Links
Andrej Karpathy (OpenAI founding member, Tesla veteran, major contributor to today’s AI frameworks) made a really good video going over the practical use of large language models. Dense, clear, highly accessible, but doesn't shy from complexity. Think of it as an excellent two-hour lecture.
We Live Like Royalty and Don’t Know It is the introduction to a great new series of essays called How the System Works by Charles C. Mann.
The great European cathedrals were built over generations by thousands of people and sustained entire communities. Similarly, the electric grid, the public-water supply, the food-distribution network, and the public-health system took the collective labor of thousands of people over many decades. They are the cathedrals of our secular era. They are high among the great accomplishments of our civilization. But they don’t inspire bestselling novels or blockbuster films. No poets celebrate the sewage treatment plants that prevent them from dying of dysentery. Like almost everyone else, they rarely note the existence of the systems around them, let alone understand how they work.
The Economist has a pretty good write-up on the danger of relying on OpenAI’s Deep Research, which basically matches my experience. It misses, however, how tools like this can help you overcome little barriers in life.
For example, this morning I used it for 2 things and it saved me tons of time:
- I wanted to find a company that comes and picks up stuff you want to get rid of and either recycles, donates, or sells it. And wanted to avoid a place that scoops up valuables and takes the rest to the dump. In about 10 minutes it found remoovit.com, which is pretty much exactly what I want.
- I was not directly threatened by the Eaton fire but we are nearby and left because the air was filled with ash and because 🤷! But we soon realized we didn’t leave the house with everything we should have. We took passports but forgot birth certificates, etc. So I asked Deep Resarch to create a checklist for a “grab and go disaster box” with all life’s critical paperwork and valuables. Again, in about 10 minutes I had a really good starting point which I printed out and will start building. Now, would it have been faster to google around and find an existing checklist? Maybe. But this was tailored to my exact needs and requests and I spent those 10 minutes making coffee and talking to my kids instead of googling.
These tools make easy the things you’re currently putting off because there’s a “toil barrier” you don’t have the energy or resources to overcome. Those things can seem trivially small but small gains add up over time.
Nike didn't invent poverty. It got there a long time before Nike -- or America, or the nation of China for that matter. Poverty is the natural state of the human condition.
For most of human civilization, we've had societies where almost everyone was desperately poor and then a few people were marginally less poor. (Poor Americans eat better and have better healthcare than European aristocrats from a few centuries ago. Plus, they don't defecate in buckets they sleep next to, which is a plus.)
The newfangled innovation brought about by capitalism isn't poor people: its rich people. Rich people who are so overflowing with money they buy frivolous shoes, enabling some Chinese peasant to get off the farm. Chinese peasants are pretty much in favor of getting off the farm -- it gives their kids a much better shot at surviving to the age of 5, and means you don't have to deal with famine.
A pretty interesting set of observations: 27 Insights From Three Years of Bismarck Brief. Partiularly the importance of functional institutions in so many different ways.
A live player is a person or well-coordinated group of people that is able to do things they have not done before. A dead player is a person or group of people that is working off a script, incapable of doing new things.
What fully automated firms will look like — “Everyone is sleeping on the collective advantages AIs will have, which have nothing to do with raw IQ: they can be copied, distilled, merged, scaled, and evolved in ways humans simply can't.”
Why is it that talking through a problem is much easier than writing it down? Why is remote, asynchronous, working great for some things but terrible for other things?
Why is conversation so easy? suggests the answer might be that humans are “designed” for communication and collaboration “because it is through dialogue that humans learn to speak in the first place”.
[…] dialogue supports interactive alignment of linguistic representations. In turn the alignment of representations has the effect of distributing the processing load between the interlocutors [conversation partners] because each reuses information computed by the other.
The Unbearable Slowness of Being is a paper that explores the large discrepancy between the brain’s sensory input capacity (around a billion bits per second) and the brain's much slower processing capacity (about 10 bits per second).
This is fascinating to me because it aligns with the ideas in Daniel Kahneman's brilliant book Thinking, Fast and Slow, which explores the same ideas from a more qualitative perspective.
It also strikes me that this validates much of what Bret Victor (et al.) is exploring with Dynamicland. If the brain works as this paper suggests, moving computation from the narrow constraints of a screen designed for serialized processing to an environment that taps into our richer, parallel sensory channels isn’t just interesting — it’s a fundamentally better paradigm because it aligns with the way our brains are built to process information.
This essay, called "History is in the making", captures something I've been trying to express to my friends for the last 10 years. Politics matters a lot less than people think, and is not worth the time we spend getting worked up about it.
An understanding of the past in which not just our intellectual successes but our technological breakthroughs occupy pride of place would be very different from the political one that dominates now. Instead of politics and war, and the growth, rise, and decline of states and empires being the focus, the central story would rather be one of human cooperation and inventiveness, innovation and scientific and technological progress and discovery, and the improvement in human well-being than the deeds (often diabolical) of those with power.
The important figures would no longer be rulers, generals, prelates, and revolutionaries but scientists, entrepreneurs, and businessmen and -women. It is people like Joe Sutter, Juan Trippe, and Alexander Graham Bell who would be remembered.
If it is the case that human ingenuity solving problems is the most potent force in history, why do so many still fixate upon politics, wars, and revolutions?
I've lived in LA for 20 years and this is by far the best description of Los Angeles I've ever read.
The whole thing is ridiculous. It’s the most ridiculous city in the world – but everyone who lives there knows that. No one thinks that L.A. “works,” or that it’s well-designed, or that it’s perfectly functional, or even that it makes sense to have put it there in the first place; they just think it’s interesting. And they have fun there. And the huge irony is that Southern California is where you can actually do what you want to do; you can just relax and be ridiculous.
“These effortless pleasures, these ready-made distractions that are the same for every one over the face of the whole Western world, are surely a worse menace to our civilization than ever the Germans were. The working hours of the day are already, for the great majority of human beings, occupied in the performance of purely mechanical tasks in which no mental effort, no individuality, no initiative are required. And now, in the hours of leisure, we turn to distractions as mechanically stereotyped and demanding as little intelligence and initiative as does our work. Add such leisure to such work and the sum is a perfect day which it is a blessed relief to come to the end of.” — Aldous Huxley
Farewell Etaoin Shrdlu is a wonderful documentary about the last day the New York Times was printed on hot metal machines. I found it mesmerizing throughout. How wonderful that someone had the foresight to document this.
“With the collaboration of Carter & Cone, Monotype, Microsoft, and Morisawa, Type Network presents the Matthew Carter Collection.”
NASA‘s “worm” logo is beautiful. And after decades of trying to get rid of it, NASA is embracing it once more. You can download the original 1976 graphics standards manual, and you can even buy a “remastered” print edition of the manual.
Frank Chimero writes about plainness and sweetness, reviewing the Untitled Sans / Serif fonts from The Klim Type Foundry.
Here is a reminder: the surest way forward is usually a plain approach done with close attention to detail. You can refine the normal into the sophisticated by pursuing clarity and consistency. Attentiveness turns the normal artful.
Plain design frequently uses defaults: Helvetica or Times New Roman, default link blue, A4 or letter-sized paper. These choices are loaded. Not only do they carry the original intention of the default materials, they are also saddled with the cultural meanings associated with choosing them. If a designer uses Helvetica, it suggests one of three things: they are ignorant of the cliché, choose to look past the cliché, or want to ironically play with the cliché. None are satisfactory if one wants to create plain design without wading into the issue. This is why we must continually invent new, plain materials. As soon as a popular and plain material becomes burdened with too many associations, we can drop the baggage and move on.
The Inter typeface family, version 4 has been released. I use Inter for overvale.com, and I love it with a passion. Looks great in printed media as well.
Inter is one of the world's most used typefaces with applications ranging from computer interfaces, advertising & airports, to NASA instrumentation & medical equipment.
Untitled Sans / Serif fonts from The Klim Type Foundry
Untitled Sans and Untitled Serif are quotidian typefaces. Untitled Sans is a plain, neogrotesk sans validated by the ideas of Jasper Morrison and Naoto Fukasawa’s Super Normal project. When Untitled Sans was nearly finished, I applied the same principles to the creation of Untitled Serif, which is drawn from the old-style genre of types: the post-Caslon, pre-Times workhorses offered by almost every metal type foundry of the time. Untitled Sans and Untitled Serif are related neither by skeleton nor a traditional aesthetic connection, but by concept only.
The year is 1518. Mexico-Tenochtitlan, once an unassuming settlement in the middle of Lake Texcoco, now a bustling metropolis. It is the capital of an empire ruling over, and receiving tribute from, more than 5 million people. Tenochtitlan is home to 200.000 farmers, artisans, merchants, soldiers, priests and aristocrats. At this time, it is one of the largest cities in the world.
Today, we call this city Ciudad de Mexico - Mexico City.
Not much is left of the old Aztec - or Mexica - capital Tenochtitlan. What did this city, raised from the lake bed by hand, look like? Using historical and archeological sources, and the expertise of many, I have tried to faithfully bring this iconic city to life.
Your Organization Probably Doesn't Want To Improve Things – Ignoring Is Bliss:
I have a friend that was denied a raise after working at a company diligently for five years - there was no budget, he was told. He was upset, and ended up spending a year traveling the world on sabbatical.
While he was away, a new position opened up in the same department, for the same job title - it was a permanent role, theoretically exactly the same job he was doing in an identical team with the same managers. They gave me that job, all the money he asked for, plus another hefty sum on top of it.
Management saying there was no budget meant nothing - it is purely a thing they are forced to say, and if the person saying it was not particularly insightful, they even believed it. The only thing an organization can do that means anything is actually solving your problem, and any words to the effect of "we'll address your concerns soon" or "we just need to do some planning around this", etc, should just be ignored wholesale.
Super Normal – Jasper Morrison & Naoto Fukasawa:
Design, which is supposed to be responsible for the man-made environment we all inhabit, seems to be polluting it instead. [...] Encouraged by glossy lifestyle magazines, and marketing departments, it’s become a competition to make things as noticeable as possible by means of colour, shape and surprise. The virus has already infected the everyday environment. The need for businesses to attract attention provides the perfect carrier for the disease. Design makes things seem special, and who wants normal if they can have special?
What Moneyball-for-Everything Has Done to American Culture:
The Nobel laureate particle physicist Frank Wilczek once said that beauty exists as a dance between opposite forces. First, he said, beauty benefits from symmetry, which he defined as “change without change.” If you rotate a circle, it remains a circle, just as reversing the sides of an equation still reveals a truth (2+2=4, and 4=2+2). But beauty also draws from what Wilczek calls “exuberance,” or emergent complexity. Looking up at the interior of a mosque or a cathedral, or gazing at a classic Picasso or Pollock painting, you are seeing neither utter chaos nor a simple symmetry, but rather a kind of synthesis; an artistic dizziness bounded within a sense of order, which gives the whole work an appealing comprehensibility.
Macintosh macOS Resources and other handy things from NASA’s Astrophysics Science Division. A bit old-school, but a venerable selection. A bit like NASA itself I guess.
A very interesting archive of historically significant computer science papers, hosted by Brett Victor.
From McKinsey's Quarterly Interview with Jony Ive:
Founders are creatives by definition — they had the idea behind a company.
To make that idea material and relevant, they need to work with a collection of people. But the nature of ideas and the creative process is so particular and unusual. It’s an activity that doesn’t naturally or easily sit within a large group of people.
When you gather a large group of people, they generally want to be able to relate to one another and to be sociable. But any process that is unpredictable does not sit comfortably or naturally in a large group setting. So people come to value activities that are predictable.
This doesn’t mean that you would willfully want to undermine those activities that are predictable. But that is the nature of creating. And one of the things I realized is that when you’re trying to create in the context of a large group of people with a whole range of different expertise, people tend to want to gravitate to those attributes or characteristics of a product that you can measure easily. If you’re trying to relate to a group of very different people and you want to appear sociable and engaged and connected, it’s much easier to talk about something that you can measure with a number. That’s why we choose to talk about schedule or cost or speed or weight. Given our very different backgrounds, that’s a comfortable and easy conversation. I completely get it.
But there’s an insidious problem with that. There is a dangerous assumption that we’re having these conversations because they’re the only important ones. But the really important conversations and preoccupations and concerns are very hard. Because you can’t assign a system of numbers to make the relative judgments that need to be made.
I used to think that this kind of conversation was a personal attack, or an affront to the practice of creating, but I’ve come to learn, over the years, that it’s just a natural, very predictable consequence of having larger numbers of people gather together to talk about developing something.
Tomorrow & tomorrow & tomorrow:
The gears of commercial networks are surveillance systems built on structures that elicit a continuous stream of confessions made public. Confessions in public become testimony; testimony summons congregations. We raise our voices in defiance or affirmation, knowing there will be consequences we don’t understand. The databanks grow.
But we built the machines around our weird amygdalas and then we went inside them and now the machine is no longer confined to a stack of software + policy + vibes; we carry it in ourselves. We haunt each new place we enter.
The Jevons Paradox states that an increase in resource efficiency results not in diminished consumption but in an over-all increase.
The Linder Theorem — the more productive we are the less we are willing to invest in chronologically expensive things — even though, objectively, we have more time for them than ever.
The Apollo program was an incredibly powerful Enabling environment, but it did not emerge from a project aiming to give scientists lots of great opportunities for personal growth. Rather, it was about putting people on the moon (and, er, saving the world from the Soviets). The enabling environment was a byproduct of that deeply meaningful effort.
Likewise, when Pixar created its revolutionary animation tools, many teams had been working on computer graphics for years, but Pixar’s systems emerged from a zealous pursuit of a storytelling dream: Pixar’s movies and technology development act as coupled flywheels.
Cathedrals! University research labs! Mathematica! They all follow this pattern.
By 2100, the Lagos-Abidjan stretch is projected to be the largest zone of continuous, dense habitation on earth, with something in the order of half a billion people.
Instead of your Life‘s Purpose — “Rather than struggle to discover a purpose or vocation, we [should] become people who can recognize and exploit opportunities to create meaning as they arise — resourceful and audacious people who live adventurous lives.
[…] the economists would be amazed. For this design — so far beyond the peak of their world’s powers — was not a gift for a King or Queen, but clearly just a child’s toy. And not just that. The very banality of the toy, the artlessness of the sculpting and the way the paint doesn’t even quite line up with the contours of the doll’s face, prove that this is not the toy of a prince — but of poor child. And they would understand that the people of our time are so wealthy, so powerful, that every one of them has access to machines with thousands of parts working in concert, and that it is less effort to build such a wondrous machine than to simply paint a doll’s eyebrows in their right places.
The Age of Average argues that from film to fashion and architecture to advertising, creative fields have become dominated and defined by convention and cliché. Distinctiveness has died. In every field we look at, we find that everything looks the same.
U.S. Stock Market Returns — a history from the 1870s to 2022
A Stanford Medicine study finds no “connection between the age children acquired their first cell phone and their sleep patterns, depression symptoms or grades.”
The clipboard on macOS doesn’t work the way you think it does.
A pessimistic stance is a safe one. There is often little to lose. Optimism is risky. It exposes us to failure. In fact, repeated failure is a given. That’s why it looks dumb from the outside.
Thor: Love and Thunder used a wild lighting rig to capture a sequence where lights were supposed to rapidly move around the set while filmed in slow motion. The technique is to use a high-speed camera whose frame-exposures are synced to store lights. This technique builds on one used in Thor: Ragnarok. It was developed by a company called Satellite Lab.
Over the last 20 years the average age of top-billed actors has risen significantly. Old movie stars are not being replaced by new ones at a sustainable rate.
Life is a constant struggle to understand what is happening and what to do about it. We just make it up as we go. You cannot prevent yourself from struggling, it is inherent in life. One of the most important things you can do is help other people with their struggles.
I feel that starting from when? I don’t know, from the ’50s or the ’60s or ’70s — at a certain point, the experience of culture — whether it’s painting or something else — the nature of the experience shifted to something that had more in common with journalism than with what we might call “an aesthetic experience,” that works of art were thought to be about something. What that “about” was, or is, is something that could be more or less easily grasped. It’s the aboutness which, for me, is a short-circuiting of the art experience.
People don’t realize that every single advanced manufacturing company — whether they’re making rockets, satellites, jets, drones, or energy for climate change — outsources about 80% to 90% of their custom parts. This isn’t like automotive where you’re printing 1,000 widgets to go on a Ford. This is ‘Hey, we need five crazy complex-geometry parts that are going on a rocket.‘ It’s about $40 or $50 billion in spend that’s coming out of these space, defense, and other advanced manufacturers going into a domestic supply chain, but it’s being routed through 3,000 or so small, owner-operated high-precision machine shops. So, in aggregate, it’s a huge industry, but it’s incredibly fragmented. You’ve really got 98% of that base doing less than $10 million in revenue.
A heresy is an opinion whose expression is treated like a crime — one that makes some people feel not merely that you're mistaken, but that you should be punished.
The global market for plastic caps and closures is worth $40 billion per year — and growing.
This study — An Efficiency Comparison of Document Preparation Systems Used in Academic Research and Development — (flawed by a small sample size as it might be) rocked my world because it shows how experts can fall into the trap of optimizing for their own requirements/wants, rather than for, you know, results.
We show that LaTeX users were slower than Word users, wrote less text in the same amount of time, and produced more typesetting, orthographical, grammatical, and formatting errors. On most measures, expert LaTeX users performed even worse than novice Word users. LaTeX users, however, more often report enjoying using their respective software.
A deep dive into macOS’s Character Picker (the thing you use to pick emojis when typing).
Why does DARPA work? — very long, but with a great summary.
Food isn’t about Nutrition
Clothes aren’t about Comfort
Bedrooms aren’t about Sleep
Marriage isn’t about Romance
Talk isn’t about Info
Laughter isn’t about Jokes
Charity isn’t about Helping
Church isn’t about God
Art isn’t about Insight
Medicine isn’t about Health
Consulting isn’t about Advice
School isn’t about Learning
Research isn’t about Progress
Politics isn’t about Policy
The economics of college vs. the economics of private tutors
How can you possibly justify a $200,000+ college expense? How can you justify a $100,000+ college expense?
This is not necessary.
The average tenure hopeful adjunct makes $40 an hour. If you were to employ her as a private tutor at the cost of $60 an hour, and had four hours with her a week, and did that for 14 weeks (that’s the length of an average college course folks) that is about $3,400.
Were you to employ three such professor-tutors, that would be about $10,200, or a bit over $20,000 a year. In four years you would have racked up $80,000 in costs. But this is still $30,000 less than the total for the ‘cost conscious’ universities. It is a quarter of what you would pay for Trinity.
Remember: this $80,000 is for private tutoring, where individual attention would give you far and away a better and more thorough education than the 300-kids-in-a-lecture-hall style of classes that dominate undergraduate education today.
But it can get even cheaper. Let’s say you take the general principle of group classes from the university. Say you can find four other people to take all of these other classes with you. Just four. Well that equals out to $680 per class, or $16,000 a person for four years of classes.
To be fair, add in $1,000-$2,000 for textbooks and a subscription to JSTOR, for a total of about $17,000 to $18,000 for four years.
Glossary of landforms: Dell, hollow, cut, plain, plain, terrace, atoll, etc.
How Method Studios created the lunar scene in Ad Astra using infrared photography.
Super Normal by Jasper Morrison & Naoto Fukasawa.
I love painting. You start by thinking you’ll get one assistant and before you know it you’ve got biographers, fire eaters, jugglers, fucking minstrels and lyre players all wandering around. They’re all saying they aren’t being paid enough and they all need assistants. Then one night you ask the lyre player to play for you and they say: ‘My lyre is all scratched up and I did ask for a lyre technician but you said not yet and if I had one I could come and play for you now.’ So you’ve got to have a lyre technician and then you better get him an Uber account too.
Marc Andreessen’s guide to Personal Productivity
The Stockdale Paradox: “You must never ever ever confuse, on the one hand, the need for absolute, unwavering faith that you can prevail despite those constraints with, on the other hand, the need for the discipline to begin by confronting the brutal facts, whatever they are.
Fast: an astonishing list of big things accomplished quickly.
An interesting review of the sources of inspiration for Frank Herbert’s Dune.
Public companies in the US are obligated, by law, to maximize the economic value of the company. That’s why hostile takeovers exist. By adding a legally binding set of ethical rules, companies might be required to to more than just add value for shareholders.
I first read about the St Nazaire Raid in Churchill’s The Second World War, and was amazed that I’d never heard of this dramatic, interesting, story.
One of my favorite Steve Jobs stories:
Jobs imagines his garbage regularly not being emptied in his office, and when he asks the janitor why, he gets an excuse: The locks have been changed, and the janitor doesn’t have a key. This is an acceptable excuse coming from someone who empties trash bins for a living. The janitor gets to explain why something went wrong. Senior people do not. ‘When you’re the janitor,’ Jobs has repeatedly told incoming VPs, ‘reasons matter.’ He continues: ‘Somewhere between the janitor and the CEO, reasons stop mattering.’ That Rubicon, he has said, ’is crossed when you become a VP.’
MOST IMPORTANT STEP FOR BUILD PRODUCT IS BUILD PRODUCT. SECOND MOST IMPORTANT IS BUILD PERSONALITY FOR PRODUCT. NO HAVE PERSONALITY? PRODUCT BORING, NO ONE WANT.
The pros and cons ofvarious tech combinations.
Scarlett Johansson, of all people, can claim one of my favorite quotes about the dark side of the internet: “the internet is a vast wormhole of darkness that eats itself.
Everything is an act of imagination. Politics is about imagining futures and having the power to bring a collective group of people with you who give you the power to make that happen. It’s what Churchill did during the second world war.
A good sentence imposes a logic on the world’s weirdness. It gets its power from the tension between the ease of its phrasing and the shock of its thought slid cleanly into the mind. A sentence, as it proceeds, is a paring away of options. Each added word, because of the English language’s dependence on word order, reduces the writer’s alternatives and narrows the reader’s expectations. But even up to the last word the writer has choices and can throw in a curveball.
I’ve written six books now, but instead of making it easier, it has complicated matters to the point of absurdity. I have no idea what I’m doing. All the decisions I appear to have made—about plots and characters and where to start and when to stop—are not decisions at all. They are compromises. A book is whittled down from hope, and when I start to cut my fingers I push it away from me to see what others make of it. And I wait in terror for the judgements of those others—judgements that seem, whether positive or negative, unjust, because they are about something that I didn’t really do. They are about something that happened to me. It’s a little like crawling from a car crash to be greeted by a panel of strangers holding up score cards.
Related: Messi Walks