the sources of an eager day one intern, and expect to throw most (or all) of them away. You should totally ask Claude to fact check, but expect it to miss some factual errors and unhelpfully nitpick others. You can even ask Claude to “help clarify your thinking.” But if you’re really just clarifying it, then you won’t use its text. Because once your thinking’s clear, you can write the text yourself, and you should.

But it wasn’t. In fact, it went away so thoroughly that I forgot about it. Just like it will this time. And wow, what a great time that will be.

A lot of happiness is probably just noticing that stuff isn’t wrong, before it goes wrong. What else has this shape?

Basic Executive Function

Sometimes it’s quite easy to clear off a table, or put away laundry, or vacuum a room, or run the dishwasher. Other times, it’s very hard. It’s quite joyful to bustle through several domestic tasks in a row; you get flow state and can immediately see the tangible benefit from your actions. There have been times when I ruthlessly attacked various sink drains in my household, simply because I felt like it. Good times!

Low Work Stress

Sometimes work is just sort of annoying and invisible. It doesn’t require heroic reserves of energy to do, nor is it terribly exciting. Nothing is on fire and it’s easy to forget about the big picture and just sort of go from one task to the next. When there’s a crisis, be it personal, team, or industry level, the boring times sound paradisical. Perhaps they are!

Having A Really Exciting Work Of Fiction

Sometimes the book I’m reading, or the video game I’m playing, or the Netflix show I’m watching, kind of sucks. I’ll still get through a mediocre or uneven game, or struggle through a show that I watch with some family member if it’s just okay. And I’ll dutifully complete a book if a friend recommended it, or I’m reading it in a book club situation. Mediocre entertainment is still entertaining.

But a great work of fiction? Or even a decent work of fiction well suited to exactly what I want? Few pleasures are as sublime. When I’m playing a great video game, I sort of think “that’s just what video games are like.” Not so! Doing Mythic+ Spires of Ascension runs really was something special.

Exercise

If you exercise regularly, wow. That’s so good for your well being. Sometimes I do, and sometimes I don’t. There are levels. Rarely, I’m super sedentary, usually because I get the flu or something and am too sick to walk. Usually I take daily walks, but that’s about it. Sometimes I also do calisthenics or noodle around with dumbbells. And occasionally, for a few months or even a precious year, I’ll have an actual, fully fleshed out gym routine. Each of these tiers is so much better than the ones below! Yet I seldom appreciate my position on the hierarchy, and how much better it is for me than the tiers I might tumble to another month.

Diet

Lately I want ice cream all the time. I don’t give it to myself all the time, but if I scan my body for what I want to eat, it is ice cream. I’m not sure what this means, but it’s probably not good.

Which reminds me that there are times in life when specific healthy meals, that I can cook, strongly appeal to me. What wonderful times those are! I bet I’ll even get another one of them, sometime in the next few years. Something to look forward to.

And More

So many other categories. Really close friendships, and specifically friend romance where you’re just totally on fire for a new person in your life. Regular social events. Frequent time spent with family. Having the house to yourself sometimes. Having company sometimes. Having a sweet baby toddling around learning new words all the time, and saying “it’s a mystery” because you said “it’s a mystery” after she asked “where mama go?”

Mama was in the kitchen. We solved the mystery. What else is there, even?

When I am asked “how are YOU" I am not sure how to answer. It is a vexing question. I’m great, man. Sincerely living the dream. But also…

Depression

There are a lot of ways to think about depression. Of course it coincides with low mood, and low mood is a big part of its essence. But there’s also a difficulty in making deals with the self, with regulating things, with being willing to take basic, clearly positive actions.

When I wanted a baby and didn’t have one, I was often depressed in the low mood sense, but there was a clear thing causing that low mood. I am happy to report this was no illusion, no hill that, once climbed, caused another hill to magically appear. Having a baby solved the emotional problem. Recommended.

In the baby era, though, I am still sometimes depressed in the more ineffable sense. It’s hard to pick up the toys, and I feel awful about the fact that it’s hard to pick up the toys, and my head hurts but it’s hard to drink water, and I haven’t slept enough but when 11pm comes around it’s hard to stop gaming, and then it’s midnight, and then it’s 1am. A month ago, it was actively exciting to get to bed on time, to try my best to feel good and reap the rewards, and to take pride in this basic regulation. Now, not so much.

Lo! A vista opens before me, as I type these words, and thus arrange these thoughts. I have spent just a little too long under the header “depression” and its ways are coloring my ways.

Prioritization

Two months ago, I knew what I needed to do to improve. I needed to look after baby overnight more often, and tidy the house. Well, I do look after baby 3 nights a week now, and the house is somewhat - but not extremely - tidier. I am not sure what’s next. It feels difficult to simply defend the territory I already have, and I wonder if pushing further on these frontiers will send me into Siberia, where gains are slim and collateral damage high.

Still, there are so many other frontiers! How to choose one? It would be very good to improve my diet, but how? Cutting down on unhealthy stuff? Cooking at home more? Integrating more protein, including protein shakes? Likewise, my exercise. But when I try to work out at the gym I get so tired these days, in a weird, bad way I didn’t used to a few years ago. Should I push through this? Try a different approach? Cut down my ambition per workout? Treat it as an extension of the dietary problem, and force down protein I don’t really want? Treat it as a medical problem, and try the roulette of specialists who will very likely say “oh yeah I dunno, long covid” and leave me with nothing in particular to do?

Nor do these frontiers feel like the important ones. The important ones feel more like “be more secure in my relationships” or “feel like I have a handle on what’s going on in the world, or else a stoic peace about not having such a handle.” But I’m not sure how to get traction there. Therapy? Feels like it could help or hurt, and demand far outstrips supply in my area. Medication? Let’s put a pin there; I have spent all the energy in reserve for “prioritization.”

Confusion

Literally speaking, I know it’s totally possible to be having a great time on the macro level, and yet struggle under some difficult micro-level mental weather. I got a nasty flu a couple weeks ago. Then my wife went off to mardi gras, and I lived with my in-laws and baby for a week with her gone. My in-laws looked after baby as I got over sickness, including overnight, which was great but also disrupted my routine and one of my recent sources of self worth. Meanwhile, work was intense including on an industry-spanning level, and when my wife got back she was immediately really sick, and now baby is sick, and I still have lingering flulike symptoms.

I’m under lots of transient stressors. I am handling these exactly like a person under stress generally would. I seek comforting behaviors. I want to game too much. It’s hard to pursue goals, or believe viscerally that stuff will soon be better.

But the weirdest part is like, this lacuna clearly exists in a really wonderful era. I feel like I should feel the wonderfulness, like I have no right to be having a bad time. Yet, well, my head hurts. Nor is it a thought I can reason with; I don’t think I should be feeling great. I think my struggling, such as it is, probably makes sense. But I feel like I should feel better. And so I feel like something is wrong with me. And also like I have spent too long under the heading of “confusion.”

Resolution

I’m good.

I have mixed feelings about the idea of “building an audience” for fiction. But I have positive feelings about my work being broadly available, so someone who likes it can browse to their heart’s content. I have received one piece of fan mail from a stranger about my online novel, and it’d be nice to invite more of that.

I can imagine it now. A “library” section of my personal website, with a carefully polished aesthetic, with links to each work that open into customized experiences with all the bells and whistles. Mobile friendly, keeping your place when you leave, optional unobtrusive backgrounds that match the themes, links in the “about” section to YouTube playlists of recommended songs to pair with any given story. I could even hold a few novels back in case I ever want to query agents, and still release tons and tons of fiction, for free, in a way I’d be really proud to share with cool people I meet at parties. “What do you do?” “Oh, I write fiction. Here it is.”

I hope someday soon, one of my weekly posts will announce this project’s completion. But that’s not this post.

Case Notes Hero

My dad’s a therapist. One of my friends is a social worker. They have to write case notes. This activity seems pretty badly suited to the human brain; you have an hour long conversation with lots of detail, and then have to summarize the important bits afterward.

Shouldn’t there be systems that could help with this? Like, OpenAI’s free whisper can transcribe streaming audio. There are tools to automatically make a transcript. And other tools to create a summary of that transcript. The whole thing could be done on a user’s device; record the conversation locally, generate the summary, delete the recording, email the summary to the therapist or social worker. Probably HIPAA would cause problems here, and even typing this sentence I am afraid that someone will Kool Aid Man through the nearest wall to arrest me on the basis of this imaginary project I may one day attempt. But in theory you could make this app, and sell it for a $20 one time fee, and help lots of social workers and therapists have a better time.

The Idea(s)

Maybe you get the idea. But in case you don’t, here are more projects:

• My personal finances could be better organized, such that I automatically have a sense of how things are going at a single glance, instead of having to log into a bunch of separate accounts and add numbers up.

• My exercise routine could be much better, if I simply kept careful track of reps, weight, etc. I’d also like to someday be able to touch my toes without bending my knees, and a rotation of different yoga courses might help there.

• I could probably have fun conversations at local meetups related to my interests, if I went to the trouble to find them. Or at least try.

• My baby is obsessed with her letters. I made an Anki deck for her as a game, but only put the few letters she knows the best. I should update it to have all 26 letters, and see if she has more fun with it now that her skills have increased.

• I could organize all these goals better, such that when I have the itch to accomplish something and miraculously have no urgent professional or parenting tasks pending, it would be lower friction to get started.

The Point

In the current AI era, all of this stuff has gotten much easier. I could probably make a pretty decent library of my fiction in a day. I could either make the case notes app or get far enough to give up in a couple weekends. I could have AI generate baby’s Anki deck for me in 2 minutes, or come up with both an automated exercise program and tracking system for myself in half an hour.

And not only is it easier to do this stuff on a basic level, it’s also easier to spruce each one up to a slightly better level. Like, if I want to go to meetups I could just check meetup sites. But I could also create a scraper that aggregates across a few different meetup sites, and sends me a custom email when one is unusually likely to meet my preferences. Which would probably take, like, an hour or two.

Basically, if it’s a software project, simply bothering to try is most of the battle. Hours of effort are quite likely to yield usable, cool stuff. And it turns out there are lots of ways to use software to make your life better. And even physical projects, of which I also have several (e.g. organizing my bookshelf), can benefit from software tooling at the organizational stages.

So the opportunity cost of just kind of vibing, or gaming, or whatever, feels much higher than it used to. I could play around with Claude Code, which feels kind of like a video game, and end up with an actual useful output. So why not just do that all the time? In every free moment? The obvious answer here is burnout, that rest is valuable in and of itself and directing a coding agent is tiring. So, okay, sure. Don’t do it all the time. But how much is correct, outside of work? An hour a day? Two hours a day? Fifteen minutes? There is no objective answer, and time bleeds by nonstop.

The Answer

I don’t have one. It feels like one of two things is true:

• Using modern AI tools will be an extremely important skill in the near future, and it’s actively useful in the present, so it should be a high priority

• Using modern AI tools is quite useful in the present, and in the near future the tools will just use themselves and so time’s running out

Either way, seems like now is the narrow window to roll my sleeves up and boss the machine around.

But also I am very tired. My job is demanding. Having a baby is demanding. I do theoretically have free time in which I can rest. Such as this very moment. I came into my room to rest. Instead I wrote a blog post. And even now, with a box for the week freshly checked, and my lovely Nintendo Switch 2 trying its best to beckon, I feel my computer’s siren call.

What are the Benefits?

First off, the book you want is almost certainly free. Project Gutenberg has pretty much every classic novel ever, and if you don’t like them there are plenty of other sources for the famous ones.

Second, old books often have unfamiliar or archaic words. Just highlight one on your phone, and in most reader apps (I’ve been using Play Books) you can see the definition in one click. And if you have a tricker query, just tab over to Google or Claude or whatever.

Third, you can read in the dark. A lot of my reading time is when I’m looking after baby overnight, and it’s too early for me to sleep. I’m going to be on my phone anyway. Why not be reading something good?

Related to this last point, there is a secret fourth reason I will get to later.

What About the Innate Majesty of Physical Books?

I read a lot of modern books. They tend to have bigger font and less punishing vocabularies, and they tend to be easy to digest, so I often binge them over just a few days. The physical form factor works great here. It’s cool to hold a bunch of connected paper that doesn’t emit light, and people should do it more often.

But the classics are often long, reward shorter spurts of careful reading, and have unfamiliar terms. Because they’re long, printed versions often have tiny font you have to squint at. And you’re not financially supporting the author by buying a copy; they’re long since dead.

So far, my policy is to buy a physical copy if I get halfway through on the digital, then take a hybrid approach for the rest. But even then, I lean hard on the phone.

Why Not a Kindle or Something?

You have a Kindle? You’re way ahead of me! Just use that. I probably won’t buy one because I feel like I have enough devices, and AI will somehow induce me to buy more soon anyway.

But Isn’t Phone Bad?

Phone can be bad! I’ve got the flu currently, and so am strangely compelled to scroll on the dedicated Subreddit for a video game series where I played the last release six months ago. This is an atrocious waste of time, with exactly enough vaguely entertaining posts to make it addictive without making it any good.

But also phone is physically available, which is much of its badness, and having a wholesome default phone activity that can soak up arbitrary amounts of time is good. When I pick up the phone to do nothing in particular, I now can think “hey, may as well read some War and Peace.” Maybe I will decide no, it’s Subreddit time, but also maybe not!

This, by the way, is the secret fourth reason: simply having an activity you can do on your phone, that you actually think you should do, is good. I’m going to check the phone. I don’t want to install a bunch of apps that limit how much I can use the phone. But I can make using the phone, on average, more wholesome. It’s easy, too. Just download a free app, download a free file from a reputable website, and boom. Literature time.

When I bought my house, the stipulations on repairs had, if taken literally, a really obvious loophole. My realtor assured me it was all standard and would be fine, that there was no sense in pushing on the wording. I took his word for it. It all went fine, as advertised. But a friend of mine used the same realtor, and ended up having to back out of a house deal because of issues with an unlicensed inspector, and ate several thousand dollars of losses.

When my wife and I were first trying to have a baby, we were interested in medical surrogacy options. The lady on the phone at the fertility clinic gushed about how our case, which we explained carefully, was exactly the sort they covered. We were on a wait list for a few months, fertility timelines ticking. And then the doctor said no, by statute it would be illegal for him to help us.

I could go on. I complained of pain in a certain area of my mouth for a while, my dentist waved it off, and then when I was flossing the filling there ripped out, and I needed a root canal. Tax software let me file a California return when unambiguously, I legally did not need to, wasting a few hundred dollars in a year when I could have used the money.

Claude Code hallucinates sometimes. It’s not totally reliable.

But, like. Have you met people?

It’s Like That One Fanfic

In Harry Potter and the Methods of Rationality,1 there’s an idea called Nihil Supernum. The relevant quotation, translated from Latin, goes like this:

No Rescuer Hath the Rescuer. No Savior Hath the Champion. No Mother nor Father. Only Nothingness Above.

It’s a meditation on heroic responsibility. There are no adults or authorities you can appeal to, when you try to achieve greatness. There’s no spectral teacher, grading your performance on effort. There’s nobody else to blame. Only you.

Hiring professionals gets you two things: one, the actual expertise and labor of the professional. But two, now there’s someone to hand the burden of responsibility. Perhaps I could have done some research and drafted my own contract, instead of hiring a lawyer, but even if I came up with something better, how would I know? When you hire a professional, the answer to “how do I know it’s any good” is that brute fact. A fancy person said so, and you paid them money. Ergo, you’re covered.

In many situations, the heuristic “trust a professional” is great. You should definitely trust a computer security professional, rather than inventing your own super cool cryptographic scheme. If you have to appear before a court, you shouldn’t represent yourself. You should follow the typical vaccination schedule, rather than read a million very official looking blog posts arguing the pros and cons of specific shots. And so on.

But another two great heuristics are “notice when something keeps not working” and “notice when you have an affordance that didn’t exist before.” The former gently suggests that if lawyers keep jerking you around (the last one I tried replied to an email asking a specific question about the fee schedule with a flat request to sign it), maybe you should consider other options. And the latter suggests asking yourself, “hey, so, could the computer handle this?”

Maybe The Computer Can Handle This

There’s something called Pre-Need Guardianship, which is sort of a formalization of godparenthood. You tell the government that if you and your spouse both die, you’d like some specific person to take care of your children. The court can say no if it turns out e.g. that that specific person is a felon, but it’ll probably say yes.

My wife and I want this. We could use a lawyer, and took the first steps, but immediately I started getting annoying vibes around cost and response time. So I asked Claude:

• Of legal instruments, how easy is this one to DIY?

• How does it actually work in my state?

• Could you draft one for me, given the details of our specific situation?

It turns out it’s very easy to DIY, since it’s really just a suggestion anyway and not a contract, so if you follow all the rules nobody’s going to be hunting down typos. The way it works is defined by a really short statute, which I could read and understand. And as far as drafting, Claude found me a template version buried in a pro bono firm’s Wordpress files, that I wouldn’t have found myself, but was clearly the right shape of thing from a trustworthy source.

So rather than trusting Claude’s composition, I just had it modify that obviously valid legal template, massaged it a bit, reread the statute itself to make sure I hadn’t overlooked anything, and, like… I think we’re good, man.

Bright Red With The Shame

I occasionally feel vulnerable, posting to this blog.2

I really feel vulnerable saying, basically “oh yes, I vibe coded a load bearing legal document for my family.” It feels like I am Animorphs-cover-style transforming into some sort of credulous, hyper-neckbearded Goofus, who will soon face retribution for my sins. Don’t I know that LLMs aren’t trustworthy? That they hallucinate? How could I be using them for anything important?

The problem is Nihil Supernum. Somebody is supposed to be above me, giving me a distracted thumbs up and charging me several hundred dollars, and then can I breathe easy. By doing the work myself - and, worse, using modern fallible tools to do it myself - I am waiving my right to breathe easy. I am saying “I’ve done due diligence, and so I stand by this document. If there’s an obvious error, it’s on me.”

That’s the extra cost. AI is pretty eager to speak authoritatively, to present itself as the guy giving you the thumbs up. But in fact, when you use AI, the buck stops with you. You have to figure out how to use what it gives you, to steer and manage it, to make sure it’s not leading you into any troublesome spirals.

But what’s the alternative? Feeling in your bones that every month, you can do more, you can grow stronger and faster and more capable, and that the cost is merely paying attention and taking responsibility, and ignoring that call? It’s easy to say “ah, AI can atrophy your reasoning, so that you stop thinking for yourself” and hard to say “yes, sure, so I will take responsibility not to let that happen.” You can ask AI to summarize a classic novel and then not read the novel, or you can ask it what specific archaic terms mean, in context.3

It’s up to you, man. Ride the wave or don’t. There’s no lifeguard, though, and the flags are flying red.

Here are some cards I got from him today, and why I chose them.

First off, consider the art. This guy is straight vibing. He does not want to do anything. He is waiting for the Earth to reclaim him.

But second, look at his moves? He just draws your deck down, doing almost no damage and refusing to retreat, and then he freaks out when you’re about to run out of cards, and enacts “crunch-time rush.”

We’ve all been there!

Same basic principle. He’s a worm. He’s Done. He will Not retreat. Unless you finally load him up with all the energy in the world, in which case he retreats all the way back into your deck. If only we knew a guy who could dig him back out…

My wife’s favorite Pokemon. Bonus points for fun attack names. Also, finally, a Pokemon that can retreat.

I played Pokemon cards at the local Borders as part of local League events as a child. I’m not proud of it, but I mostly preyed on one specific impressionable kid by inventing a somewhat different game using Pokemon cards that I could always win, and just farmed him for League points. I’m not sure what he got out of it. Maybe I was good company.

Anyway, this Garbodor now represents that aspect of my childhood behavior. And I got so mad at my sister for cheating at Monopoly. I’m sorry.

Also as an adult one time I was curious about competitive Pokemon cards and watched some of a stream, and for whatever reason Garbodor was by far the best card at that time and everyone just spammed a move called “Trashalanche” which I thought was very funny. Probably my version of the game was better, come to think of it. Why should I repent?

You might think I just had garbage on the brain after my fraught Garbodor acquisition, but no, Porygon2 featured heavily in a role playing game I ran after my wheeling-and-dealing-at-Borders era. Specifically, an arc in that game I played with my sister. Who is my nephew’s mom, so it felt appropriate. Yes, this is the sister who cheated at Monopoly. Her son is honorable, though. He won’t repeat our sins.

Okay, now I just had garbage on the brain.

I love this Pokemon. It’s a nose that smells North. That’s right. A nose that’s a compass. No notes.

I also like how this one in particular looks so happy. I don’t even think Rolling Tackle means to do damage. I think he just gets too excited with all the Iron Collecting and tries to give the opponent’s Pokemon a hug.

My wife hates Nosepass. In this, she is wrong.

My nephew was under the impression that since all my cards are super old, I should have a VMax card. Or maybe just a… V card? I don’t know. I don’t know what this is. My hairline is receding. I don’t understand how the art takes up the whole card. Or how a basic Pokemon can have 190 HP.

Anyway, thank you, dear nephew. I am now a resident of the modern era. A bit nervous about this “Lost Zone” business, though. We didn’t have that in my day.

Whoa, does this guy go hard. That he is weak to plants is unaccountable. Perhaps they threaten his gains at the gym. Or give him a feeling of peace that is incompatible, after so many years of brutal conflict, with who he has become?

Anyway I took this card because it reminded me of When I Win The World Ends, which is a great read.

DreamWorks Face.

In the general case, I am not entirely convinced. My own forays with AI assisted software production have benefited a lot from my actual knowledge; the models bark up the wrong tree a few times before you get them there, even for simple stuff. But in the specific case of Chrome extensions, yep, ask and you shall receive.

I’m not aware of a simple canonical post about this, with examples. So here’s mine.

Prerequisites

You’ll need a Claude pro account. They cost $20 a month. Once you have one, setting up Claude Code is really easy. In fact, you can just ask Claude how to do it. Or follow the instructions here.

What To Actually Do

1. Make a folder somewhere on your computer.

2. Open a terminal in the folder (on Windows you can do this in the right click menu).

3. Type “claude” in the terminal (without the quotation marks).

4. Ask for the thing you want, and give Claude permission to mess around in the new, empty folder you just made.

It’s possible Claude will ask for other weird or fishy stuff, if it gets on the wrong track. You can just veto that if so. For most things you might ask for, it’s not much of a risk.

When Claude is done, it’ll tell you how to load your browser extension into Chrome. Which takes about 10 seconds, and is very easy. Then you can test it, and iterate as needed. I’ve done the whole process more than once. It takes like five minutes. And then your web browsing experience is permanently improved.

Examples

The proof’s in the pudding! So here are three extensions I’ve made.

Custom New Tab

I often use pastebin to write out little to-do lists for the day, or keep track of goals. I decided it’d be nicer if whenever I opened a new tab in Chrome, I could immediately see my goals, and had a little text field to write stuff down. Within a few minutes, Claude Code achieved this for me.

Handy! I can hear skeptics chiming in now, saying “yes, but to-do lists are easy peasy sample projects that everyone knows AI can do.” To which my reply is “yes, which is exactly why to ask for one.”

Spreadsheet Populator

There’s a Google Sheet that I regularly add rows to. In the past, I had to henpeck a bunch of specific fields from multiple other tabs, as the last step in some long and boring of record keeping.

I put examples of the relevant webpages in a directory for Claude (you can just download webpages, it’s perfectly legal), and it made an extension where I click a button, and all the data is copied to a table row I can paste. Saving me five minutes a few times a week, into the indefinite future.

Wolfe Glick Censor

I like Wolfe Glick videos. He plays competitive Pokemon, and has high production values. But he’s also an ambitious YouTuber, and apparently what makes people click on videos is the title “I WON TOURNAMENT X.” Thus, I always know the ending in advance of watching one of his videos; if he says anything other than “I WON” I know he’s going to lose.

So I had Claude make a browser extension to do this:

It also hides how long each video is, and how far along I am when watching, so I won’t know by there being five minutes left that he’s about to be eliminated.

There are still some bugs on this one; it only works reliably on his channel. Is it worth the more than five minutes it might take to fix that? I don’t know! But on the very first try it got a lot of the benefit; I can mosey on over to his channel and have a spoiler-free experience from now on.

Go Forth!

It is an exciting time. Think for a few minutes about what has annoyed you online lately. Maybe you can solve it.

With full spoilers, here’s the plot:

A woman’s twin sister has been killed, in connection with a shady tech billionaire and his coterie of hangers-on. So she poses as the recently dead sister and shows up to the next shady tech billionaire get-together, along with a famous detective, to ferret out whodunit. There are lots of machinations and plenty of ridiculousness, but it turns out that the crime itself was extremely simple: tech billionaire has a reputation as a genius, but he’s actually an idiot who stole his idea from murdered twin sister, and he killed her when she tried to block him from releasing an unsafe product to drum up hype.

So, a glass onion. The appearance of layers, but really there’s nothing much to see; the fact that this guy is an idiot is concealed by his wealth and prestige, but once you understand that he’s stupid, the mystery dissolves.

It’s clear wish fulfillment. Twitter lets us see extremely rich and successful people behaving stupidly, and we have various examples in history of men taking credit for women’s scientific contributions. Thus, use the latter to justify the former: these guys who everyone thinks are brilliant are actually idiots, piggybacking on real geniuses being maltreated behind the scenes. Then, you can have a story about someone giving one of these idiots what he deserves, exposing him as a fraud and restoring moral credit to the unsung hero who truly deserves it.

My wife thinks I just don’t like wish fulfillment. I think I do like wish fulfillment. But only the good stuff.

My Beef

Billionaires sometimes say and do stupid things. Many have inherited their wealth, and history shows nothing if not the fact that powerful people’s children often turn out to be huge dumbasses. But even self-made plutocrats often have huge blind spots, or seem extremely dumb in one area even as they dominate in another. Also, since self-made billionaires are automatically famous, most aspects of their lives are scrutinized, and by selective presentation (or just selective paying attention), it’s possible to make any given one look pretty dumb. It’s easy to roast Zuckerberg for The Metaverse, for example, as VR really isn’t taking off right now.

But, like, tech billionaires are dumb in certain ways. The tech founder is an eminently mockable sort of guy, and stuff to mock is all out there on the public internet, for any interested dilettante to soak up. For example:

• They take crazy risks, citing the chance of extreme upside, and undertake projects most people think will catastrophically fail

• They break rules and norms, often up to actual laws, hoping they’ll entrench their positions before regulators catch up

• They jump wholeheartedly into what they see as the bleeding edge, which leads them to make embarrassing mistakes (like falling into AI psychosis)

• They assume they can understand anything from first principles, and fire off into centuries-long discussions with blithe reinventions that leave experts tearing out their hair

• They do drugs like modafinil, testosterone, or good old fashioned amphetamines, to maintain their edge and maintain the optimal risk tolerance, which is sometimes not so optimal in other areas of life

• They do weird sex stuff, often entangled with weird legal stuff

I have no idea how prevalent these behaviors are among the tech billionaire class; perhaps not very! But there are multiple plausible examples for each, which is enough to create a bona fide stereotype. There’s something there to mock! But Glass Onion totally declines to engage with the stereotype, or give us a villain that actually resembles that type of guy. They make their tech billionaire a salesy bullshitter wielding LinkedIn malapropisms, which is a totally different subspecies.

It’d be one thing if I were an insider, and was mad that a mass market movie represented my tribe incorrectly. But I think my read of tech billionaires and their foibles is just from, like, distantly paying attention. Not close attention, even!

When you satirize a type of person well, it simply rings true. I’ve never met the people Arrested Development is making fun of, yet from vague osmosis I totally believe the portrayal, and individual jokes (“What could a banana cost? $20?”) plug directly into my actual beliefs about how those people suck. Or take Silicon Valley, which my parents found very entertaining, but which also pulled off a perfectly serviceable Roko’s Basilisk reference. You’ve got to pay attention to how people really are, or at least how they seem to be, to mock them right. Especially when the story is wish fulfillment, and culminates in knocking such a person down. It might be satisfying to see a tech billionaire laid low by actual tech billionaire foibles. I don’t know! But it’s not very satisfying to crush a sloppy caricature of a different type of guy, with “tech billionaire” on his nametag.

Part of the frustration is that Knives Out 1 did such a good job at this. The rich family seething for inheritance was totally believable, and it was awesome to see them defeated by an honest woman! Knives Out 3 does pretty well too, though I have my quibbles with the supporting cast.

The Good Stuff

So, what wish fulfillment do I actually like? There are at least two options. One is to pay loving attention to your caricatures and stereotypes, such that they feel real. It’s illegal for me to say so in 2026, but Harry Potter is pretty good about this; the bumbling government feels like a real bumbling government, the old money hate group feels like a real old money hate group, etc. Or Ender’s Game. Nobody wants to actually be Ender, but the power fantasy of being the special smartest boy and saving the world with pure gamesmanship shines all the brighter, for the fact that Graff is a credible spook and Bonzo Madrid a terrifying, realistic bully. Or Middlemarch, which I am reading now, which spends pages and pages occupying a cowardly status-seeking doctor’s perspective, as he convinces himself of his fundamentally noble nature before, when the chips are down, he sells out his friend to the local bigshot. The author is roasting him, and setting him up later to have a life way out of line with his stated values, letting the reader feel superior. But it’s done with affection, treating the character as a well-loved plaything. Oh, you.

The other option is to just go totally bananas, and invent new types of guy. It helps to invent an entirely new type of world for this, too. Aragorn’s story does tap into fundamental archetypes, but his actual situation is totally untethered from reality. (Who among us has not been granted conditional semi-immortality by our main squeeze, but also lusted after by the princess of the horse girls, and then needed to go to the ravine to recruit various ghosts who owe our family a blood debt?) It doesn’t matter. You get into the Lord of the Rings rhythm, and Aragorn’s many, many victories feel like your own.

Nor does art have to be highbrow to do this well! Trails Through Daybreak does a great job, and it’s extremely pulpy. The protagonist has the made up job of “spriggan” and is obsessed with public baths, his pickup truck, and candy, which seem like interests selected from a bingo card. But the world he exists in is also very weird, with more factions and weird bureaucracy than you can keep straight, and because both the world and the character are well constructed it’s easy to settle into his perspective. Everyone has a crush on Van Arkride, which is the oldest wish fulfillment in the book, but because he’s a three-dimensional character and he fits into the world properly, this is actually believable. Which is my whole ask, really. Give me wish fulfillment, sure, but give it the weight to suspend my disbelief!

All this to say, my social sphere is normies. Or as I prefer to call them, ordinary people. They aren’t ordinary in the sense of being unremarkable or unimpressive; I find plenty to be impressed by! But they aren’t hyper-analytical, hyper-ambitious, or readers of rationalist blogs. (Other than this one, because I browbeat them. Hi, friends!)

I’ve also got another social sphere. I spend a lot of time on LessWrong, in both lurking and editorial capacities, and sometimes head out to Berkeley to soak up the hyper-analytical, hyper-ambitious, in-one-sense-my-people vibes.

So when I read Jenn’s recent banger, I wanted to respond.

Jenn’s Recent Banger

Jenn’s post is called In My Misanthropy Era. I summarize it thus:

Jenn has deeply rooted beliefs about the fundamental value of all people, in part inculcated by SJW Tumblr. But she also has an inner edgelord. She’s been reading the great books, which has been fueling the inner edgelord, since e.g. Schopenhauer has not-so-nice things to say about the common man. She doesn’t want her inner edgelord to run rampant, so she went to a general public philosophy meetup to get a taste of what normal people are like. She didn’t like it, and found herself fucking with them rather than genuinely engaging, which made them like her more and her like herself less. Hence, a dilemma: she wants to empathize with ordinary people, but actually interacting with them makes her respect them less.

I am sure my compression does some violence to the source, so if you’re a blog post addict like me, read the whole thing here.

My Perspective

I think I understand what Jenn is getting at. At a pool party this summer, I met a fellow tech worker. He is a software engineer. He’s a smart guy, he made an impressive and touching custom cake for his girlfriend, and he’s better than me at the only video game I know we’ve both played. No shade. But when I asked him if he used AI for work, and he replied that he’d only really tried it once, I felt at sea. I cannot even imagine this. No matter how much I use AI, I always feel like more and better usage would be a clear professional boon; I don’t begrudge myself my current level because I have other priorities and time is finite, but anything less than once or twice a week just sounds crazy. In software?

Much like Jenn had a negative experience seeing people fail at basic reasoning, I had a negative experience seeing someone in my field not keeping up to date with the latest tools. Specifically, my mind reached out for judgment, and found it wasn’t sure on what level to judge. Here is a person with fundamentally different latent assumptions, with totally distinct mental habits and strategies, at least in the domain of work. I can’t expect him to work like I do, nor can I totally dispense with my judgment, because ultimately I do believe that it’s correct to stay on top of AI, and that not doing so is, in expectation, a quite costly mistake.

So the result is this strange dissociation, a failure to engage that ping-pongs between guilt/shame, a vague feeling of unearned superiority, and thrashing rationalization to close the gap. All this can happen in the span of just a few seconds, then the conversation moves on, and it’s like it never happened.

Do you notice how I said “in expectation” two paragraphs back? Most of my local friends do not know what that means. Which is another example of a fact that is hard to fit into my brain. Because I stress out about not remembering off the top of my head the difference between UDT (the U is “updateless,” right?) and CDT, and that stress contains an implicit assumption that things like expected value are absolute table stakes. I have formed (much of) my thinking into a shape, and there are people out there who have not done this project at all, and in fact have formed their own thinking into different shapes rather foreign to me, where I am ignorant even of their most primitive basics. I do not, for example, know the difference between knitting and crocheting. Perhaps my sin is greater, as I have never seen a crocheted gift fail to spark delight.

But Also, My Other Perspective

You might think that when I go to Berkeley, and am among fellow rationalists, I am free from alienation. If you were right, that would be very bad news for my pocketbook, as I would have to move to one of the most expensive locales on Earth. Luckily for my pocketbook, however, you would be wrong. This is where I differ from Jenn, I think; I do spend enough time in the rationalist bubble to have them rub off on me, and to viscerally judge things against their value system. But I also spend enough time in the normie sphere, that I make visceral normie judgments, too! And wow, are there a lot of normie judgments to make in Berkeley.

Like, my most recent visit to the bay, I encountered someone I’ve worked with more than once online. He’s also dating a friend, and we’ve met in person before. He didn’t recognize me, which, no problem. But then he also commented that I “look different than he thought.” Which was slightly unnerving, so I asked how he’d thought I looked before. He just said “worse.”

This guy was one of my hosts during my visit! I was spending a month on his turf, and he immediately revealed that he incorrectly thought I was super ugly, right up until that moment. He did not soften or modulate this observation in any way. What?!

Or I was sitting around a fire, with people talking about genetically engineered babies, and someone ventured that it was perhaps morally obligatory to genetically enhance your children via embryo selection, since you could select for traits that would make them better off. This from the same crowd that’s often worried about low fertility, with no apparent thought to the contradiction; (most) people don’t want to do IVF when they don’t have to!

Or take the frequent torrent of rationalist blog posts about the horrible clawing distracting power of devices and algorithms. Digital minimalism is well and good, and being intentional about devices is fine, but most normal people I know are perfectly fine with their level of YouTube, Instagram, etc. consumption. The idea of fretting about it intensely is just like… weird. Extra. Trying too hard. Because most people aren’t ultra-ambitious, and the opportunity cost of a few hours a day of mindless TV or video games or whatever just doesn’t really sting.

Like the pangs coming from the other direction, these moments of alienation are of a common shape. I don’t feel like the rationalists are making a definite error, but rather feel like there’s a collection of vibes and pre-theoretical socially-transmitted inclinations that drive certain patterns of behavior, and those patterns of behavior are, to a certain set of inclinations that I’ve learned (that of normal people), off-putting or insane.1 You see the most ambitious and successful people on Earth, dazzling superbeings of metric-based achievement, simultaneously the most worried about their own precarity along every dimension, and like… what? How? Make it make sense.

A Solution?

No, not really. I guess the value of my perspective to Jenn (I flatter myself) is that truly integrating with ordinary people will in fact dual-boot your social operating system, such that you can be vaguely unnerved by each partition from the perspective of the other. It sounds like some cutesy both-sides LLM slop, but in my experience it is just literally true.

Also true, though, is the fondness. My normie friends think something like Inkhaven, an expensive blogging retreat that requires you to take substantial leave from your job, is totally insane. I can share in their humor, but I also find Inkhaven wonderful and worthwhile, and the fact that it seems insane from one perspective throws its grandness into sharp relief from the other. Likewise, the hours spent in meandering conversations about video games with my unambitious friends is all the sweeter for the knowledge that somewhere else in the world, the same conversation would (and did) feature some guy suddenly and explicitly introducing annoying norms to guide the conversation his way, thus ruining it, for me, completely.

Not everyone needs to multiclass. Jenn may be best off living entirely in her chosen and well-curated bubble. But I don’t think an evening among a slice of humanity is enough to learn to live like them, and I don’t think anything less than learning to live like them, wholeheartedly, can break alienation.

You may notice that T is off the board. It’s one of my baby’s favorite letters to remove. It also can stand for testosterone, which my baby has partially removed. From me.

Well, Maybe

If you’re not a link clicker, the short version:

New fathers showed a 26% drop in morning levels and a 34% decrease in evening levels, compared with single nonfathers, whose morning and evening testosterone went down by 12% and 14%, respectively (a decline attributable to the passing years). The study also revealed that testosterone levels were lowest in men who reported spending the greatest amount of time spent caring for their children.

Studies are coarse instruments, and there are many other factors setting human hormone balance. But I have changed since becoming a parent, in ways consistent with lower T. Shall we name them?

Less Competitive

I like board games for many reasons, one that I’m good at winning them. But when I’ve played such games post-baby, I feel less joy from victory and less sting from loss. Moreover, my desire to play board games is reduced; it feels a bit like some of my motivations (socialize with friends, mental exercise) are intact, while another (crush rivals) is more-or-less removed.

I’ve similarly chilled out when it comes to video games; I haven’t been making it to Smash Bros tournaments since baby was born, and that’s only partly from being busy. Like with board games, the activity itself sounds fun, but the struggle for rank is not much motivation.

I still care about the same about making money, but my focus in that arena is much more on how I can spend the money for the benefit of family, and less on abstract number-go-up stuff.

More Socially Self-Conscious

One of my great pleasures in life is holding forth. I am, after all, a blogger. But in the last year, I’m timorous about going on too long. I still expound at length for my wife, particularly when I think she’ll enjoy it, but in conversations with friends I feel more internal pressure to cede the spotlight. Most conversations I will naturally think “hmm, have I listened enough?” This hasn’t been a deliberate area of growth, either! I’ve always known it was something to work on, I’m just better at it now.

I noticed this most strongly when I did a long, reflective speech for my family on my birthday, and it felt like a rare and precious indulgence. People used to say I’d make a good cult leader. I don’t think they’d say so now!

More Minivan Accepting

We got a teensy SUV recently, and I wouldn’t call it gender euphoria per se, but driving that thing around makes me feel like a million bucks. It’s just such a family car, absolutely begging to be decked out in car seats and, later, sports paraphernalia. It makes me long for a ponytail. Go team!

(I used to find that shape of car kind of unbearable. I tell myself I’d still cringe at having a bigger one, but, well, give me a few years.)

More Flexible

I’ve always had a really hard time with changes to my environment. I still do. But I’m much less rigid than before. I’ve gone from the person in my marriage whose rancor and panic prevented home improvements, to the person in my marriage browbeating my wife into buying felt organizers.

I’ve also been a night owl forever. But I decided to start waking up at 8am every day, two hours before my prior wakeup time,1 to make it easier for me to spend the night with baby three nights a week. Magically, today, I woke up at 8am with no alarm.

Or, when I felt like rebranding this blog, or pausing habit tracking, I just did it. A couple years ago, I might have been too attached to how things were, even if it didn’t fit me anymore.

Happier

I just feel better, overall. Even with the sleep deprivation. I will attribute most of this to the baby herself, who is very sweet. But I also feel vindicated. I’ve never been obviously all that feminine, but I have always had this itch that boy stuff - or at least certain boy stuff - is a drag. And lo and behold, turn the T knob down a few notches, and I feel more like myself.

Ever since I was a small child, I’ve been bummed out that I couldn’t actually bear children. It just seemed like making a person was obviously the coolest possible thing to do, and it was insane to me that, at random, only some people could do it.

That envy is quieter now. The baby is right there! She’s born! It’s time to do the work.

On paper, everyone knows this, and it’s called inflation. “Penny candy” used to be a literal description of certain physical objects, and “dollar menus” were popular at fast food chains in my millennial memory.

Viscerally, though, there’s a temptation to say phooey, and resent the fact that a bigger number is required to meet basic needs. I posit that this temptation gets its spark from human nature; we don’t respond very well when things get better on one dimension (rising salaries) and worse on another (rising expenses), even if the combination is obviously good on paper (salaries rising faster than expenses). But it gets its tinder from social contagion, and thereby becomes a bad habit.

Socially, if you comment on how stuff is expensive these days, you can’t lose. You’re pointing out a problem that everyone has, so your interlocutor is almost guaranteed to be sympathetic. If you’re struggling financially, you’ll feel less alone. If you’re doing fine in that department, you avoid seeming out of touch. So it’s pretty easy to install a habit where, whenever you buy anything, how much it cost becomes an automatic object of small talk. Unlike the weather, though, or other benign topics, “stuff is pricey” mostly serves to make people feel bad. And in aggregate, this is actually a problem, since it contributes to people thinking “the economy” is bad any time prices are rising, and when people think “the economy” is bad they want to rock the boat and make facially insane political choices.1

Doing Your Part

I can think of a few ways to break the cycle.

The easiest is to mention prices less. Once you understand that there’s a silly bug in human cognition, such that it’s hard to accept that things are nominally more expensive every few years, then the topic is pretty boring. Cheap fast food items are $5 rather than $1. Eventually they’ll be $10, then $20, and then you die. So it goes.

A second option is to delight when things are cheap. I try to trigger childlike wonder in myself whenever I’m in a grocery store, at least once. This place really exists? I can get that much rice for that little money? There are more ways to meet caloric needs on a shoestring budget than grains of sand would fit in the universe, all available in a five minute drive. Or take video games. $60 for 60 hours of high-quality entertainment? If you’re busy, you can access gamer bliss for months on the cost of a few Chipotle orders. Nor was it always thus! Yesteryear’s games were, in real terms, dramatically more expensive. Also yesteryear we were children, and at the mercy of our parents. And so on. Gratitude is always available, should you look, including in the realm of prices and values.

The hard part, of course, is when other people complain about prices to you. I think the trick here, if you believe there’s a bad social contagion afoot, is not to offer much reward. You shouldn’t contradict the person; they may be struggling, and what’s the point of fighting about this? But you can change the subject, say something bland and noncommittal, or take a stoic line, like “inflation do be that way.” It’s a stupid hill to die on, stuff not actually being that expensive.2 But no need to feed the beast.

I raked the leaves for my neighbor while he was out of town. But when I returned, it turns out that my neighbor is stressed out by the presence of a leaf-pile that he has to get rid of, and wouldn’t have been stressed by a diffuse carpet of leaves. If he’d liked it, I’d be providing value and thus it’d be in his interests to find a way to provide value in return. But since he doesn’t, there is no such obligation, and in fact since I’ve provided negative value I should probably make it up to him.

If this were the software the human brain - okay, fine, my specific human brain - were running, then in the aftermath of such a leaf debacle I’d be looking for other ways to help my neighbor, and paying closer attention to his likes and dislikes to make sure the next attempt goes better. And I guess that does happen. A little.

But another thing that happens is insecurity. Once my brain internalizes that I’m not useful, it starts looking for other strategies. Such as “sickly convalescent.” This isn’t conscious, and it certainly isn’t endorsed! But I do notice that when I’m doing a great job in reciprocal relationships and feeling like a pillar of my community, I interpret my health more optimistically. Whereas when my gifts keep bouncing back and my efforts to help end in failure, I interpret every little physical problem as a significant issue. If I’m a little bit sick, after all, I can’t be expected to live up to ordinary interpersonal standards!

It should go without saying that this is a horrible adaptation; being sick feels bad, and the actual ways to dig out of the hole are:

• Succeed at social bids

• Give up on them

Or some combination. By feeling sick (or depressed, or anxious, or whatever), I (and perhaps you) avoid both options and prolong the problem indefinitely. I could fulfil my nebulous and challenging obligations rather than failing to do so, of course, that’s the sort of person I am, I just happen to be under the weather, or having a mental health issue, or whatever.

The fix is delicate attention to the object level, a deep reservoir of self-love and self-knowledge (but not narcissism or rumination, natch), and acceptance of uncertainty and change. Oh, and sufficient exercise and sleep. Ho hum. That’s how you fix everything! Aren’t there more exciting strategies available?

Well, one would be to deny all potentially somatic illness, and treat all feelings of helplessness or weakness as malingering. This seems like a great way to not catch literal cancer until it’s stage four, or to end up so majorly depressed that you lose your capacity to hold down a job. This is the preferred strategy of the protagonists of Westerns or characters played by Tom Cruise. It works sometimes, but the risks seem pretty bad, because illness (even mental illness) can alas be real.

Another would be to become a spoonie, just permanently sick and obsessed with the experience of being sick, and socialize exclusively with other sick people who take for granted that until further notice, nobody should expect anybody to contribute anything. Also not great.

A third option - so very often my favorite option - is to think about it more. To be more intentional and systematic about trying to be useful, and to reflect explicitly on when and how things go badly or conditions change. If you could do this entirely in your own head, it would be perfect, and we could all be maximally useful superbeings. But dagnabbit, it requires actually checking with the people you’d like to be helpful to, and hashing things out the right amount, rather than too little (emotionally closed off) or too much (bugging and fussing and meddling).

A lot of the battle for me is just noticing at all. Having it as a hypothesis that if I’m paying more attention than is warranted to my sinuses, I might just try knocking out a chore or two, and that my body saying “oh, I couldn’t possibly” is at best ambiguous evidence that that’s true.

Share