Notes

Here's an idea: Compare and contrast the forms of skepticism and pessimism demonstrated in these posts and take the opposite of each to see where the benefits of the tools might be found:

Robin Sloan in Claude revision report, May 2025 when trying to use Claude to edit a fictional short story:

None of Claude’s edits met my standards, which are very high. The model’s sug­ges­tions made the story duller rather than sharper; they clob­bered pre­cisely the bits of lan­guage (often weird-angled) that were the most fun; they pushed the story’s shape and tone pow­er­fully towards conventionality.

and

Dustin Curtis Thoughts on Thinking:

Thinking is compounding–the more you think, the better your thoughts become. But now, when my brain spontaneously forms a tiny sliver of a potentially interesting concept or idea, I can just shove a few sloppy words into a prompt and almost instantly get a fully reasoned, researched, and completed thought.

Okay, further expanding on the earlier idea: Let's connect this with some collected thoughts on the internet is bad vs. the internet is a tool and there’s better ways to use it.

Let's see who's headed in the direction of finding better ways, or at least conscious of and avoiding the bad ways:

Alan Jacobs in Kingsnorth’s Machine:

Some people, of course, know what to do: they opt out. And that’s why we don’t hear from them: we remain in the places that they’ve opted out of. But I continue to believe that it’s possible to use the internet healthily, without unplugging altogether.

and

the Standard Critique of Technology in From Tech Critique to Ways of Living:

We live in a technopoly, a society in which powerful technologies come to dominate the people they are supposed to serve, and reshape us in their image. […] The proper response to this situation is not to shun technology itself, for human beings are intrinsically and necessarily users of tools. Rather, it is to find and use technologies that, instead of manipulating us, serve sound human ends … Those healthier technologies might be referred to as holistic (Franklin) or convivial (Illich), because they fit within the human lifeworld and enhance our relations with one another.

and

Maggie Appleton Dark Forest & Generative AI:

This leaves us with some low-hanging fruit for humanness. We can tell richly detailed stories grounded in our specific contexts and cultures: place names, sensual descriptions, local knowledge, and, well the je ne sais quoi of being alive.

The final edge we have over language models is that we can prove we’re real humans by showing up IRL with our real human bodies.

In Markets for Lemons and the Great Logging Off, Lars Doucet proposed several knock-on effects from this offline-first future. We might see increased fetishisation of anti-screen culture, as well as real estate price increases in densely populated areas.

and

Mike Caulfield in AI Is Not Your Friend:

I am proposing that rather than act like an opinionated friend, AI would produce a map of the landscape of human knowledge and opinions for you to navigate, one you can use to get somewhere a bit better.

But perhaps the more profound question is why we’ve decided to consume the combined knowledge and wisdom of human civilization through a straw of “opinion” in the first place.

and

Sally Hall in Is AI ruining my brain?:

AI hasn’t ruined my brain yet, but I’m going to keep checking to make sure.

and

Finally, back to Robin Sloan again in 'Ask for what you want' is a key skill for the 21st century:

My alternative: get your wish, then stuff the genie back into the lamp.

Paths from here

• My Personal A.I. Policy: what I decided to do about it
• On Writing to Understand: letting the tools do the thinking
• Give a crap: the Who Cares Era, and caring anyway
• Bespoke Software: where the tools earn their keep
• Tools may change, but taste is hard to fake.: what the tools can't fake

1. I will use A.I. to help people who need help.
2. I will teach other people how to use A.I. in ways that help them while not hurting others.
3. If you can't talk about how A.I. can both help AND hurt, or have a more nuanced conversation than just "A.I. is BAD" or "A.I. is GREAT", I'm not here for the conversation. Like any powerful tool, there are a lot of ways it can hurt people, but how can it help people? Let's talk about it.
4. I will learn from the tool's capabilities, interrogate its output, and use the results as a guide or roadmap in my own processes. I will not blindly publish any machine learning model's output without care. If I can't be bothered to Give a crap, why should anybody else?
5. I will not use A.I. to automate away my own creativity, humanity, understanding, decisions, research, and process. These are things I love doing and experiencing myself. So often, the process or journey or experience along the way is the point. It's what makes a human life worth living.
6. Whatever I make with A.I.'s help (writing, design, code, etc.) has to carry my voice. A.I. can find sources and sketch options, but the choices that make a thing mine are mine: what it is, what it's called, how it looks and reads, what it says. I don't put my name on anything I haven't shaped into something I'd stand behind. Tools may change, but taste is hard to fake.

See also A.I. Good vs. Bad for more on the reasons behind this personal policy.

Paths from here

• On Writing to Understand: understanding as the point
• Agency: the decisions no one else can make for me
• Dive Deeper: the journey is the point

I enjoy metaphors. This is a record of (most but not all of) the sources and decisions behind how the garden looks, gathered together in one note. You may also find much of this same information on the About these notes page and in the Tool shed.

The printed folio

• Every page is a riso-printed sheet on a gridded table: two inks, Hunter Green and Sunflower, overprinting to olive, with Bright Red kept for stamps. The ink names come from the Riso ink list.
• Ornament lives in small places (a drop cap, stage sprigs, the arched window, the snail in the footer), and the CSS stays plain underneath.
• Rest of World's Digital Divinity series, and how they made it: the arched window, the table of contents with dotted leaders, the stacked page edges.
• Knowing Machines: plain, brutalist labels.
• Matthew Butterick's Typography for Lawyers: the margin rail and sidenotes.
• Melina K's Brothertiger Fall Tour 2022 poster: the window, the hairline frame, the red ledger.
• Michael Flarup's The Future is Colourful and Dimensional: depth that belongs on a screen, which he calls "Diamorph".
• Stamp ink fades as a note goes untended. It's decoration only: the date is always there in words.

Growth, not dates

• Notes are sorted by how grown they are: found, draft, in bloom, each with a two-ink sprig (seed, sprout, bloom). The stages approximately follow Maggie Appleton's growth stages. See Digital Gardens.
• Each note keeps a permanent N° in planting order, set in the number-plate face GL-Nummernschild.

The ground the notes grow in

• Questions plot: a note's Questions sit in a cross-section of soil, each question a seed, after Christopher Silas Neal's illustrations for Kate Messner's Up in the Garden and Down in the Dirt. A rabbit sleeps in its burrow at the edge.
• Plat map: Neighbors is a survey of lots, one per site, each with a survey call made from its owner's name, after Carlos Vanegas and colleagues' Procedural Generation of Parcels in Urban Modeling, found through jamessimo's Making my dream city builder.
• Seed packets: the tags page is a bed of seed packets pushed onto stakes, and each tag's own page is the back of its packet.
• Fingerpost: "Further in" is a fingerpost planted on a trail map, walking down the path as you page through the cards.

The tool shed

• The idea came from Victoria: a place for the code made for the site. See Tool shed.
• Made here hangs on a pegboard. Each piece of code hangs as the garden tool closest to what it does: the sync is a watering can, Slugify is secateurs, the plat map is a hand fork. A folded index card on the board, Why these tools?, gives each reason.
• The pegboard is packed like the cards on the index, so tools hang at staggered heights, with every hook on a hole.
• Borrowed tools are power tools still in the cases they were lent in, pastel cases with a luggage tag tied to the handle saying who lent them, standing on plank shelves on cut-wood brackets.

Dark mode

• Dark mode is its own palette, not the light one inverted. The inks overprint by light instead of by ink (screen rather than multiply).
• In the tool shed, the cases are the three inks at dusk, matched in lightness and outlined in a lifted tint of their own ink rather than cream. The tags are dark kraft, the twine and screws brass, the wood walnut, and shadows fall darker, not lighter.

Paths from here

• Digital Gardens: the idea the whole garden comes from
• Tool shed: the code behind it, and the tools borrowed
• Limitation as driver of creativity: two inks, and what that rules out

Robin Sloan in February of 2020:

This messaging app I built for, and with, my family, it won’t change unless we want it to change. There will be no sudden redesign, no flood of ads, no pivot to chase a userbase inscrutable to us. It might go away at some point, but that will be our decision. What is this feeling? Independence? Security? Sovereignty?
Is it simply … the feeling of being home?

then

again Robin Sloan in an update September of 2026:

The AI industry’s immediate future has more to do with exotic finance and gnarly politics than the underlying technology. I’m not invested in that future; most days, I think a minor bust would be a good thing. But, that underlying technology, in its development and flourishing this past year — yes, really that recently — is profound. It reframes computing entirely, replaces a stack of API reference docs with this simple invitation:
Ask for what you want.

then

My alternative: get your wish, then stuff the genie back into the lamp.

The launch of Anthropic's Claude Opus 5.5 model on 2026-10-22 has made the pipe dream of bespoke software even more possible. You can theorize, plan, design, build, and launch your very own custom app just for you in a single sitting.

Over the course of my first two weeks planning, designing, building, and deploying The Dashboard with this model, I've found this to be right on the money.

and

David P. Reichert via Jason Kottke on 2026-08-23:

As a case study, I’ll be reporting on a series of small machine learning experiments, executed with various degrees of autonomy by Anthropic’s Claude. If you’re sceptical of claims around AI, but also haven’t really engaged with recent developments in agentic coding… continue on! (If not, also continue on).

David's article is a long read but worth every second to get a sense of where things stand.

To begin to answer the questions posed by Sloan, for me the things I've built recently (The Dashboard, the Whisky page, this garden) more so than any software I've built before (and I've built a few things) absolutely provide that feeling of being home.

They all come with a semblance of independence (all can be redeployed anywhere I want whenever I want because they're built in a way that enables such transportability, avoiding platform lock-in), security (nobody can make my content disappear without warning or recourse), and sovereignty (I decide what does and does not change, and how often) that other platforms, and even other projects I've helped build in the past haven't.

However, not everyone sees it as coming home, instead offering tempering perspectives: Maggie Appleton in Building Lodestone:

Without them, you are no more capable than when you started. You may even be less skilled as you slowly forget how to do tasks without the help of AI assistance.

and

Mandy Brown in Toolmen:

if those in power cannot prove that a great many people are already inferior then they will bring that inferiority about by forcing them to use a tool that diminishes their intellectual and creative capacity

Where I've landed, for now: I use the new tools to build faster with the tools I already know and rely on, not to replace them. This garden is built on Eleventy, the Whisky page is HTML, CSS, and a bit of JavaScript, and I understand how both of them work.

Matt Webb, in A few technical words about Upsideclown, chooses his tech with one test:

will I be able to fix it, half drunk, ten years after I've lost all the tooling.

For these, the answer is yes. If the AI tools went away tomorrow, the garden would still build, and the Whisky page could go back to the plain HTML list I wrote by hand for years before I made it more interactive. The new tools make the work easier and faster, but nothing depends on them. That's a personal preference based on the past couple decades of enjoying using and working on the open web.

Paths from here

• Feral Upstarts: amateurs finding what’s next
• Building Things Here: making at a human scale
• Agency: control of your own tools
• A.I. Good vs. Bad: the case against, and for
• Tools may change, but taste is hard to fake.: new tools, same judgment

Via Robin Sloan’s April 2025 newsletter, Alan Jacobs on temporal bandwidth in No Time But The Present quotes Thomas Pynchon’s Gravity’s Rainbow,

“Temporal bandwidth” is the width of your present, your now. . . . The more you dwell in the past and in the future, the thicker your bandwidth, the more solid your persona. But the narrower your sense of Now, the more tenuous you are. It may get to where you’re having trouble remembering what you were doing five minutes ago.

It's also where this note gets its name. Jacobs on what happens when the Now is all there is:

You can’t stop playing the game, but its rules keep changing. There’s no time to think about anything other than the Now, and the not-Now increasingly takes on the character of an unwelcome and, in its otherness, even befouling imposition.

Alan goes on to say,

Increasing our temporal bandwidth helps us address the condition of frenetic standstill by slowing us down and at the same time giving us more freedom of movement. It is a balm for agitated souls.

I’m reminded of Colin Fleming on Melville’s Bartleby for The American Interest (via Venkatesh Rao)

"ghosting himself with his own life."

Rao's own read on where that leaves us:

The Arrow of Freedom™ now points in the other direction. Escape now lies inwards.

This line from Nettle by Ethel Cain:

Time passes slower in the flicker of the hospital light

particularly resonates as my father was in the hospital in June 2025 with a heart condition, when anxiety extended and expanded my sense of the now, blocking out the ability to see beyond the present moment and plan for the future.

Back to Jacobs:

Navigating life in the internet age is a lot like doing battlefield triage. There are days we can’t even put gas in our cars without being assaulted by advertisements blared at ear-rattling volume. And so we learn to be ruthless in deciding how to deploy our attention. We only have so much of it, and often the decision of whether or not to “pay” it must be made in an instant. To avoid madness we must learn to reject appeals for our time, and reject them without hesitation or pity.

His remedy is to "break bread with the dead":

I want to argue that one of the best actions we can take when faced with this paradoxical misery is to listen to the voices of those from long ago and far away—to “break bread with the dead.” And while I do not wish to suggest here that reading old books is a cure for depression, the expansion of our temporal bandwidth, to which reading old books can make a vital contribution, just might be a hedge against depressive inclinations and might provide a port, for however brief a time, in the storm.

That jibes with Clive Thompson’s Rewilding your attention.

Temporal bandwidth works at the scale of a whole career, too. Will Larson in A forty-year career:

There is, however, a simple formula for checking if what you’re doing is sustainable: how long does it take on a vacation or weekend until you stop feeling anxious, and how anxious do you start feeling when you think about returning to work on Monday? If work anxiety is a constant companion, then change your situation even if it feels like a step back in the short-term: your success depends on sustained impact, not spikes.

Work anxiety that follows you into the weekend is the narrow Now in another form: it crowds out everything that isn't Monday.

What's been widening my temporal bandwidth is Readwise Reader and structured notes. Reader is where long reads wait until I have attention to give them, instead of getting constantly triaged in the moment. Highlighting means the good parts don't disappear. Most of the quotes on this page came out of Reader (tagged now or re-wilding). And this garden is the structured notes part. Going back through old highlights and putting them next to each other is my own way of breaking bread with the dead, even if the dead in question is me from a few years ago.

Paths from here

• Agency: control of your own attention
• Ways to Live a Life: loss, and the everyday carrying on

TK is an old editing mark for to come: a placeholder for something that isn't written yet. We spell it TK, not TC, because the letters T and K almost never appear side by side in English, so a search for TK turns up the placeholders and nothing else.

Around here, a TK is an outward-facing reminder to myself. This part of the note is still growing, and there's something here that I want to come back to in the future, but I haven't quite put it into words yet.

Paths from here

• Digital Gardens: notes that keep growing

• Destin and the Smarter Scrubber
  • making new things is hard
  • making things in this country is expensive but not impossible

but

Marc Andreessen in It’s Time to Build:

Why aren’t we building Elon Musk’s “alien dreadnoughts”—giant, gleaming, state of the art factories producing every conceivable kind of product… all throughout our country?

Andreessen may have a point: building physical things here has gotten harder, and Destin’s scrubber shows how much work it takes. But who does the building matters. Many of the platforms funded and expanded over the last fifteen years got markedly worse for the people using them once investors demanded their returns.

Anna Wiener in Does Tech Need a New Narrative?:

If Andreessen’s exhortation to build was a call for “aggressive investment in new products, in new industries, in new factories, in new science, in big leaps forward,” it was also a call to power. … The sorts of projects Andreessen proposed don’t really make sense under the venture model; they’re unlikely to be as profitable as the products that have come out of Silicon Valley since the popularization of the commercial Internet. Building may need a different set of incentives, investment models, and values.

so

Maybe we can separate out the “wishful, ahistorical thinking” (Anna) from the elements of truth in what Marc said here, and tie in with smaller and more personal efforts instead of grand power-grabbing schemes.

Human scale is where most of us already build: a website, a tool, a fix around the house, things one person or a few people can make, understand, and repair. It doesn't need a gigafactory or a venture fund behind it, just someone who cares enough to make the thing and keep it working. People who've been making things on the web for a long time keep coming back to that.

Here's Matt Webb in A few technical words about Upsideclown: > It seems to me that, sometime in the last 17 years, the web forgot the simple pleasure of making, and appreciating what’s made, together.

and

Dan McKinley in Choose Boring Technology: > We should generally pick the smallest set of tech that covers our problem domain, and lets us get the job done.

These resonate with me as I've continued to build out my own little corner of the web. I make the conscious choice to use the smallest set of tech that solves the problems I encounter while building, instead of platforms built by others who have a vested interest in maintaining power and control.

Paths from here

• Agency: choosing my own tools over someone else's platform
• Bespoke Software: software at a human scale
• Feral Upstarts: who builds when it isn't the incumbents
• Limitation as driver of creativity: the smallest set of tech, as a constraint
• YouTube Favorites: more from Smarter Every Day

Start at Jony Ive in a talk with Stripe's CEO Patrick Collison:

I believe that when Somebody unwrapped that box and took out that cable, and they thought somebody gave a care about me. I think that’s a spiritual thing.

and

Contrast Dan Sinker's (negative) observations on lazy/sloppy AI usage in The Who Cares Era with Nathan Zed's video essay How Trying Became Cool Again where he explores how music artists have begun trying again in their presentation and performance.

then

See Dan Sinker's closing line in The Who Cares Era:

In the Who Cares Era, the most radical thing you can do is care. In a moment where machines churn out mediocrity, make something yourself. Make it imperfect. Make it rough. Just make it.

and

Mandy Brown in A Unified Theory of Fucks:

Good work is the art of giving a fuck about the living. And all of us, every day, are faced with good work that needs doing, and good work that we can do.

and

Venkatesh Rao in Pretending to Care, Pretending to Agree:

There are two basic types of masks: masks of pretending to care are exit masks, and masks of pretending to agree are voice masks. … it is only hard to pretend to agree when you care. If you don’t care to begin with, pretending to agree adds no additional strain.

and

Tobias Revell in Box 010: Being Nice to The Reader:

"Just like a designer, you, as a writer, have a job to make that effort as easy as possible for your reader so they can get to the actual substance of what you’re describing."

and

Adam Mastroianni in 28 Slightly Rude Notes on Writing:

Writing is a costly signal of caring about something. Good writing, in fact, might be a sign of pathological caring.

and

Matt Webb in Tech Has Graduated From the Star Trek Era to the Douglas Adams Age:

"(And btw being unserious, which is what I’m advocating here, doesn’t let us off the hook re: paying attention to what our tech does in the world.)"

but

Dave Ackerman in conversation with Lu Wilson in Dialogues on Natural Code:

Shipping code “to make money” is different than “to help society”, no
matter how tech hypocrites may try to conflate them.

Paths from here

• My Personal A.I. Policy: caring about what I publish
• Neologisms: making the reader’s effort easier
• A.I. Good vs. Bad: the duller, more conventional version
• Tools may change, but taste is hard to fake.: taste as a kind of care
• YouTube Favorites: experts doing things expertly

Stephen Wolfram, in considering his book Idea Makers, writes:

And I have to say I’ve found it a lot of fun to try to discover those intellectual missing links—that let one see just how someone went from here to there to manage to arrive at a particular idea.

The story itself is usually interesting—both historically and personally. But I’ve also found that knowing the story of how an idea came to be almost invariably lets me understand the idea itself more deeply.

and

Instead of heading straight for an abstract goal, what often happens is that the trajectories of people’s lives cause them to work on solving some practical problem. Needless to say, most people will just be satisfied if they solve the practical problem, and will go no further. But some will try to generalize it, and build up an abstract intellectual framework around what they have done. And that, in my observation, is where a great many important new ideas come from.

Paths from here

• On Writing to Understand: using your brain, not just your tools
• Dive Deeper: going deep to understand

"The world is full of revelations."

(Haruki Murakami, Hard-Boiled Wonderland and the End of the World)

Here's Jorge Hillebrand regarding his encyclopedic Star Trek knowledge in Proof that Patrick Stewart exists in the Star Trek universe, an interview by David Friedman on his blog Ironic Sans:

I think I know a lot about the visual aspects of Star Trek, but I also know where to look for additional information, so I haven’t memorized that all!

Knowing where to look (and how to look) is more important than just knowing things. Unless you're a Jeopardy contestant.

Put another way: You don't always have to know everything if you know where to look for the things you don't know.

Part of knowing where to look is knowing what you're looking for, sometimes long before you can name it. Here's Paul Graham in How to Do Great Work:

Unanswered questions can be uncomfortable things to carry around with you. But the more you're carrying, the greater the chance of noticing a solution — or perhaps even more excitingly, noticing that two unanswered questions are the same. Sometimes you carry a question for a long time. Great work often comes from returning to a question you first noticed years before — in your childhood, even — and couldn't stop thinking about.

and

Tiago Forte in Progressive Summarization:

what you read is good and useful and very important, you’re just reading it at the wrong time. ... The challenge is knowing which knowledge is worth acquiring. And then building a system to forward bits of it through time, to the future situation or problem or challenge where it is most applicable, and most needed.

That's a lot of what this garden is for me: a place to keep the questions I'm carrying and the things I read at the wrong time, so I know where to look when a new line suddenly connects to an old one.

Science fiction keeps coming back to this too. Here's Stephen Baxter’s Xeelee:

"Rees, the secret of a Scientist is not what he knows. It’s what he asks."

and

Isaac Asimov in Foundation's Edge:

What counts is that I have discovered the question to ask and a good question is, of course, the key by which infinite answers can be educed.

but

Lastly a word of warning by way of Ursula K. Le Guin in The Left Hand of Darkness:

“Well, we come here to the Fastnesses mostly to learn what questions not to ask.” “But you’re the Answerers!” “You don’t see yet, Genry, why we perfected and practice Foretelling?” “No—” “To exhibit the perfect uselessness of knowing the answer to the wrong question.”

Asking beats knowing, but only if it's the right question. A perfect answer to the wrong one just sends you looking in the wrong place.

Paths from here

• On Writing to Understand: knowledge, ability and understanding
• Dive Deeper: what you find when you keep going

Don't forget to use your brain. Your tools don't have to do your thinking for you, that way leads to an atrophied ability to truly understand. And understanding isn't just a stepping stone to getting things done.

It's also a muscle. Hand the work off and it withers. Keep at it and it compounds: the more you understand, the easier it gets to understand more. A lot of that work happens in writing, where putting a thought into words is how you find out whether it holds up.

Put another way: better writing starts with better thinking, and better thinking happens on the page.

Here's Hank Green in the We Are Here Newsletter on 2026-10-16:

I don’t know where we’re heading, but as a science communicator, I want to be clear that investigating the world does not just confer upon us new abilities. I think of our investigations into the world as giving us three great gifts:

1. Knowledge — just knowing /that/ things are.
2. Ability — using the knowledge to allow us to do new stuff.
3. Understanding — actually building a model of how things work.

That third one… I think some people forget it, or they believe that it is only a step we take in order to get to ability. And yes, I’m glad it does that, but also it is a great human drive. And some portion of us should love and celebrate it the way others love and celebrate great achievements in sports or music or literature. The fact that science often does give us new tools with which to affect our world can make it seem like that’s the only reason it exists, but true, deep human understanding that can be passed from person to person is, on its own, a little bit the point.

and

Robin Sloan in ‘Ask for what you want’ is a key skill for the 21st century:

“Your spaceship”, in this analogy, isn’t only your software, but your mind. Making these apps, I felt my programming muscles wither in realtime, the timelapse of the rotting apple. The idea of writing code by hand now feels … not impossible, but just SO excruciating. That feeling honestly frightens me, and I intend to keep it firewalled into this domain.

and

Paul Graham in How to Do Great Work:

If you do work that compounds, you'll get exponential growth. Most people who do this do it unconsciously, but it's worth stopping to think about. Learning, for example, is an instance of this phenomenon: the more you learn about something, the easier it is to learn more.

and

from the same essay:

Talking or writing about the things you're interested in is a good way to generate new ideas. When you try to put ideas into words, a missing idea creates a sort of vacuum that draws it out of you. Indeed, there's a kind of thinking that can only be done by writing.

and

Ted Chiang in Exhalation:

And words were not just the pieces of speaking; they were the pieces of thinking. When you wrote them down, you could grasp your thoughts like bricks in your hands and push them into different arrangements.

and

Adam Mastroianni in 28 Slightly Rude Notes on Writing:

lots of people think they need to get better at writing, but nobody thinks they need to get better at thinking, and this is why they don’t get better at writing.

Paths from here

• Ideas and their origins: where new ideas come from
• Knowing where to look: asking, not just knowing
• A.I. Good vs. Bad: tools and the thinking they replace
• My Personal A.I. Policy: not automating away understanding

What happens when we measure the wrong things, or measure the right things too hard?

Marilyn Strathern in ‘Improving ratings’: audit in the British University system, generalizing economist Charles Goodhart’s original law:

When a measure becomes a target, it ceases to be a good measure. [...]

Proof of performance and productivity requires outputs that can be measurable and thus made visible. This subverts the integral role that time with no visible output plays in both teaching and research.

and

Jack Clark in Import AI 411 on the Chatbot Arena (LMSYS) leaderboard:

The cautionary tale of LMSYS is an example of what happens when you over-optimize for making a number go up on a benchmark and therefore cause the benchmark itself to lose meaning.

Just like Google’s PageRank caused search rankings to lose meaning over time as SEO sought to game the system.

Will Larson in Privilege’s upward-facing window regarding measuring engineering team effectiveness:

(The most damaging variant is when folks recharacterize work they find difficult into low status work, recasting the hardest work as the least valuable.)

The exact slate of critical-but-unrewarded work depends on your particular environment, but I’ve seen teams that discredit the value of note-taking, event scheduling, user interviews, writing tests, planning – for any activity you believe is essential, there is a team out there that self-rates as extremely high performance and has decided that activity is total garbage.

The danger isn’t only that others measure the wrong thing. It’s that we pick measures that leave out our hardest work.

Matt Webb in A few technical words about Upsideclown:

What can we count so that we’re over the moon when there’s twenty of it?

Paths from here

• Privilege and Personal Blind Spots: more from the same Larson essay
• Ad Reads vs. Produced Commercials: what The Algorithm™ is optimizing for
• A.I. Good vs. Bad: the case against, and for

From my partner Victoria on 2025-05-26:

Ad reads feel intimate and familiar, commercial feels cheap and tacky and mass produced

See also: Clive Thompson on Rewilding your attention

and

Ali Montag in The Inner Ring of the Internet, quoting Vox’s Rebecca Jennings:

What we’re seeing is the lowest common denominator of what human beings want to look at, appealing to our most base impulses... The result is, well, mediocre.

Paths from here

• Agency: choosing whose voices get your attention
• The Now and the Not Now: ads at the gas pump, and triaging attention
• Goodhart’s Law: what The Algorithm™ is optimizing for

A lot of what shapes our days is controlled by someone else. Attention is the obvious one: feeds tuned to keep us scrolling, notifications timed to pull us back in. Clive Thompson calls the way out rewilding your attention. Then there's money, the tech we depend on, and politics. They're control systems in the engineering sense, feedback loops that measure what people do and push them toward someone else's goal, and we're constantly being measured.

We can't opt out of most of them. What we can do though is notice where we still have a hand on the wheel: what we read, what we make, which tools we use and how, and to some extent where our money and time go.

That's where agency lives for me. It isn't dependent upon having control over the systems. It's making my own choices within them, and making them on purpose.

I think of Ariel in Moonbound by Robin Sloan after visiting a Wizard to alter his substance:

Hughes told Ariel to lift his arms. Below each armpit, on the cool, smooth flesh of his flanks, the wizard had signed her work: 
So the number of authors grew. Ariel peered from one armpit to the other. He liked these new marks. They seemed to stand for taking control (left armpit) and making decisions (right armpit). They made him feel more adult.

and

Mike Rugnetta in an interview with The Verge

Sometimes the most helpful thing is to be reminded that you do have agency within your own life, though it may be remote, forgotten, unused, underdeveloped. 
I think of this advice as a reminder to look for the ways I have control over my life and use them in service of my goals. For me, that is what “taking life seriously” means; it is something no one else can do for me.

also

Of course, as mentioned at the top of this note, CJ Eller in Rewilding Your Attention:

rewilding attention is an active practice. One must not only pursue those tiny signals but share them as well, whether that means writing about them on your blog or by word of mouth.

and

Finally, Dene Grigar speaking in The Future of Text III:

I think it’s that kind of intellectual laziness that drives people to just allow themselves to be swept away by algorithms, and not intervene on their own and take ownership over what they’re consuming.

Paths from here

• The Now and the Not Now: rewilding your attention
• Building Things Here: building at a human scale, not on someone else's platform
• Bespoke Software: control of your own tools
• My Personal A.I. Policy: the decisions I keep for myself
• Ad Reads vs. Produced Commercials: whose recommendations you trust

Austin Kleon in Show Your Work:

Because they have little to lose, amateurs are willing to try anything and share the results. They take chances, experiment, and follow their whims. Sometimes, in the process of doing things in an unprofessional way, they make new discoveries.

and

Robin Sloan Surrendering to the Surface re: Jony Ive & Sam Altman:

(In all other respects, I agree with Jason Snell’s assessment: with sincerely all credit where due, this is not the team to figure out the successor to the smartphone. Now is the time for feral upstarts, not designers emeritus.)

and

Anna Wiener Does Tech Need a New Narrative?:

An earlier Silicon Valley mythology of post-counterculture nerds in electronics-cluttered garages grew sleeker, slimmer, and better capitalized.

The thread through all three: the people with the least to lose (amateurs, upstarts, outsiders) are often the ones who find what’s next, while incumbents polish what already exists.

Paths from here

• Building Things Here: human-scale building, not moonshots
• Limitation as driver of creativity: making more with less
• Bespoke Software: the home-cooked app

Constraints keep showing up as the thing that makes the work interesting, not the thing that stops it.

Rachel Kriebel, in an interview discussing the inspiration for her typeface Ultramega, on Soviet designers who created the Futura typeface:

They were trying to be efficient. Maybe they didn't have money, but they were still trying to be creative. I'm not gonna preach too much, but I love it when artists work within limitations and they still make something really interesting and cool.

and

Cap Watkins notes how forcing compliance with a style guide sparks inventive work in A Decision Tree for Designers:

"It’s surprising what people can do when forced to adhere to constraints, and I personally believe even just the first two questions generated not only better design but very interesting applications of styles we already had available."

then

L. M. Sacasas reflects on physical limits and the friction of the work in Feeding on Illusions, quoting a passage from Mass Driver in From human hands on the friction that changes the work:

ChatGPT will never invent a new calligraphy technique, because it has no wrists to feel uncomfortable. A tool that shields us from the friction of the work is compelling, but if we don’t experience the friction, we will never change the work.

and

Finally, Hank Green in A Beautifully Foolish Endeavor:

Put your problem, your assets, and your limitations in your head, and shake them together until something falls out.

Paths from here

• Feral Upstarts: amateurs who try anything
• Building Things Here: making things that last, on a budget
• Sinistral: a writing system for the left hand

Will Larson on privilege, and the confidence it took before he could see his own:

Noticing your privilege is a lot of work. It would be rewarding to say that I was always thoughtful on this topic, but it took me a long time to grow confident enough in myself that I could allow myself to acknowledge the inequality that I benefit from. For me, self-confidence was a prerequisite to self-awareness, and I imagine it’s similar for many others. ... If you don’t have some grasp on your own privilege, you’ll struggle greatly to support others as an effective manager, inspire others as a leader, or contribute as a partner in your own life.

and from the same essay, why no one's view of it is ever complete:

in the spirit of the Hegelian dialectic, most perspectives are flawed representations of another more complete perspective

C.S. Lewis in The Magician’s Nephew names both halves, where you stand and who you are:

For what you see and hear depends a good deal on where you are standing: it also depends on what sort of person you are.

but

Where you stand can also hand you some of the confidence Larson had to grow into. Here's Ian W. Sainsbury in The World Walker, on the people who were simply given it:

Seb had always slightly envied those who seemed so sure of themselves, of their place in the world. That sense of entitlement. Some people just seemed to have it, while others had to nurse a sense of self-worth so fragile, it sometimes seemed barely present.

bell hooks in The Will to Change describes a sureness that comes from simply being, not from a title or from what you do:

In patriarchal culture males are not allowed simply to be who they are and to glory in their unique identity. Their value is always determined by what they do. In an antipatriarchal culture males do not have to prove their value and worth. They know from birth that simply being gives them value, the right to be cherished and loved.

Without it, the gap shows in who you look down on. Shani Zhang in 21 Observations From People Watching:

I can see how much someone accepts themselves by looking for intense distortions in the way they are interacting with the world. Find the range in how they treat people; if there is a split difference in their stance towards people they admire, and people they look down on. I never met a person who looked down on others and unconditionally accepted themselves.

and how that hides itself, even from you:

Some people don't like themselves. They hide this from themselves by thinking they don’t like other people.

which is why Frank Herbert in Chapterhouse calls the noticing a constant job:

Humans are born with a susceptibility to that most persistent and debilitating disease of intellect: self-deception. The best of all possible worlds and the worst get their dramatic coloration from it. As nearly as we can determine, there is no natural immunity. Constant alertness is required.

Confidence borrowed from where you stand hides the very ground you're standing on. The kind of confidence hooks describes is what lets you look right at it and see it for what it is, which is a critical step in dropping the delusion that you don't benefit from privilege yourself.

Paths from here

• Goodhart’s Law: more from the same Larson essay

If I can’t show up as my true, authentic self, then I have to have something else to hide behind, usually some form of expertise or experience.

Paths from here

• Privilege and Personal Blind Spots: why it takes self-confidence to see your own blind spots

Michael Flarup in The Future is Colourful and Dimensional, via Daring Fireball, on generative AI and icon design:

Tools may change, but taste is hard to fake.

The type foundry Mass-Driver in From human hands., on why taste needs a human behind it (and citing the same left-handed writing system as Sinistral):

When we ask Midjourney to draw an ox, it does not think of home. ChatGPT will never invent a new calligraphy technique, because it has no wrists to feel uncomfortable. A tool that shields us from the friction of the work is compelling, but if we don’t experience the friction, we will never change the work.

and

on what gets lost when the tools set the taste:

If we allow these tools to dictate our visual culture, we must accept that it will be vitrified by machines which consider a few billion webpages equivalent to the sum total of human existence. I don’t accept that.

and

where the easy road goes:

AI might be a scenic path with no hills or hindrances, but it does not lead to the goal we’re chasing. It leads to a fucking desert.

The same goes for words. Here's Adam Mastroianni in 28 Slightly Rude Notes on Writing:

The beauty ain’t in the necklace. It’s in the neck. … Maybe that’s my problem with AI-generated prose: it’s all necklace, no neck.

and

what the neck is:

it’s not about filling in the biographical details, it’s about realizing that the thing you’re reading comes from somewhere. Good writing is thick with that coming-from-ness.

Same goes for code, even code an agent writes for you. See Robin Sloan's third rule in 'Ask for what you want' is a key skill for the 21st century:

Don’t let the AI agent name the app. You pick the name.

More on Robin Sloan's and Maggie Appleton's takes on what these tools can and do flatten are in A.I. Good vs. Bad.

Paths from here

• Give a crap: taste as a kind of care
• A.I. Good vs. Bad: what the tools can and can't fake
• Bespoke Software: new tools, same judgment
• My Personal A.I. Policy: rule 6, my voice in whatever I make
• Sinistral: a hand the training data never had

References

• Maggie Appleton, A Brief History & Ethos of the Digital Garden
• Maggie Appleton, Digital Gardeners on GitHub
• Tiago Forte, Forte Labs: Progressive Summarization
• Chris Aldrich, BoffoSocko: Wikis
• Tom Critchlow, Of Gardens and Wikis
• Maxime Vaillancourt, Digital Garden Jekyll Template
• Anne-Laure Le Cunff, Ness Labs: How to set up your own digital garden
• Mike Caulfield, Hapgood: The Garden and the Stream
• Tom Critchlow, Wiki
• xandra, 32-Bit Cafe: Resources List for the Personal Web
• xandra, 32-Bit Cafe: Creating Your Own Website
• Mark L. Irons, Patterns for Personal Web Sites
• swyx, Digital Garden Terms of Service

Paths from here

• Neologisms: naming the thing
• TK as reminder: how a growing note is marked here

From Jeremy Culhane's Last Meal (@ 39:41):

You're like, oh, um, yeah, I guess like I never realized that was something that was hurting me from 15 years ago. That every single time I do a show, that little impulse is felt in my body without my head knowing. And that's what triggers this whole thing.

Did you like find that sort of like primordial wound of like, oh, this is where the want to be respected, the want to, uh, be funny. The want to make other people happy to try and make myself happy comes from. There's so many, you know, it's, it's so interesting.

It's a, well. I always picture it not like, where you can dive, you can keep diving deeper and deeper and deeper and it's bottomless, but it's meant to be bottomless. And that's the beauty of it, is that you can keep on diving. It doesn't mean that there's pointlessness to diving. You have to keep on diving because that's the beauty of the journey of being an artist is how deep can you dive into that well, and what do you find there? And then what medium do you use to show the world your beautiful discoveries?

So, the journey of going deep is fun. So if you can land on something, like, I landed on an experience where, when I was in, uh, middle school, my mom, uh, I, it was a harvest festival, Halloween, and I really wanted to be this character Pink. It's a woman that I had made up, not Pink, the musician. It was just a woman I had made up that I thought was like, this is awesome. This is funny. It was like early drag basically was what I was doing. We went to Aardvarks, which was this old like costume place. She found me like a, a beautiful, funny gown. We put on a big wig and I went to my Catholic school and I thought it was gonna crush.

I thought like, this is so fun and funny and I'm, I became her. And I got, every single person looked at me different than I'd never been looked at in my life. From kids to parents. Parents were looking at me going like, don't talk to him right now. And that's something that we landed on of like, oh, that was so deeply affecting the psyche of who I, and how I present forever. That it was like, oh my God. Like, there's this part of me that loved dressing in drag. There's this part of me that loved embracing femininity wholly. Yeah. I'm not trans. I've definitely thought that whole experience and there's a beauty to that. I don't personally feel trans. But there is a part of me that is so deeply feminine that I think everybody has. Yeah. That Catholic society and all society really doesn't let us embrace, you know?

So. I think like investigating that gender identity and performance, especially as to alpha males and I'm dead serious, is really, really important and something that we don't grow up being taught. Absolutely. And so that was a core memory of going, there's this whole part of me, this whole fun and it was so fun. It was so joyous. It wasn't guilt-ridden. And then society had made it guilt-ridden and shamed that going through the process of diving deep in discovering that very specific moment has made this other whole feminine part of me explode and embrace and re-remember the joy that I felt through that femininity.

And I love it. And that is only the beginning of going like, oh, that was a core moment. What are all the things that led to that? And what are all the other moments connected with, me not ever feeling, feeling guilty when I experience joy. Which I think is a very modern thing, is if you feel joy, you feel guilt because other people aren't having it.

also

An old journal entry from 2005, a Go Deeper image of a scuba diver seen from below and thoughts surrounding it at that time, but in a new context now outside of the strictly religious interpretation that always felt too limiting.

and

10,000 Emerald Pools by Børns

Paths from here

• Ideas and their origins: finding where things come from
• Knowing where to look: the questions that lead somewhere
• My Personal A.I. Policy: the process is the point

Idea: You don't have to name a group of related things in order to talk about them together or what are some other ways of referring to a group of related things? (e.g. hyperlinks or tags) or do neologisms make our thinking more or less accessible to our readers?

Inspired by running across a few difficult-to-parse articles referencing some Weird™ neologisms which the author seemed to presume familiarity on the reader's behalf

Maggie Appleton on Neologisms, e.g. Digital Gardens Austin Kleon on Eno's "Scenius" in Show Your Work

Paths from here

• Give a crap: being nice to the reader
• Digital Gardens: a neologism worth keeping

Over the course of a single hour I built (née vibe coded) a Dashboard for the private security company I've worked for over the past year and a half. Then over the next week I shipped a laundry list of features and improvements. It has, quite literally overnight, become the single most indispensable tool in increasing the operational efficiency of the company.

The problem

Dispatchers are responsible for tracking every guard clock in and clock out across the entire company. Here's how they did that... TK Dispatch process

The results

TK Where we're at now

Watching experts do things expertly (these guys Give a crap)

• Four Keys Book Arts: From E-Book to Real Book
• Baumgartner Restoration: Why Bother?

Learning about science and interesting things:

• Smarter Every Day: The Mind-Blowing Machines that Stamp Millions of Metal Parts

Paths from here

• Give a crap: experts who care
• Building Things Here: Destin and the Smarter Scrubber

Phaedra Charles, a left-handed type designer, on how handwriting and calligraphy are taught for right hands, and the writing system she made for the left.

See: Observations on Sinistrality by Phaedra Charles and also: A Guide to Sinistral Hand

Paths from here

• Limitation as driver of creativity: a writing system built for the other hand

On loss, and how the everyday carries on anyway.

From Sputnik Sweetheart by Haruki Murakami:

"So that's how we live our lives. No matter how deep and fatal the loss, no matter how important the thing that's stolen from us-that's snatched right out of our hands-even if we are left completely changed, with only the outer layer of skin from before, we continue to play out our lives this way, in silence. We draw ever nearer to the end of our allotted span of time, bidding it farewell as it trails off behind. Repeating, often adroitly, the endless deeds of the everyday. Leaving behind a feeling of immeasurable emptiness."

And from Hard-Boiled Wonderland and the End of the World:

Loss was not a skill, not a measure of a life.

Paths from here

• The Now and the Not Now: time slowing in the flicker of the hospital light