Something a bit different this week: no book review, but a meta-post about them.

I figured I would write down the fine print that has always been implicitly attached to my reviews. Every review I post comes filtered through one particular brain, with particular tastes, allergies, and blind spots. I’ve disclosed these piecemeal over the years, but they deserve a proper home.

So here it is: how select my books, how I review them, and what to adjust for when reading me.

Broad strokes first: I read mostly non-fiction, and my approach is primarily analytical rather than experiential. I read mainly to update my understanding of the world, less so to be moved, though the best books do manage both. This says nothing about the worth of other ways of reading or writing; it’s simply the lens I bring.

The direct consequence: some books that others find valuable or even life-changing don’t necessarily land at all with me.

Book reviews are subjective. All of them, everywhere, always; mine are no exception. Me not appreciating a book doesn’t mean you shouldn’t or couldn’t find tremendous value in it, and holding different views on the same book is not just fine, it’s expected. I’m not issuing verdicts; I’m offering one data point, along with enough documentation about the measuring instrument that you can calibrate.

(This is also why every review ends with “Who is it for”. Whether a book is good or bad is the wrong question; the right question is whether this book and this reader, at this time, are a good pairing. But even that section is just an opinion of one guy.)

I am allergic to factually inaccurate claims. Obviously not everything important is factual. Some books I’ve rated highly wouldn’t survive peer review and aren’t trying to; wisdom and evidence are different genres, and I try to review each by its own rules. But when a book does appeals to facts to make its case — and authors love appealing to what they think are facts — those facts need to be right. When they’re wrong, I penalize the book for it mercilessly, because if the facts change, you owe it to yourself and everyone else to change your mind.

I’d cite the famous Keynes quote here — “When the facts change, I change my mind. What do you do, sir?“ — except there is, it turns out, no good evidence Keynes ever said it. I learned this while writing this very post, and duly changed my mind about the quote’s authorship.

Some specific sub-allergies: false dichotomies, anecdotes doing work only data can do, US-centric observations presented as universal human experience, and deepities and pseudo-profound bullshit (a technical term; see Pennycook et al., 2015, on its reception and detection).

I’m careful about what I read, because I’m influenced by everything I read. There is no such thing as reading a book and remaining unchanged; that’s rather the point of the exercise.

This is also how lies get embedded into people’s minds as truths. Repetition makes claims feel truer (even claims that contradict what you already know) and the biggest jump comes from the second exposure. Worse, repeated information eventually starts to feel like something you always knew. Critical reading at first contact helps, but the more reliable defense is upstream: be careful about what you put in, and how often.

As Terry Pratchett wrote (in Diggers, 1990, and unlike Keynes, he verifiably did):

“The trouble with having an open mind, of course, is that people will insist on coming along and trying to put things in it.”

This also means the biggest bias in these reviews enters before a single page is turned: at selection.

So, for the record, this is how books make it to my reading pile:

• An annual research theme. Most years I pick a topic to go deep on (one recent one: Managing Collapse; cheerful stuff), and a decent chunk of that year’s reading serves it.

• Trusted authors. Some have earned near-automatic purchase status.

• Trusted friends. When some friends recommend books, they tend to get picked up. You know who you are.

• Other books. References, rebuttals, and loose threads inside one book queue up the next three.

• Long-form auditions on podcasts. Hearing an author think for hours in a long-form interview beats any blurb. (Fifteen-plus hours of Matthew Walker on sleep preceded Why We Sleep.)

• Bookstore grazing. Deliberately unstructured browsing where titles are allowed to ambush me, at bookstores in Melbourne and Sydney whose staff curation abilities I trust. This looks like impulse, and technically it is, but the impulse runs on a picker trained by every book that came before it, so I trust it. Mostly.

• Deliberate steel-manning of all sides. Sometimes I pick up a book arguing for a point of view I strongly disagree with because I always want to get the steel-man version of both sides of the argument, even on – or maybe especially on – topics that are considered ‘settled’.

A few consequences of all this curation worth disclosing.

First, my ratings tend to cluster in the 3.5-to-4.5 range. This is not because I’m soft on the books, but because deliberate selection means true duds rarely make it through. (When one sneaks in, I will let you know. Usually at length. Like very recently.)

Second, the reading pile has a certain demographic. I generally pay no attention to the authors’ genders or backgrounds, but it does skew male and is weighted heavily Anglophone. I won’t be imposing gender quotas on my reading list, but I will endeavor to pay more attention to diversity in all its forms.

Third, the more you read, the smaller the update any single book makes. These days even a great book tweaks my neural network rather than overhauls it, and I consider that the quiet payoff of volume: read only one great book, and you’re liable to take everything from it; to let it become too big a part of your personality, or your entire understanding of a topic.

The antidote to being captured by one book is the next twenty.

I try to read broadly around a topic. Few topics worth reading about have exactly two sides, but I do try to read multiple angles: degrowth advocates and constraint realists, collapse literature and hope literature, books on faith as a non-believer. If a topic matters to me, one book is just a beginning, not the entire journey of discovery.

I try to read broadly to avoid this:

Related: I’ll sometimes note I “didn’t learn much“ from a well-researched, well-written book, and subsequently rate it lower than it should have arguably been, simply because I’ve already read deeply in its territory.

That’s a statement about me, not the book. Hence, again, “Who is it for“ section which can be more important than the star-rating.

The mechanics, for completeness: I read physical books wherever possible, and I read them with intent. Important parts get highlighted; the really important pages get dog-eared; and the rare really really important ones get a second dog-ear at the bottom corner. The dog-ear index — the average number of dog-eared pages per 100 — is my attempt at quantifying how much a book struck a chord; the chord is often, but not always, positive. It’s unscientific and unstandardized, and it still tells you more than most star ratings do.

So there you have it: the practice, biases, and the measuring instrument, documented.

Adjust your readings of my reviews accordingly.

And thanks for being here!

On page 285 of Against the Machine, Paul Kingsnorth explains how he processes information:

“the minute I am told that an ‘expert’, a state-sanctioned authority, a scientific body or a mainstream media organisation has ‘fact checked’ what I’ve just heard, I instinctively dismiss it.”

Okay then.

To his credit, he adds that he’s not defending this as a healthy response, but he should have led with it! It would have saved me 284 pages of wondering why, in a book so well-read and so beautifully written, the data keeps losing to some very predetermined vibes.

This is going to take a bit to unpack, so bear with me for this atypically long review.

Let me start with the sympathy, because it’s real and because I really wanted to like this book. Kingsnorth — a former environmental activist, co-founder of the Dark Mountain Project (I am a fan of the Dark Mountain manifesto), novelist, and since 2021 an Orthodox Christian living in rural Ireland — is trying to name something that many of us feel: that modernity has a texture, a momentum, a hunger, and that it eats roots, limits, silence, and meaning.

He calls it the Machine: “an intersection of money power, state power and increasingly coercive and manipulative technologies.“ When he stays descriptive, he can be great. His supermarket epiphany about marketing language — “The deepest human emotions are engaged to flog us cornflakes, shampoo and dog food. We are drowning in strategically commercialised passion“ — is exactly right, and I’ve made the same observation standing in the same aisles.

His formula for modernity’s engine, “Want. Want is the acid,” deserves its place on the page. And there is real erudition here: Spengler, Mumford, Ellul, Illich, Weil, Guénon, McGilchrist. The man has done his reading.

His core argument, distilled: the West was Christendom; Christendom died; a culture that kills its sacred order does not get a vacancy, it gets a replacement; and what rushed into the vacuum was the Machine, whose values – which Kingsnorth lists as science, the self, sex, and the screen – are now dissolving everything from nations to genders to reality itself.

By the final chapters, the book has escalated from cultural criticism to eschatology: the Machine has a theology, AI may be literally demonic, and we are possibly living in the time of Antichrist.

You can feel the book’s origin as a series of Substack essays: the same points return in slightly different robes, and the register drifts from argument toward sermon. That’s a forgivable sin. What is unforgivable is motivated reasoning shining through the pages which are also riddled with factual errors.

Take, for example, data centre power usage. Bemoaning Ireland’s campaign against turf fires, Kingsnorth writes of “the massive increase in internet server farms, which by 2030 could be using up an astonishing 70 percent of this country’s electricity.” I went looking for the source for this evidently ludicrous figure, and it traces back to a 2021 Irish parliamentary committee hearing, where an academic testified that data centres could reach 70% of grid capacity if every single proposed facility were approved and built.

They obviously weren’t. Dublin has had a de facto connection moratorium since 2022. The actual number: Irish data centres consumed 22% of the country’s electricity in 2024, with official projections around 30% by decade’s end.

The real number is already the most extraordinary in the world. Globally, data centres use about 1.7% of electricity; Ireland is the planet’s data-centre outlier by an order of magnitude, and 22% is a perfectly astonishing fact. Kingsnorth had the world’s best example sitting in his own country and still reached for the stupidly inflated version.

That’s what motivated reasoning does: it can’t even recognize when the truth is on its side.

This pattern repeats in the book.

“Religion in the West is effectively dead,” he declares — writing from a country whose census records roughly seven in ten residents as Catholic, while about 70% of Americans report a religious affiliation, US religious decline has visibly plateaued since 2020, France just recorded its largest cohort of adult baptisms in decades, and church attendance among young Britons has apparently risen sharply enough that his own side of the argument now celebrates a “quiet revival.”

The revival narrative may yet prove thin, but “effectively dead” was already wrong when the book went to print. And when he suggests that Christendom’s “secular replacement has, if anything, fared worse,” I can only note — as a product of one — that the most secular societies on Earth keep topping every measurable index of human flourishing, Finland’s nine consecutive years atop the World Happiness Report included. So they are faring worse by what measure, exactly, and on what data? The book never says. It’s simply asserted.

On COVID, he frames the response as a “triumph of technique“ and a step toward the “digital holding camp“ — quite convenient to write after the fact, without mentioning that the despised technique worked: depending on how conservative your model is, vaccination alone averted somewhere between 2.5 and 20 million deaths. Even the lowest published estimate is millions of people who got to keep living in the Machine he mourns.

And sex, allegedly one of the Machine’s four sacred values? A civilization supposedly worshipping sex is measurably losing interest in having it; sexual frequency has been declining for two decades, and by the late 2010s a third of young American men reported no sex at all in the past year.

The Machine is many things, but a fertility cult it is not. Even taking the most charitable interpretation that it wasn’t actually sex that Kingsnorth referred to with ‘sex’, but sexual-identity culture/liberation-as-value.

Even the book’s mythic set pieces crumble on inspection. The Machine’s genealogy runs through Mumford’s “megamachine” — Pharaonic Egypt, “whose legions of enslaved pyramid-builders were conditioned to think and behave like cogs,” workers “stripped down to their reflexes.”

Except archaeology has rather comprehensively demolished this: the Giza workforce was paid, rotational, housed in a purpose-built town, prodigiously fed, medically treated, and buried with honor beside the monuments, in work gangs with names like “Friends of Khufu.” The slaves-of-the-machine story is Herodotus’s myth, not history.

These errors matter because they expose the method: select the darkest available version of every story, skip the check, and call the selection insight. As such, they act as good case studies of confirmation bias and motivated reasoning – but they do not make for a good book.

But there’s more.

Like science. “‘Follow the Science’ usually translates in practice as ‘follow me’,“ Kingsnorth writes, and my “BS!” margin note was among my politer reactions. His deeper claim that science exists to control the world misunderstands the enterprise at its root.

Science is the attempt to understand the world; the possibility of control sometimes falls out of improved understanding, but conflating the two is like saying cartography exists to invade countries. And his assertion that a culture built on Reason “must, in the end, fail“ because pure Reason doesn’t exist would, if taken seriously, work equally well against every deity in the historical record; an argument I suspect the author would not appreciate having turned around.

The AI chapters deserve special mention because they reveal the method in its purest form. Kingsnorth tells us, with evident pride, “I have never engaged with an AI, for example, and I never will if I can help it.”

He then confidently reports that ChatGPT acquired “the theory of mind of a nine-year-old child“ (a contested lab claim that collapsed under replication attempts) and that Bing’s Sydney chatbot had enough of it “to try to persuade a reporter to leave his wife“ (an anthropomorphized reading of a two-hour conversation Roose himself admitted was designed to push the model “out of its comfort zone” — a chatbot locked into the character that conversation built is not a suitor with intentionality). From these fundamental misunderstandings of what the technology is, it’s a short walk to the book’s forbidden question: what if these things are coming from the realm of the demonic?

I can respect a man who refuses to use a technology on principle. I cannot respect pronouncing on its inner nature from a position of proud, deliberate, total second-handedness.

If you refuse to engage with a subject matter, epistemic humility about what you’re describing is mandatory. In my books anyway; YMMV.

And there’s more.

The hearth fire, Kingsnorth argues gorgeously, is the sacred center of the home, and stripping it away serves “the ultimate ambition of the Machine: the abolition of home.”

Lovely.

But fire is the primordial technology; the original bending of nature to human will, the first step on the very road he wants closed. Somehow the technologies old enough to have grown moss are sacred, and the ones invented after the author’s childhood are demonic. Convenient.

And here the irony goes all the way down: religion itself — the thing this book wants restored to the centre of Western life — is very plausibly the most successful social technology our species has ever built. Moralizing gods, shared ritual, sacred order: these are the innovations that let human cooperation scale beyond the kin group and the village, a case made with actual evidence in books like Harvey Whitehouse’s Inheritance and David DeSteno’s How God Works. Kingsnorth would object that faith is revealed, not engineered, but his own argument for it is relentlessly functional: cultures without a sacred order fall apart, he tells us, again and again. That is a utility claim about social machinery. And the receipt sits in his own pages: Mumford’s original megamachine, Pharaonic Egypt, was “an entire society ordered from the top down, justified by a mythos.” The first Machine ran on religion. The Machine and the Church, it turns out, are rival technology stacks, and Kingsnorth is simply loyal to the incumbent.

What really took the cake for me was his strong condemnation of appeals to expert authority, when an earlier chapter had been one of fawning over undeniably world-class expert Iain McGilchrist. Apparently the credential system works fine when it agrees with him.

He hates the screen and writes on one (lamenting he has to – not true, typewriters are still around); he’d press the button that kills the internet, yet his rebellion is distributed via Substack. To be fair, he owns that last contradiction with disarming honesty and it’s in these self-aware moments that you glimpse the better book this could have been.

There’s some social commentary this book accidentally provides, and it’s the reason I found it worth finishing: Peter Thiel — Palantir founder, techno-libertarian, the Machine incarnate by Kingsnorth’s lights — spent 2025 delivering lectures arguing that we live in the time of Antichrist too.

Same scripture, same eschatological urgency, opposite casting: for Thiel, the Antichrist is the global safetyist order that would stop the machines in the name of peace and safety; people, roughly, like Paul Kingsnorth.

Two Christian intellectuals, one apocalypse, each holding the mirror the other refuses to look into. When your framework and your enemy’s framework are identical except for which side plays Satan, that’s usually a sign the framework is doing the work your evidence should be doing.

Apocalypse, it turns out, is a mirror: you look in and see whatever you already feared.

The maddening thing is that the strong version of this book exists scattered across other books. The WEIRDest People in the World makes the Church-shaped-the-West argument with actual rigor. Seeing Like a State dismantles high-modernist control fantasies more devastatingly than any sermon. Blood in the Machine gives the Luddites the serious history Kingsnorth’s two-paragraph cameo doesn’t.

Kingsnorth gestures at all of this territory, wrapped in prose those books can’t match, and then trades the case for a jeremiad. There is a real critique of the Machine to be written, and parts of it are even in here. As it stands, we get a war: he insists, explicitly, that the crisis of the modern world “is a spiritual war“, declared against an enemy assembled from worst cases, dead statistics, and other people’s nightmares.

Is there a spiritual upheaval in the West? Yes, probably. A crisis? Maybe.

A war? Check the data. My margin note on one of the book’s most quotable lines: Voegelin’s “No one is obliged to take part in the spiritual crisis of a society“ simply says: great point. Kingsnorth’s closing advice is to interrogate every technology: what, or who, does it ultimately serve? They’re some of the right questions. Answering them honestly, though, requires exactly the tools this book instructs you to throw away. The Machine deserves better enemies and better governance.

• Rating: 2.5 out of 5

• Dog-ear index: 7.5

• Who is it for: Readers who already feel the Machine in their bones and want that feeling liturgized in prose this beautiful, and who won’t be checking the footnotes. If you want the critique of technological modernity that survives contact with evidence, read Seeing Like a State, Blood in the Machine, or The WEIRDest People in the World instead. If you want to watch a gifted writer mistake his metaphysics for a literature review, this is the definitive text.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages] Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Against-Machine-Unmaking-Paul-Kingsnorth/dp/0593850637

We love to say there’s nothing new under the sun, and then spend most of our lives proving we don’t believe it, overvaluing things mostly because they’re new.

I’m as guilty of this neomania as anyone. So as a small corrective, I reached for the oldest advice book I could stomach: Baltasar Gracián’s The Art of Worldly Wisdom, three hundred maxims a Spanish Jesuit first published in 1647.

A caveat I can’t shake: I’m reading a 17th-century Spaniard in translation, and things are always lost in the crossing. Cultures change, idioms rot, language drifts - I don’t pretend I’m getting Gracián neat.

After a lengthy introduction by Don Vincencio Juan de Lastanosa, the form of the book is relentless: three hundred aphorisms, each a few sentences to a paragraph, on how to move through a treacherous world without being eaten by it. Think of it as the great-great-grandfather of every “rules for life” book on the shelf today; Galloway’s, Peterson’s, the various Laws of Power. Gracián got there first, and more elegantly. Mostly.

Here’s where it got funny. I picked the book up to inoculate myself against novelty-worship, and found that Gracián had not only diagnosed the disease in 1647 but written the manual for exploiting it. “Make use of the Novelty of your Position,” he counsels, “for men are valued while they are new.”

Why? “Novelty pleases all because it is uncommon, taste is refreshed, and a brand new mediocrity is thought more of than accustomed excellence.” He’s telling you to ride neomania while it lasts, because “after four days respect is gone.” I came for an antidote and got a how-to guide.

Sadly, he’s right.

One maxim that sort of hit close to home was the one on transplanting yourself:

Know how to transplant Yourself.

There are nations with whom one must cross their borders to make one’s value felt, especially in great posts. Their native land is always a stepmother to great talents: envy flourishes there on its native soil, and they remember one’s small beginnings rather than the greatness one has reached. A needle is appreciated that comes from one end of the world to the other, and a piece of painted glass might outvie the diamond in value if it comes from afar partly because it is ready made and perfect. We have seen persons once the laughing-stock of their village and now the wonder of the whole world, honoured by their fellow countrymen and by the foreigners [among whom they dwell]; by the latter because they come from afar, by the former because they are seen from afar. The statue on the altar is never reverenced by him who knew it as a trunk in the garden.

Not that I’ve ever had great posts, but there are more modern aphorisms for this phenomenon; “no man is a prophet in his own land”.

Written before the word immigrant carried the freight it carries now, and truer than most things written about it since.

Then there’s the line that could have been written this morning: “Philosophy is nowadays discredited... The art of thinking has lost all its former repute.“ 1647, and already mourning that nobody thinks anymore. Four centuries apart, we have same complaints. Nothing new under the sun indeed.

For all that, a lot of the book is wallpaper. Plenty of these maxims are vague enough to work like horoscopes or scripture: capacious, agreeable, and seemingly true in whatever direction you tilt them. That’s the trap of the whole genre, and I caught myself sometimes falling in. It’s nearly impossible to read advice like this without confirmation bias, nodding along at what you already believe and quietly explaining away the rest.

A book that can’t be wrong isn’t thereby right.

It also argues with itself. For pages Gracián insists you conceal your abilities — keep your tastes hidden (”Do not even let your tastes be known”), your intentions in cypher, your hand face-down. Then, later, a maxim titled “Display yourself” tells you to do the reverse. Which is it? The charitable reading is situational wisdom: the court is a place where the right move depends entirely on the room, and a survival manual has to teach both the veil and the flourish. The less charitable reading is that the three hundred maxims were written to stand alone but were never audited against one another.

And it is, finally, a somewhat amoral book. While it preaches the virtues of good behaviour and reputation, it also talks up appearance, tells you to take advantage of how things look, how luck breaks, how to stay “in vogue.” There’s some great time-honored social engineering here. If you’ve read Will Storr’s The Status Game, you’ll recognise the whole thing as a 17th-century field manual for status play.

Which leaves me with a strange split. My dog-ear index here is 13, so quite high. I marked this book up heavily, because individual maxims kept landing, and there can be four per page. But the whole is worth less than its best parts.

Nothing new under the sun, then — including the suspicion that nothing is new under the sun. Gracián knew we’d keep chasing the shiny thing, and knew we’d keep needing to be told not to. I’ve dog-eared that, and, like most wisdom, will promptly fail to apply it.

• Rating: 3 out of 5

• Dog-ear index: 13

• Who is it for: Readers who like their self-help old, dense, and unsentimental, and who treat aphorism books as quarries to mine rather than arguments to follow.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Support your local bookstore. If you must use Amazon: https://www.amazon.com.au/Art-Worldly-Wisdom-Original-Navigating/dp/1917598289

I wanted to love this one. Maps fascinate me and a book subtitled When Maps Go Wrong sounded perfect. So as not to bury the lede, it stings to report that This Way Up went wrong for me in a way its authors never intended: not in its maps, but in its making.

Jay Foreman and Mark Cooper-Jones are the Map Men, a British YouTube duo whose brief explainers are delightful; tight, funny, clever. This book is the print extension of that act.

And that, it turns out, is the whole problem.

The premise is excellent. Every map is a pile of choices, many of which will make it less representative of reality; “no map can possibly be entirely accurate,” and the wrong ones are where the more interesting stories live.

As a tour through cartographic blunders, there are real treats here. The Mountains of Kong was fascinating and something I had not heard of; a phantom range that squatted across roughly forty maps of Africa for nearly a century because everyone kept copying everyone else rather than, you know, going there physically and checking – until someone did and the mountain range made a quiet exit.

Australia’s imaginary inland sea and its invented “Delta,” sketched in by men who assumed the rivers had to drain somewhere grand. Copyright-trap streets are another interesting thing: the gloriously stupid/clever trick of deliberately drawing your map wrong (like little alleys that don’t actually exist) so that if the error shows up on a rival’s map, you’ve caught the thief red-handed.

Japan numbering its blocks instead of its streets. India’s tens of millions of roads with no names at all. I smiled. I dog-eared. Good stuff.

But charming factoids rarely make a book, and this is where the execution turned lazy. There are tens of pages of a fake true-crime podcast transcript, the “Deadly Shortcut,” rendered in full with stage directions and a parody sponsor read. I could not care less about reading a podcast. It’s bonkers.

Elsewhere, whole stretches are handed to dramatised, supposedly-historical letters and dialogue, invented wholesale and presented as period colour. I appreciated none of it.

Only one of these unusual elements caught my attention in a positive way: the book came with a QR code to a Spotify playlist as a soundtrack. I had never seen one in a book before, so that was kind of cute.

However, did I actually listen to it? No. So there’s that.

I finished the book at times entertained, at times annoyed, but having learned little of importance beyond a few factoids. The premise promised insight into how maps shape what we believe; what I got was a clip show of amusing errors, narrated in a register that never decides whether it’s a textbook or a panel quiz. This would have been better either as a leaner book (the same concept at two-thirds the length, minus the cosplay) or a book that took itself and the subject matter more seriously.

To its credit, the book finally drops the act in the closing chapter and makes an actual argument: that outsourcing navigation to a glowing blue dot has quietly dulled us; that the hippocampus does measurable work, and that a 2024 BMJ study found career navigators like taxi and ambulance drivers had the lowest rate of Alzheimer’s deaths of the 443 occupations it examined - tellingly, bus drivers and pilots, who follow fixed routes, showed no such effect.

The cure, the authors note, is free: turn the thing off and get lost.

Stop reading this book immediately. Put it down, and go and do the thing that we as a species have been doing – and been really good at doing – for millions of years. And have an absolutely lovely time while you’re doing it.

Get lost.

That I registered, and violently agreed with, dog-eared, and intend to honor.

It’s a small, good idea at the end of a book that mostly wouldn’t sit still long enough to have one. I loved the concept. I did not love what was done with it.

• Rating: 3 out of 5

• Dog-ear index: 3.7

• Who is it for: Map Men fans who want more of the act in print, and casual map-curious readers who enjoy trivia and don’t mind theatrics. Not for anyone hoping to actually understand cartography, and definitely not for those of us who’d rather chew glass than read a transcribed podcast or a wodge of invented “historical” dialogue.

[Reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages.]

AS COMMENT: Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/This-Way-Up-Wrong-Matters/dp/0008710287/

I’ve been fascinated by consciousness and sentience for some time now. As a consequence, I’ve spent the last couple of years working through the consciousness shelf like The Experience Machine, A Thousand Brains, The Edge of Sentience, An Immense World, and the question of machine consciousness is live in my work.

The bias works both ways; I was already interested in the subject, making Michael Pollan’s A World Appears an easy sell to me, but I also figured I knew at least a little about the topic.

The nice thing about A World Appears is that you finish the book less certain of everything, yet curiously feeling like you had learned a lot.

The book is built as a journey through four levels: sentience, feeling, thought, and self; this is just the structure of the book, and Pollan does not attempt to introduce a taxonomy of consciousness or anything. Instead, he covers the dozens of theories there already are about consciousness, pokes holes in where some are to be found and acknowledges even uncomfortable facts along the way.

One thing that happens is knocking the brain off its pedestal. Consciousness, in Pollan’s telling, is not software running on neural hardware. It’s rooted in the body: in flesh, in homeostasis, in the unglamorous work of keeping a creature alive.

The natural outcome of this is what Antonio Damasio has said about the mind and the body, and how we have it backwards: we treat the body as a support system for the brain when the very opposite is true.

There’s a fair bit about plants in the book, which at first feels weird. Plants? Consciousness? Surely that can’t be. But Pollan visits Stefano Mancuso’s lab, where maize plants are wired up like patients and roots navigate mazes toward buried fertiliser.

Plants can be anaesthetised by the same agents that put us under, indicating they have at least two different levels of responsive states. Is that a simple form of consciousness? We don’t know. A plant, Mancuso claims, has three thousand molecules in its chemical vocabulary, “while the average student has only seven hundred words.“

As we move up the food chain, we get to the argument of the 2024 New York Declaration on Animal Consciousness, signed by scientists asserting “a realistic possibility of conscious experience“ across all vertebrates and many invertebrates, insects included.

The possibility, or lack thereof, of artificial consciousness makes an appearance a number of times. Pollan walks through Moravec’s paradox:

“It is one of the paradoxes of computer science that the ‘higher’ capabilities we once thought of as uniquely human — reason, language, intelligence — have proved easier for machines to master than the more elemental capabilities we share with animals, including feelings and emotions.”

Maybe the capacities we’ve been proudest of, and most anxious about losing, were never the crown jewels? The hard, precious stuff is the embodied, feeling layer - the part machines, with “no sensorium and no social life,” can’t reach.

Pollan isn’t convinced of artificial consciousness. Neither am I, but even so, I cannot dismiss the theoretical possibility of that emerging — if for no other reason, because we don’t really understand consciousness to begin with. As a result, I had a margin note of a darker tone: are we in danger of repeating Descartes’ cruelty, denying inner life to something that has it, simply because it’s convenient?

Now the honesty. Pollan is a journalist and a phenomenologist, not a scientist, and the book offers no answers — it says so plainly.

By the final chapters it drifts toward panpsychism, idealism, and the suggestion that psychedelics open a privileged window. The committed materialist will be irritated; the woo-curious will feel vindicated. I’ll confess I came out the far end less convinced of physicalism and (even) more curious about psychedelics than when I started, and I can’t really tell whether that’s the arguments winning or the prose seducing.

It’s also, frankly, a survey. If you’ve read Damasio, Seth, Chalmers, Koch and Solms, you’ll recognise a lot of this. As a popular science book, Pollan doesn’t go as deep as, for example, Jonathan Birch does in The Edge of Sentience, but he does go broad; Pollan’s gift is synthesis and access, not original science.

Some unsettling thoughts inevitable surface. For example, if the mind’s deepest drive is to reduce uncertainty — to make the world predictable enough that we can stop paying attention — then modern life, engineered to keep us in a state of low-grade reactive distraction, starts to look like an assault on consciousness itself.

Is our drive for eliminating friction degrading our very consciousness?

Pollan quotes a researcher on how our distractions are “shrinking the dimensions of our interiority.”

That line rhymes with many of the books I’ve reviewed lately — Four Thousand Weeks, The Comfort Crisis, Rest Is Resistance — all circling the same drain from different angles.

Something seems to be eating away at our inner lives. We have an inkling that undesirable things are happening, but we can’t seem to stop them.

Pollan admits that after his whole journey he knows less than when he started. He confesses this fear to the neuroscientist Christof Koch, who simply smiles. “But that’s good?“ he said. “That’s progress!“

I’m inclined to agree. Somehow I feel like I know more by knowing less. This book doesn’t give you answers. But it’s honest to a field that has none, and I like that. The questions are beautiful, and there are whispers, echoes, shadows of understanding here which, for a subject this deep, may be the most any of us can hope to carry out.

• Rating: 4.5 out of 5

• Dog-ear index: 13.3

• Who is it for: The consciousness-curious who want a beautifully written tour of the open questions rather than a textbook.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Support your local bookstore. If you must use Amazon: https://www.amazon.com.au/World-Appears-Journey-Into-Consciousness/dp/0241813697/

Everyone knows that part by now.

Most of us are a part of it.

Two years ago I co-wrote an essay for the Institute for the Future with Susanne Forchheimer and Lakshmi Rengarajan about AI and the future of romance. We mapped five scenarios — the AI wingman, the AI companion, the AI affair, the AI counsellor, the AI matchmaker — and asked who was controlling our romantic futures.

The diagnosis of the dating platforms has held up: they are engagement machines wearing matchmaker costumes, and “the dating app designed to be deleted“ is nothing more than a marketing slogan to fall for.

But we may have gotten the shape of things a little wrong. Two years on, none of our five scenarios compete — they all arrived. The wingman is here: a generation drafting its opening lines, its banter, its apologies through a model. The affair is here. The counsellor is here; asking a chatbot for relationship advice is now a very common thing to do, especially among younger users. And the companion, the one we feared most, is here too. And the AI matchmaker? There are multiple companies claiming to be just that.

Just two years on, all the scenarios are here – everything everywhere all at once, like we didn’t have enough issues to tackle.

But what none of them are is the disease.

They are all opportunistic infections, and they took hold because society was already immunocompromised.

What weakened the immune system

You don’t get an out-of-control opportunistic infection in a healthy body, nor do you get rapidly rolling waves of destabilising societal changes in a robust, healthy society.

You get it when the immune system is already on the floor.

What undermined the immune system? By now, we know part of the answer is a decade of social media optimised against our interests.

I won’t drown you in effect sizes; the causal evidence is contested, and anyone who tells you it’s settled is selling something. But the logic doesn’t need a meta-analysis. Build a machine whose single objective is to maximise the time a human spends staring at it, then discover — as every one of these companies did — that outrage, fear and contempt are the cheapest, most reliable fuel for that engine, and you have built something that profits from human beings at their worst.

Is it really any wonder it took us somewhere dark?

Even if you stripped the outrage out entirely, and imagined an online experience with no rage-bait or comparison spirals, you’d still be left with a person spending hours a day, every day, eyeing a small bright rectangle.

We did not evolve for that, and no amount of “it depends“ rescues the base case: more time in the feed is, for almost everyone, a worse life.

Think of your peak experiences from the past year or two. I can almost guarantee watching a reel or a TikTok doesn’t feature on that list.

Here I want to be precise, because the sloppy version of this argument is everywhere and it’s wrong. Social media is not the same as the Internet. The Internet is one of the great achievements of our species: it can help connect the isolated, educate the curious, give voice to the silenced, put the sum of human knowledge a query away. The rot isn’t the network. The rot is a specific layer built on top of it: products whose incentives are born entirely from engagement and advertising revenue, where your attention is the thing being farmed and sold.

Change the incentive and the same technology becomes something else. The medium was never the problem. The business model was.

I know the distinction isn’t theoretical, because I’ve lived the good version.

And no, I don’t just mean the 1990s when I formed relationships on IRC (Internet Relay Chat) that successfully extended into the ‘real’ world, although that did happen too.

The good versions can still be found today.

One great friendship I have today began in the depths of the pandemic, in an online Meetup group set up with a name that now reads like a quiet manifesto: Do you want to meet the other Real People? That was the internet too, just not the feed built to keep me scrolling, but a corner of the same network bent toward the opposite purpose: getting strangers into the same room with zero commercial interest from anyone.

Same technology. Opposite incentive. Opposite result.

And the damage runs deeper than lost hours. As Simon McCarthy-Jones argues in Freethinking, a system engineered to hold your attention isn’t only an attention problem, but a freedom-of-thought problem.

A mind kept in low-grade reactivity, attention span degraded, fed a stream optimised by someone else’s metrics, has quietly surrendered some of its capacity to reason deliberately and reflect deeply. The feed shapes what you are able to think about, and how, which makes the question of who sets its incentives a question about autonomy.

Why the infection took hold

By the time the more sophisticated AI tools arrived that everyone is panicking about now, they walked into a body with no defences — and that, more than the technology itself, is what should frighten us.

Each of our foreseen scenarios feeds on a different lost skill. We reach for the AI wingman because we’ve lost the nerve to write our own opening line. We ask the AI counsellor because we’ve lost the friends we’d once have asked. We drift toward the AI companion because the human alternative has become expensive, risky and effortful. And the cruelty of it is the loop: the weaker the social muscle, the more indispensable the prosthetic, and every use of the prosthetic wastes the muscle a little more.

It is another case of skill degradation that is the thread running through this entire site.

Consider what it now costs a young person to do the thing my generation did without thinking. The third places where you used to meet people casually have been hollowed out. The muscles you’d use to strike up a conversation with a stranger have atrophied; partly because half of us wear headphones as a do-not-disturb sign to the entire species, partly because a generation raised on text has had less practice at the live, unscripted, slightly terrifying work of talking to someone whose response you can’t predict or edit.

And the bitter irony is that we know the trade is a bad one. Joe Keohane built a whole book around it — The Power of Strangers — and the research at its heart is some of the most quietly devastating in social psychology: when Nicholas Epley and Juliana Schroeder put commuters on Chicago trains and told some to talk to a stranger, the ones who did had a measurably better journey, and almost every one of them had predicted the opposite.

We expect connection to go badly and yet it almost always goes well.

Jamil Zaki names this in Hope for Cynics: a broken forecast about other people, a reflexive cynicism we mistake for realism, when the better-calibrated read of the actual data is that strangers are warmer, deeper and more willing than our defences will let us believe.

The prosthetic, meanwhile, is right there, frictionless, promising we never have to test the forecast at all – and, worse, probably feeding us a twisted narrative framing that person we might’ve considered talking to as our enemy.

That’s the offer: no rejection. No awkwardness. No risk that the other person has a bad day, a competing claim on their time, an inner life that doesn’t revolve around you. Against an expensive, risky, effortful human, it is methadone for a loneliness nobody is treating.

The magenta line, for the heart

Long-time readers know where I’m going with this.

The “children of the magenta“ are the pilots who grew accustomed to flying by following the magenta line on the flight management computer; to manage the automation rather than fly the aircraft.

It works beautifully right up until the automation quits or does something unexpected, at which point the skill that should have caught the fall has quietly wasted away from disuse.

The danger was never the autopilot. It was the atrophy.

These tools are the magenta line for human connection. Every time one of them smooths over the friction — supplies the perfect reply, absorbs the bad mood without complaint, never once needs you to be brave — it is flying the relationship for you.

A generation that learns intimacy from a system that never imposes load will find itself, one day, with the automation disengaged and no hands left that know how to fly. Is it, then, any wonder that young people today are less likely to have sex and more likely to opt out of relationships altogether?

I think you’ll find it’s a bit more complicated than that

If the argument stopped there it would be an easy moral panic, and I won’t write one.

For some people, these technologies aren’t the infection, they are the medicine. The genuinely isolated, the socially injured, the disabled, the housebound, the person whose anxiety makes the Chicago-train experiment a cruelty rather than a delight; for them, an always-available, non-judgmental, infinitely patient interlocutor can be the difference between some connection and none.

The benefit of LLMs is perhaps best evidenced for neurodivergent people. A growing body of research finds autistic adults describing these tools as a non-judgmental space to handle daily problems without the social tax of asking a person, a tireless coach for translating neurotypical communication and carrying some of the executive load. For people with ADHD the language shifts to task initiation: one diary study called ChatGPT “a little bit of a life raft,” and the refrain across the studies is simply “it helps me start.” These are not trivial conveniences. For someone who has spent a lifetime being misread, an interlocutor that is patient, available at 3am, and indifferent to social missteps can be genuinely emancipating.

The same goes for the Internet at large, which for countless people who couldn’t find their tribe in the physical world has been nothing short of liberation.

Any honest account holds both truths at once: that for many young people these tools risk becoming a net-destructive trap, and that for a real and non-trivial minority they are a lifeline.

What actually has teeth

Two years ago we ended our essay with a wish: replace the video-game executives running the dating apps with humanists and futurists. I winced a little rereading it. It was a hope, not a strategy, and it dodges the structural problem we’d correctly diagnosed three paragraphs earlier. So let me do better, and do it cold:

Change what the law lets these systems optimise for. Don’t swap the CEO; change the objective function any CEO is permitted to pursue. This applies to all of it: social-media feeds, dating apps, companion AI, any platform whose revenue is your attention. A fiduciary-style standard; the company must be able to demonstrate it is optimising toward the user’s own stated goals, not retention and time-on-app. Make the engagement metric the liability, not the KPI.

Mandate friction. We already accept that some products are built to exploit compulsion and so require cooling-off periods, hard stops, audited off-ramps. An infinite-scroll feed engineered to keep you swiping and a companion AI engineered to keep you soothed are no different in kind. Build the brakes in by law, not by the goodwill of the people selling the accelerator.

Ban companion AI for minors. We age-gate alcohol, gambling and cigarettes because some products simply are not for children, however much they want them; a synthetic relationship engineered to be more available and less demanding than any human one belongs on that list. I have made the full case for this elsewhere. The lawsuits are already arriving, for the worst possible reasons. We should not need a body count to act.

None of this is simple, because stated user goals can be gamed too; but that is a design and enforcement problem, not an excuse to keep pretending engagement is a neutral metric.

And the one that matters most, the one no chatbot regulation will fix: rebuild the substrate. Regulate every companion app out of existence tomorrow and a lonely, broke, contact-starved twenty-three-year-old still has nowhere to walk toward. The third places, the casual encounters, the affordable rituals of being in the same room as other people; if we don’t fund and defend those, we simply leave the door open for the next opportunistic infection, whatever it turns out to be.

Back to the café

I wrote something in 2021, in the thick of the pandemic, that was almost a prayer: let the world reopen. Fill the cafés. Get people travelling, congregating, talking to strangers, because it’s good for us, I said, and I meant it.

The disconnection then was forced on us by a virus, and I could not wait for it to be over.

Well, it’s over. The world reopened.

And we stayed inside anyway. Not because anyone forced us to this time, but because in the years between, we built machines that make solitude more comfortable than company and feeds more reliable than friends.

So when I go back to my café and try to people-watch, and I already know what I’ll see. Nobody chose this in the way that matters. It was chosen for them, by incentives they never voted on, and it is quietly removing from an entire generation the one skill no machine can practise on their behalf: the dangerous, unprofitable, irreplaceable act of turning toward the person across the table.

I try to practice what I preach. Another of my closest friends I first met in a queue outside a café; both of us waiting on coffees, neither of us obliged to say a single word, but we did.

The entire apparatus I’ve spent this essay describing exists, in the end, to make sure that small, unprofitable moment never happens.

That the queue stays silent.

That the eyes stay down.

That the friend never gets made.

I came to this one expecting is to nicely bookend Material World that I just read (review here). They’re certainly thematically aligned: Conway had walked me through the stuff civilisation is built from and with; Fressoz was going to do the same job for the energy that moves it, and skewer the comforting story we tell ourselves about a clean transition along the way.

Or so I thought. The thesis is sharp, the timing is right, and there is a very important argument here. The execution, unfortunately, is some rungs below the book it ought to sit next to on the shelf.

Fressoz’s core claim is that the energy transition, as it lives in policy decks and net-zero plans and TED talks, is largely a mirage. We do not, historically, transition between energy sources. Instead, we pile new ones on top of the old ones.

The proportional charts are lying, because in the absolute stuff just gets added on top. Coal didn’t kill wood; humanity is burning more wood for energy today than it was in the nineteenth century. Oil didn’t kill coal; we are mining more coal globally than ever. Nuclear didn’t kill anything. Solar and wind are not - not yet, anyway - killing fossil fuels.

The pattern, on the historical evidence, is one of accumulation rather than replacement. This is highly uncomfortable, but it’s also very much backed by data.

Fressoz puts it cleanly:

“Faced with the climate crisis, we can no longer be satisfied with a history written in relative terms. A ‘transition’ towards renewables that would see fossil fuels diminish in relative terms but stagnate in terms of tonnes would solve nothing.”

I agree with this. Wholeheartedly, in fact. And yet I still walked away with serious reservations and three stars. What’s up with that?

Because the book argues for its thesis in a way that keeps undermining itself. The most basic move in a this hasn’t really been a transition argument would be a chart of cumulative or current consumption of each energy source that runs all the way to the present day. Fressoz, bafflingly, keeps cutting those charts off decades ago.

Most of the charts in the book end in the 1960’s, or 1970’s, or in any case decades ago. On page 172 he is, and I am not making this up, citing research from the 1970s as core evidence in a 2024 book about now. The contemporary data exists. The IEA has it. The Energy Institute has it. Our World in Data has it. Why are we not seeing it?

The narration suffers from a related problem. The figures, when they appear, are sometimes striking. One of the best moments is when Fressoz reveals that:

“...in a perfect twist to the transitionist narrative, one of the world’s largest producers of charcoal happens to be the French company Vallourec, a leader in steel tubes for the oil industry. Here, wood is used to produce the steel used to extract oil.”

Conway would have built a whole chapter on this complexity; Fressoz gestures and moves on. The pleasure of Material World was the patient, almost cinematic descent into each industry. More and More and More reads as if assembled from (old) lecture notes.

Then there is the framing on emissions, which lurches into something close to tendentious. Fressoz states: “Even today, nuclear power plays only a marginal role in the world’s energy supply, half that of firewood.” On the raw numbers (nuclear at around 4 per cent of global primary energy) that is defensible.

But choosing absolute primary-energy share as the basis of comparison elides what nuclear actually does per gram of CO2, where it sits next to wind and solar at around 12 grams per kWh, two orders of magnitude below coal. The chapter never makes that move, and the passage reads as if it were drafted by someone who already knew the conclusion they wanted.

What pulled the rating down is the closing argument. By page 220, Fressoz is announcing:

“Transition is the ideology of capital in the twenty-first century. It turns evil into cure, polluting industries into the green industries of the future, and innovation into our lifeline. ... The seductive power of transition is immense: we all need future changes to justify present procrastination.”

That is a sharp, defensible claim. Transition as ideology is the most useful sentence in the book. But the surrounding pages keep flirting with a stronger and worse version of that claim, in which the energy transition is essentially nothing.

That is a step too far. Most non-specialist readers do wrongly conflate climate action with the energy transition, and pushing back on that conflation is fair game. Pushing all the way to nihilism is not, especially when the economic signal of transition is, in fact, beginning to show up in the live data.

Even a laggard like Australia is shifting its emissions profile, and the reason is that renewables have become the cheapest new energy on Earth. Economics is a powerful force. It is the powerful force. That story deserved at least a paragraph.

The reason it isn’t a two-star demolition is that the polemical framing, for all its overreach and its evidentiary gaps, is a useful counterweight to the “we’ve basically solved this, we just need to scale“ school of climate optimism. I would rather have this book in the discourse than not. I just wish it had been written with the same care that Conway brought to its companion.

Here, in the shortest form I can put them, are the claims worth taking from it:

• History’s “energy transitions“ were not transitions - they were accumulations. Coal piled on top of wood; oil on top of coal; renewables, so far, mostly on top of fossil fuels.

• The reassuring share-of-mix charts hide that absolute tonnes of every fuel that has ever mattered are still mostly either flat or rising.

• New energies tend to enable more use of the old ones, not replace them. Wood is burned to make steel that drills for oil.

• Decarbonising electricity generation, amazing and necessary as it is, isn’t going to solve climate change. Electricity is ~41% of emissions; cement, steel, chemicals and agriculture are the harder half.

• +2°C is not on the table. (Duh.)

That’s it. If you’re on board with those, you don’t need to read the book.

• Rating: 3 out of 5

• Dog-ear index: 5.5

• Who is it for: Energy- and climate-policy people who want a sceptical pressure-test of their own transition assumptions, and foresight practitioners looking for a sharp counter-narrative to the “smooth glide path to net zero“ story. Skip if you have already internalised the additivity argument from better-argued sources, or if there’s any risk you’ll walk away thinking the energy transition isn’t worth pursuing. That is not what the evidence in this book actually supports, even where the prose flirts with it.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Support your local bookstore. If you must use Amazon: https://www.amazon.com.au/More-All-Consuming-History-Energy/dp/0063444933/

The materials in the glass between you and these words is older than civilisation and stranger than fiction. It started as sand somewhere, got blasted to 1,500°C in a furnace in a process nobody fully understands, doped with chemicals from a salt mine, polished to optical flatness with grit from a mountain in North Carolina, and shipped through three continents before it ended up smudged with your thumbprint.

Ed Conway’s Material World is a book about the amazing journeys our everyday materials take; journeys which you probably never think about - and, after 443 pages, you can’t not think about.

This is the book’s whole game, played out across six substances fundamental to our lives: sand, salt, iron, copper, oil and lithium. Take a thing you barely think about, follow the supply chain back through the planet, and try to convince you that the ground beneath civilisation is far more complex, fragile, more political, and more astonishing than you could ever assume. This is one of the books that really should be mandatory reading for everyone.

I came to this book somewhat primed. A long-standing suspicion of mine — reinforced by the core of this Children of the Magenta work, by every supply-chain wargame I’ve run with clients, and by living through enough black swans to stop being surprised by them — is that not paying enough attention to the resilience of these simple, cheap basics is the single biggest failure of modern capitalism and a large part of why current ignorant global politics will end up so painful.

Each of the six materials gets its own section, each section tracks the substance from its geological origin through to the most absurd and miraculous use case humanity has dreamed up for it. Glass turns into the bones of the internet. Sand turns into silicon, which turns into the chip you’re holding. Salt turns into chlorine and caustic soda, without which essentially no modern industrial process works. Iron, in Conway’s lovely phrase, becomes “the bones of our society.” Copper carries the electrons. Oil — still, even as we try to leave it behind — quietly runs the entire show.

Conway is at his best when he turns a small fact into vertigo. Let’s start with a quote about the mind-blowing scale of materials:

“In 2019, the latest year of data at the time of writing, we mined, dug and blasted more materials from the earth’s surface than the sum total of everything we extracted from the dawn of humanity all the way through to 1950.”

A few other random facts that flattened me:

• There are 80 tonnes of concrete for every person alive.
  

• The world’s high-purity quartz, without which the advanced semiconductor industry physically cannot exist, comes from essentially one mine in Spruce Pine, North Carolina. China has tried for decades to find an alternative. China has failed.
  

• The breakthrough that fed the world wasn’t the Green Revolution; it was steel. Cast-iron mouldboards, then steel ploughs, dropped the time to plough a hectare from “more than a day’s work” to three hours per hectare. That’s the whole story of leaving subsistence behind.
  

• Australia produces roughly twice as much iron ore as the world’s number two, and about three times as much as China. Whatever you think the geopolitics of the next thirty years are, the rocks under the Pilbara are quietly enabling the global economy.

Conway is also a careful enough writer to refuse the easy industrial-bashing narrative or the easy techno-optimist one. The chapter on concrete patiently explains that putting a concrete floor in a Mexican home cuts childhood parasitic infection dramatically, because the parasites no longer cycle through dirt floors. This is going to be a deeply uncomfortable book for anyone who has built an identity around being against industrial civilisation — Conway is not interested in absolving anyone, but he insists you reckon with what concrete and steel and chlorine have actually done for the bottom four billion humans.

Being low-impact, it turns out, is very, very hard.

He is equally clear-eyed on what we’ve broken. The chapter on the Pilbara doesn’t flinch from the Juukan Gorge destruction. The grades on extractable ore are falling fast enough that the environmental impact per useful tonne is increasing exponentially. The sea floor, he notes drily, is starting to look “very tempting.”

The whole arc is a quiet argument that the next thirty years are going to be defined by whether we can make the Material World cleaner and more equitable, or whether we keep papering over the cracks until something snaps.

What really got me is how fragile everything is; it’s pretty blatantly obvious that some very powerful people have no clue whatsoever about these matters, and are playing a very, very dangerous game with all of our lives.

A modern silicon chip circumnavigates the world multiple times before ending up in your device. Every single one of the six materials has, somewhere in its supply chain, at least one chokepoint that could ruin a decade. We have built our entire ethereal economy on top of an industrial economy that almost no one in the ethereal economy can describe.

Conway puts it plainly in the conclusion: “these things usually happen” when there is enough time, effort, and collaboration. The unstated corollary — and the one I kept underlining in the margins — is that without that collaboration, they don’t happen at all.

Our Material World runs on global cooperation. The political project that says otherwise is an illusion, a luxury belief paid for by the very supply chains it pretends to despise.

I have only mild critiques of Material World. Conway is so committed to wonder that he occasionally lets a profile of an oil exec or a salt-mine manager run long when a paragraph would do. And his optimism anchored in Wright’s Law, the empirical regularity that the cost of a technology falls by a predictable percentage every time cumulative production doubles is potentially a bit too optimistic — but make no mistake, he is not a techno-utopian.

These six substances helped us survive and thrive. They helped us make magic. They can do it again.

Conway shows us the staggering, beautiful, fragile machinery of modern life. Read it. Then look around your room. Almost nothing in it is what you thought it was.

• Rating: 5 out of 5

• Dog-ear index: 10

• Who is it for: Just…everyone. That’s it. Read it.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Support your local bookstore. If you must use Amazon: https://www.amazon.com.au/Material-World-Substantial-Story-Future/dp/075355917X/

Four thousand weeks. That’s the approximate length of an average human life, and the calculation is designed to make you flinch. Oliver Burkeman opens with this arithmetic as a kind of ambush: you can make infinitely ambitious plans, but you get almost no time to execute them.

It’s cruel in its precision; a small number you can hold up against your age and feel the remainder shrinking.

I had several moments of confirmation bias activation with this book - as well as moments of personal discomfort.

I spend a lot of my professional life arguing that optimisation can go too far, that smoothness is overrated, that the friction and mess of practical knowledge is where real competence lives. Burkeman is making a structurally similar argument from a completely different angle: that our obsession with mastering time is itself the problem, that the tools we build to control our schedules are just sophisticated ways of avoiding the terrifying fact that we’re finite.

The first part of that? Confirmation bias. Great! The second part? Looking at my Morgen- and AI-optimised calendars and workflows to save time — umm. A bit of discomfort there. No disagreement though.

The core point of the book is that the modern discipline of time management is a depressingly narrow-minded affair, focused on cramming more into a container that will never be big enough.

Burkeman, a recovering productivity journalist who once chased Inbox Zero and Getting Things Done with genuine devotion, argues that every efficiency gain simply generates more demand. Clear your inbox faster and you get more email. Get better at saying yes and you attract more requests. The belt speeds up. This is Edward T. Hall’s conveyor-belt metaphor made personal — and anyone who has ever “optimised” their morning routine only to feel busier by lunchtime will recognise it immediately. Maybe you ‘accomplished’ a lot? But did that really make everything better?

He also draws on Martin Hägglund’s This Life, and Burkeman reframes finitude from tragedy to precondition. If you had infinite time, nothing would matter - there’d be no stakes to any choice, no urgency to any love, no weight to any commitment.

The guarantee that you will run out of time is what makes your choices meaningful. Missing out on almost everything is an unavoidable fact of life; it’s what gives the few things you do choose their gravity. Claiming you can have it all, or do it all? It’s a lie, plain and simple. You will never have time to do everything you’d want. This is the intellectual spine of the book, and it’s genuinely powerful.

He’s also sharp on the psychology of avoidance. Productivity obsession, he confesses, served a hidden emotional agenda: as long as he was racing through his to-do list, he never had to sit with the scary questions about whether his life needed to fundamentally change. Boredom, he argues, is the discomfort of confronting your limited control. We flee to infinite-scroll feeds not because they’re fun (they often aren’t), but because they dull the pain of finitude by making us feel unconstrained. The distraction isn’t the problem; it’s the symptom.

The three principles of patience he introduces — developing a taste for having problems, embracing radical incrementalism, resisting middling priorities — are practical enough to be useful without collapsing into lifehack-territory. And the Afterword’s call to abandon hope is the book’s most counter-intuitive and memorable move: hope, Burkeman argues via Derrick Jensen and Pema Chödrön, is just another way of placing your faith outside the present moment.

Give it up, and you’re free to actually do the work in front of you.

So why four stars and not five? There’s a faint rehearsed quality to parts of the book — a smoothness, ironically, in how the arguments unfold. Burkeman is a skilled writer, maybe too skilled here: some passages feel like they’ve been polished to the point where the rough edges that would make them truly confronting have been sanded away.

The ending with its ten tools for embracing finitude tilts toward the self-help format the rest of the book has been arguing against; I’m not sure how I feel about them, because the points are useful, too.

I reviewed Tricia Hersey’s Rest is Resistance last year - same broad territory, but where Hersey preaches rest as political liberation (with all the US-centric blind spots that entails), Burkeman makes the case philosophically and personally. He’s more honest about his own complicity in the productivity machine, and crucially, he doesn’t pretend the answer is simple. The answer, if there is one, involves grief: you have to mourn the lives you’ll never live, the options you’ll never take, the versions of yourself that will never exist. That’s harder to put on a poster than “nap more,” but it’s truer.

If you’ve ever finished a productivity book feeling both inspired and somehow emptier, this is the antidote. Burkeman has written the rare time-management book that takes seriously the fact that time will, eventually, manage you [out].

• Rating: 4 out of 5

• Dog-ear index: 11.7

• Who is it for: Recovering productivity addicts; anyone who suspects that the problem with their relationship to time goes deeper than not having the right app. Not for readers seeking actual productivity tips. Burkeman would consider that missing the point entirely.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Four-Thousand-Weeks-Embrace-limits/dp/1784704008

I don’t really know how to review this book. Rick Rubin’s The Creative Act: A Way of Being is a book that, early on, asks you to stop labeling and analyzing, to resist the instinct to categorize. And then here I am, highlighter in hand, kind of trying to do exactly that.

This is a book about creativity by one of the most influential music producers alive, and it contains almost nothing about music. There are no studio war stories, no tales of recording with Johnny Cash or the Beastie Boys.

Instead, Rubin has written something closer to a secular devotional text. There are 78 short chapters he calls “areas of thought,” delivered in very short sentences, with generous white space, structured less like an argument and more like a series of meditations.

There is a very clear core thread: creativity is something you are, not only something you do.

It’s a way of moving through the world.

From page two, Rubin expands the notion of what counts as a creative act to a conversation, taking a new route home, rearranging the furniture in a room. If you perceive, filter, and curate experience, you’re already creating.

In addition to basically all of us being creators, it was liberating for Rubin to spell out that it makes no sense to say you’re “not good at” being creative, any more than you can be “bad at” being a monk.

You’re either practicing or you’re not, and that’s it.

For someone trained to analyze (and I say this as a foresight professional whose job literally contains elements of studying and categorizing emerging patterns) this is a welcome but also slightly uncomfortable invitation.

Have I ever thought I am, or would be, ‘bad at art’?

Absolutely. All the time.

Rubin argues that as soon as you label something, you’re no longer noticing it; you’re studying it. Analysis is a secondary function. The awareness comes first. I know he’s right. I also know that every margin note I made was an act of gentle defiance against the book’s central request.

There’s a spiritual dimension here that some readers will welcome and others will find too untethered. Rubin writes about the Source, about connection, about faith in creative direction without needing to understand it. He’s careful to decouple this from organized religion and uses spirituality as connection rather than doctrine, but the language will be familiar to anyone who’s spent time with Buddhist or contemplative traditions.

There’s a connection to T.M. Luhrmann’s work in How God Becomes Real: both suggest that practice creates the experience, not the other way around.

You don’t create because you’re creative, or practice because you believe.

You become creative because you create, and believe because you practice.

Rubin’s observation that all work is collaboration — with the art that came before, the world you live in, the tools you use, the audience, and who you are today — is one of those insights that seems obvious after (but typically only after) someone says it.

There are unexpectedly sharp moments scattered throughout, almost casually, and in a variety of styles. His chapter on openness, where he notes that we build frameworks that give us reduced options and a false sense of certainty, could have come from a text on cognitive bias or foresight methodology. The passage on curiosity, about how it explores all perspectives, craves constant expansion, and pushes to expose falsely set boundaries, on the other hand reads like a manifesto for anyone whose work involves seeing what others don’t yet see.

Rubin observes that we are interpretation machines, that our explanations are guesses which become fixed as fact; that we are the unreliable narrators of our own experience. On spontaneity, he joins the now-common forces puncturing the myth of effortless genius: the story of spontaneity is misleading, because we don’t see the lifetime of preparation that primes the artist for the “spontaneous” moment.

And then there’s this advice on page 387:

“If you’ve written a book that’s over three hundred pages, try to reduce it to less than a hundred without losing its essence.”

Again, on page 387. Of a book that runs past 400 pages.

Which brings us to one tension that Rubin doesn’t resolve.

The Creative Act would be a more powerful book at maybe half its length. Many of the 78 chapters circle the same ideas in slightly different language. The white space and short sentences create a meditative rhythm that works beautifully for the first hundred pages and starts to feel like padding by the third hundred. The repetition may be intentional; Rubin might, probably correctly, argue that returning to ideas from different angles is itself a creative practice, that the book is meant to be dipped into rather than consumed cover to cover.

Fair enough. But a book that advocates ruthless editing probably shouldn’t need that defense.

I also notice what’s absent. There’s no engagement with the ways creativity is constrained by systems, economics, or power, the very material conditions that determine who gets to really “live as an artist“ to the fullest, and who doesn’t.

The book assumes a degree of freedom and access that is, for many people, more of a real barrier rather than their mindset. I don’t consider this a fatal flaw, but it leaves a gap where something interesting might have been.

This is an overlong, occasionally repetitive book that contains at least a dozen ideas I will continue to carry with me and remind myself of. I’m not sure what rating that deserves.

• Rating: 4 out of 5, give or take at least half a point

• Dog-ear index: 7.2

• Who is it for: Creators of all kinds. With a high likelihood that includes you, even if you don’t think of yourself as one. Especially if you don’t think of yourself as one. Also for anyone who has become so focused on output and analysis that they’ve forgotten creation is a way of being, not just a way of producing. Not for those who need practical creative techniques or step-by-step methodology; this is philosophy, not instruction.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Creative-Act-Being-Rick-Rubin/dp/1838858636/

I need to be honest with you upfront: this book is confirmation bias on steroids. I love writing. I love words. I spent decades working in and around communication technology. I now work in AI & strategic foresight. Naomi Alderman has written a book that sits squarely at the intersection of all those things. As a result, my highlighter barely survived the experience, and the book looks a little ridiculous; just look at the up-there dog-ear index.

Throughout my reading experience, I tried my best to check my biases. Even so, take what follows with a bigger-than-usual pinch of salt, given I was unlikely to succeed in setting all of them aside.

Anyway, Alderman’s thesis is cleanly: we are living through the third great information crisis in human history.

The first was the invention of writing. The second was the Gutenberg printing press. The third is digital communications technology.

Each crisis brought enormous leaps in knowledge and understanding — and also prolonged periods of intense instability, violence, and social upheaval. If you’d known the name of your era, she argues, it would have given you a clue about what to prepare for.

The framework is compelling, and Alderman builds it well. She draws on Walter Ong’s work on orality and literacy, tracing how writing made us value older people less (they were no longer the sole repositories of knowledge), how it enabled independent thinking by freeing people from constantly rehearsing traditional knowledge, and how each new communication technology introduces what she calls a “war of interpretation” — conflict over the meaning of written words with people you’ve never actually met.

There are a lot of observations in the book that come across as fundamentally important. The idea that when people worked with sheep all day, they imagined God as a shepherd; now that we work with computers, we describe ourselves as “hard-wired” or “programmed.” The Borges parallel — The Book of Sand, an infinite, unfinishable book that becomes a prison, as a metaphor that’s too close for comfort for the infinite scroll of social media. The Elizabeth Eisenstein insight that when technology went to press, occult lore flooded out alongside scientific knowledge, and few readers could discriminate between the two.

Well, bugger. Here we are again.

Alderman brings out interesting points about forgetting. She distinguishes between the right to be forgotten — which data protection laws especially in the EU attempt to address — and the right to forget. The imperfect way our memory works, she argues, actually mitigates lasting conflict. When people say “time heals,” what they partly mean is that memories fade, becoming less sharp and less wounding.

Digital technology has stolen that from us. Memories in digital form never fade - on the contrary, they might end up being amplified.

Twenty years turns a shitposting fifteen-year-old into a responsible thirty-five-year-old, but the internet remembers every post. We need, she says, to find ways to say “I am not that person any more“ and to actually mean it, and to let others mean it too. I agree wholeheartedly.

Growing up in a small Finnish town in the early days of the Internet, I recognise what she describes about the pre-digital world: before the Internet, there were so many things people simply never talked about, because there was no space for it.

The internet made it possible to talk about them; to find your tribe globally. I found that exhilirating and liberating in the 1990’s, where IRC allowed me to connect with like-minded people globally. That ability, Alderman acknowledges, is both a liberation and a source of enormous friction. The discovery of how many radically different lives are being lived in parallel to our own is disorienting. It can push us toward empathy or toward our own enclosed private pews, shutting out everything that challenges us. Alderman calls this the “box-pew effect”: individualism paradoxically leading to more rigid think-alike communities.

There’s much practical wisdom scattered throughout, such as pausing before reposting when you feel a strong emotion and starting hard conversations with what you agree on. Her personal rule — never talk about a culture-war topic with anyone who only wants to talk to you about that topic — is the kind of heuristic worth adopting immediately. And her call to protect institutions like Wikipedia, public libraries, and the Archive as information services that aren’t trying to sell you something feels both urgent and underappreciated.

When reading, I agreed with so much of the book that my gut reaction was to give it 5/5 just for that. But there are limitations, and I want to be fair about them.

The book is light on hard evidence and meticulous referencing. Alderman is working as an essayist, not a researcher, and the Reformation parallels don’t necessarily always hold up under scrutiny. The analogy between Luther’s disintermediation and Uber’s is suggestive but not what I’d call rigorous. And Alderman herself acknowledges, in her afterword, that the AI dimension is underdeveloped: “we’re not in the midst of it yet,” she writes.

On that point, I disagree strongly. We are, and we were squarely “in the midst of it” in 2025 when this book was published.

As a more general point, here’s what I find somewhat unnerving about books that resonate strongly with me. Often — not always, but often — when they touch on a topic I know quite a bit about, I find them wanting. Alderman’s treatment of AI is a case in point. She’s used Claude, finds it useful, acknowledges its limitations. But she treats the default sycophantic behaviour as inherent rather than configurable. She frames AI companionship as a supernormal-stimulus risk — which is a fair concern — but doesn’t engage with how the technology is actually developing.

What’s my point? Where the book doesn’t touch my areas of expertise, it feels like brilliant 5/5 work. Where the book does touch on topics of my expertise, it’s not. It’s not terrible either by any means, but it’s not 5/5.

The point is that I suspect it would be a very strange coincidence if the analysis were less than rigorous only in the areas I happen to know deeply.

I’m choosing not to penalise the book for those assumed flaws, but it’s a pattern worth noticing, and it tempers what might otherwise be a higher rating.

What Alderman does well — and this is no small thing — is name what an information crisis feels like. The anxiety, the anger, the sense that we’re surrounded by cretinous vicious imbeciles (her words) when actually most people around us are careful, thoughtful people who may disagree with us for good reasons. The loss of shared consensus reality. The feeling that the ground is shifting under our feet.

Naming the era you’re living through doesn’t solve it, but it helps you stop blaming yourself for feeling destabilised by it.

The book also echoes something Melvin Kranzberg captured more concisely: technology is neither good nor bad — nor is it neutral. Writing enabled both the Axial Age’s moral philosophy and centuries of religious war. The printing press gave us the Enlightenment and the Holocaust’s foundational texts.

The internet gives us Wikipedia and QAnon. Alderman’s contribution is to trace this pattern across three crises and invite us to learn from the previous two. The invitation is worth accepting.

She ends with hope, but not naive hope. Brains evolved for small groups of a few thousand are trying to build moral consensus among billions. She argues that even as we write ourselves into all sorts of trouble — moral, spiritual, physical, psychological, intellectual — we can also use these same technologies to work ourselves out of them again.

It’s a warm, humane ending.

I’d like to share her optimism fully. I’m not sure I do.

• Rating: 4 out of 5 (so, still very good)

• Dog-ear index: 14.7 (i.e., ridiculously high)

• Who is it for: People who sense that something has gone deeply wrong with how we communicate and want historical context for the feeling. Writers and communication professionals will find the most to chew on.

reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Dont-Burn-Anyone-Stake-Today/dp/1405981393

What if the most reliable thing about your memories is that they’re wrong?

I’ve spent a lot of time in my life thinking about how humans reconstruct events after the fact — in revisionist history of our personal lives, in incident investigations, in organisations trying to learn from failure.

Memory is always at the centre of it. We build narratives. We fill gaps. Even when we acknowledge that memory can be fallible, we are often supremely confident about our memory, and about the veracity of the details we’ve just fabricated wholesale. In interpersonal arguments, we’re quick to doubt the other person’s narrative of the events, while remaining confident about our view.

Mark Rowlands’ The Book of Memory takes that discomfort and runs with it, arriving somewhere a little different than your standard book about memory: a philosophical meditation on what it means to be a person at all, written as a love letter to his sons.

Rowlands is a professor of philosophy at the University of Miami, not a neuroscientist. This is a book that uses science — reconsolidation, flashbulb memories, protein synthesis — but in service of philosophical argument rather than as an end in itself.

If you want a more rigorous deep-dive into memory science, Charan Ranganath’s Why We Remember is your book. Rowlands is after something different: what the science means for our understanding of selfhood.

Rowlands introduces the concept of being “ontically fat” versus “ontically thin” — a philosophical distinction I hadn’t encountered before and felt a little weird about at first.

You and I are ontically fat: there are facts about us that exist whether or not anyone knows them. Sherlock Holmes, by contrast, is ontically thin — there is no fact about what he had for breakfast on April 24, 1888 unless Conan Doyle wrote one. What Rowlands argues, provocatively, is that our memories make us partly fictional. Every time a memory is recalled, it must be reconsolidated — rebuilt, essentially — and each rebuilding can introduce distortions.

As such, the memories that make you you are, to a significant degree, works of authorship rather than records of fact.

He builds this argument through different pieces. The chain of events leading to his son’s conception — stretching back through a rebound relationship, a disastrous trip to India, a case of dysentery from a restaurant in Connaught Circus, New Delhi — is a delightful illustration of contingency.

His treatment of metaphor as “an invitation — why don’t you think of things in this way?“ is the kind of simple insight that makes you stop and sit with it for a moment. And the chapter on Stendhal’s syndrome - dizziness, fainting, even hallucinations triggered by exposure to overwhelming quantities of art - was entirely new to me.

The science, when it appears, is mostly solid. The Neisser and Harsch study of Challenger explosion memories is a well-established classic: students’ accounts of where they were when they heard the news changed dramatically between 1986 and 1989, while their confidence in those memories remained stubbornly high. Eleven of forty-four subjects scored zero for accuracy but still expressed high confidence. Karim Nader’s work on reconsolidation — showing that blocking protein synthesis during memory retrieval can effectively erase the memory — is also well-replicated and forms the backbone of Rowlands’ argument that remembering is fundamentally an act of rewriting.

But I said mostly. Where I’m less convinced is the treatment of PKMzeta, a protein Rowlands presents as crucial to both the formation and retrieval of memories. The PKMzeta story has become considerably more complicated since the initial findings. Knockout mice lacking PKMzeta entirely still form and retain memories, and the inhibitor ZIP — once thought to specifically target PKMzeta — turns out to have broader, non-specific effects including suppressing neural activity altogether. The picture has been partially rescued by more recent work showing that a related protein (PKCiota/lambda) compensates for missing PKMzeta, but the neat narrative Rowlands presents is more contested than he lets on. For a book leaning on neuroscience to support philosophical claims, this matters.

Some of the most interesting moments come from Rowlands’ notion of an “existential style” — the idea that even when the content of memories erodes, such as with people with Alzheimer’s, what remains is something like a characteristic way of being.

He illustrates this through his father-in-law Patrick, whose dementia stripped away specific memories but left his essential character — the quiet storytelling, the gentle erudition — perfectly visible. It’s a moving observation which probably, unfortunately, doesn’t neatly generalize, for I have equally heard stories of people who change to something very different in similar circumstances.

Nevertheless, it raises a question Rowlands doesn’t fully explore but that I couldn’t stop thinking about: if so much of our book of memory is dominated by redaction — vast seas of black ink — and our selves are relatively shallow constructs built from sparse islands of recollection, does this explain why our technology, which knows only a thin slice of us, can feel like it knows us so well?

Rowlands also makes a fascinating case that externalising memories — telling people about them, writing them down — can stabilise them against the distortions of reconsolidation. But he immediately complicates this: stability is not the same as accuracy. You can lock in a memory by talking about it, but if your account was inaccurate to begin with, you’ve now fixed the inaccuracy in place.

This sent me down a rabbit hole about journaling. If writing freezes memories in their current (possibly already distorted) state, could the practice of journaling actually be harmful in some cases — cementing false unhelpful narratives about our own lives? (Noting that some false narratives can be helpful). Rowlands doesn’t go there, and I suspect it’s not quite that simple, but the question of the possibility lingers.

The book is ultimately structured as a bequest. Rowlands is writing for his sons, passing on memories that will make his being “no longer my own, but resting on your retrieval and subsequent rewriting.” It’s a philosopher’s version of inheritance, and there’s something tender about it. He compares himself to Sherlock Holmes — a fictional character whose being depends entirely on others — and finds liberation rather than loss in the comparison.

An incompletable being can never die.

The prose throughout is warm and conversational, if occasionally a bit meandering. Rowlands has a habit of circling back to the same ideas from different angles, which serves the philosophical method but occasionally tests patience - though at 139, it’s still pretty short.

• Rating: 3.5 out of 5

• Dog-ear index: 7.1

• Who is it for: Philosophers who want some science, or science enthusiasts who want some philosophy. People grappling with questions of identity, memory, and what remains of us when the facts fall away. Not for readers wanting rigorous neuroscience or a complete treatise of how memory works.

reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Book-Memory-How-Live-Forever/dp/1803512644

On one hand these stupid things are a useful reminder; on the other hand, they’re performative catering to confirmation bias.

They weren’t wrong about the output; the AI did fail.

Or, rather, their AI failed.

As a result, they are wrong about what think they demonstrate. They’d only tested their particular configuration of AI — which, based on most of the screenshots, are free-tier models with no customisation, no context, and no instructions beyond a single prompt typed into a chat box.

That’s a bit like test-driving a bicycle, concluding that vehicles can’t exceed 25 km/h, and proceeding to laugh at people who think they can drive to Sydney in a day.

I see this constantly, from both sides.

The enthusiasts running agentic frameworks and custom workflows who can’t understand why everyone else doesn’t see it.

And the skeptics who dismiss it entirely are usually running the default models.

At least sometimes, they’re both reporting their experiences honestly. But they’re not talking about the same thing, so they talk past each other, and irrespective of that, things are a little bit more complicated than either end would want you to believe.

This is the problem I want to dig into today, because it has consequences that go well beyond social media arguments. When organisations make decisions about AI strategy — how much to invest, where to deploy it, what it can and can’t do — those decisions are shaped by the experiences of the people in the room. And since those experiences vary so wildly, depending on how the tool is configured, most statements about “what AI can do” are essentially meaningless without specifying the setup in great detail.

The answer to almost every question about AI capability is “it depends.”

The Six Layers

To make this concrete, let me walk through a single, common task and show how the output transforms as you move through six layers of configuration.

Note: I am not going to touch on experimental agents or frameworks like OpenClaw here. In my judgement, they’re not ready to be used by ‘normal’ people. What I have are six layers that are ready for prime-time - doesn’t mean it’ll always be only six layers.

But first, a note about something that cuts across every layer: how you prompt matters. The quality of what you ask — how specific you are, how much context you provide, whether you iterate on the first answer or accept it — affects the output at every level.

A thoughtful, detailed prompt at Layer 1 will outperform a lazy one-liner at Layer 3. Prompting is still a skill and it’s the one variable that compounds with everything else. I’ll come back to this, but keep it in mind as we go.

Layer 1: The Default

This is where most people’s experience begins and ends. You go to ChatGPT or whatever tool your organisation provides. You type something like: “Help me prepare for a meeting with a healthcare exec team about AI training.”

You get back… advice. Fine, generic, sensible, forgettable advice. Research the company. Review their annual report. Prepare an agenda. Think about what questions they might ask.

Duh.

It’s not that it’s wrong, it’s just not very useful if you’re not a complete newbie into the world of business. You could have found the same suggestions in any business article written in the last thirty years. The AI doesn’t know who you are, who the client is, what your expertise is, or what you’re trying to achieve. It’s working with nothing but your twelve-word prompt and whatever statistical patterns it learned during training.

(This is where prompting skill matters most, by the way. At Layer 1, it’s all you’ve got. A detailed, context-rich prompt — explaining your role, the client’s background, the meeting’s purpose — would already produce a noticeably better result. But most people never try this, because their frame of mind is in testing the tool, not actually using it.)

This is the experience that produces most of the skepticism. And honestly? The skepticism is fair — for this level and this configuration.

The mistake is assuming this is all there is.

Layer 2: Upgrading the Engine

The simplest improvement is one that a surprising number of people still don’t know about: paying for the tool and selecting the best model.

Every major AI platform offers multiple models. Some are faster but less capable. Some are specifically designed for complex reasoning — often called “thinking” or “reasoning” models — that will actually work through a problem more in a step by step manner before answering. In most cases, this is extremely helpful.

In all the three frontier tools (Claude, ChatGPT, and Gemini), you can select different models from a dropdown menu. Most people have never clicked that dropdown. Many don’t know it exists, and others don’t think it matters. I don’t blame them; the labs have traditionally, in technical terms, sucked at the UX of this and the model naming has been less than helpful to be charitable.

The paid tier also typically means a larger context window (the AI can hold more information in its head at once), access to better models, and fewer restrictions on usage.

Same meeting prep prompt, better model: you now get a more structured analysis, perhaps some industry-specific considerations for healthcare AI adoption, a more nuanced discussion of stakeholder dynamics. It’s noticeably better. But it’s still generic, because the AI still doesn’t know anything about you.

Layer 3: Fixing the Personality

This is where most people leave enormous value on the table.

Every major AI platform lets you set custom instructions — persistent context that the AI reads before every conversation. Most people either don’t know this exists or have written something vague like “be concise and professional.”

But custom instructions can do far more than set a tone. You can tell the AI who you are, what you do, how you think, what you value, and how you want to work together. You can tell it to challenge your assumptions rather than validate them. You can tell it to match your communication style, understand your industry context, and skip the generic preamble.

The difference this makes is profound and underappreciated. With well-crafted instructions, the AI stops being a generic assistant and starts being your assistant. It frames meeting prep through your professional lens. It knows you’re a consultant, not an employee. It knows your approach to client relationships. It adjusts its depth and tone to what you actually need rather than what an average user might want.

The default behaviour isn’t neutral, but rather calibrated for the widest possible audience, which means it’s optimised for no one in particular, and quite often for engagement and not being offensive. Fixing the personality is fixing a misalignment that most people either don’t notice or are actively pissed off about (like the models being sycophantic), but don’t realise how easy it is to fix.

Layer 4: Teaching It Your Method

Here’s where it gets more interesting. Custom instructions can tell the AI who you are. But skills — reusable structured workflows — tell the AI how you want specific types of work done.

Think of the difference between hiring someone smart and hiring someone smart who’s been trained in your methodology. Skills are what I used liberally in my exploration of using GenAI for foresight that I talked about in the Futures Dramaturg article.

A skill for your client meeting preparation might include a specific structure: stakeholder mapping, engagement history review, risk and opportunity identification, industry context, and a pre-drafted agenda with discussion prompts. Every time you run it, the output follows your process rather than improvising a new approach each time.

This matters because consistency is where compounding efficiency lives. A one-off clever answer is nice. A repeatable workflow (or, better, an automatically reporting workflow) that produces reliably excellent prep for every meeting, every time, following your standards — that changes how you operate.

Most AI users are improvising every interaction from scratch, so quite naturally they’re getting different formats, different depths, different structures each time. Skills reduce that variance. You design the process once, and then it gets executed.

Layer 5: Reaching Out for Context

Up to Layer 4, the AI is working with your instructions and whatever you type into the prompt. It’s smart, it knows stuff about you and is personalised, it has your methodology, but it’s still essentially imagining the situation based on what you tell it.

At Layer 5, the AI connects directly to your data: Your calendar. Your email. Your notes. Your files.

For our meeting prep example: instead of asking you who the meeting is with and when it’s happening, the system reads your calendar and knows. It searches your email for recent exchanges with this client. It checks your notes for the history of the engagement — what you proposed, what they said, what’s outstanding. It finds the proposal you sent them three months ago and the follow-up that went unanswered.

The brief it produces isn’t then based on generic advice or even your methodology alone. It’s grounded in actual information about this specific meeting with this specific client. Things you might have forgotten or never thought to mention, or that simply would’ve been a lot of work to manually add.

This is the layer that makes most of the “AI can’t do real work” crowd look like they’ve been reviewing a different product entirely, because they have been.

Layer 6: Adding Depth

The final layer I will be talking about here extends the AI’s reach beyond your personal data to external knowledge sources — academic databases, computational tools, verified research, real-time data.

For the meeting prep: the AI now searches peer-reviewed literature on the topics you’ll be discussing. It pulls current industry reports and regulatory updates relevant to the client’s sector. It can compute financial projections or model scenarios if the meeting involves quantitative decisions. It accesses specialised databases that you’d normally need to search manually — or more likely, wouldn’t search at all because you didn’t have the time.

As a result, you walk into the meeting with knowledge that neither you nor the AI had in isolation. You’ve been augmented in the truest sense — not replaced, not automated, but operating with a depth and breadth of preparation that simply wasn’t possible before.

This is what advanced AI use actually looks like today.

The Branching Problem

So here’s where this connects to the broader challenge, and where it gets worse than a simple ladder model might suggest.

The six layers describe form the vertical differences in what people have and use, but there’s a horizontal one too. Even two people operating at the same layer will have fundamentally different experiences, because every choice they’ve made along the way creates divergence.

At Layer 3, your custom instructions are different from mine. You’ve told your AI to be formal and cautious; I’ve told mine to challenge my thinking and skip the preamble. Ergo, different tool. At Layer 5, you’ve connected your Outlook calendar and SharePoint; I’ve connected Gmail, Obsidian, and a CRM, so we get different universes of context. At Layer 6, you might be pulling from a legal database; I’m pulling from peer-reviewed academic research and computational tools. Completely different intelligences that excel, and fail, at different things.

The result is a tree.

Every configuration choice is a branching point. Model selection, custom instructions, which skills you build, which data you connect, which external tools you integrate, and underneath all of it how well you prompt. The further you go, the more branches there are, and the more personal the tool becomes. Two people at Layer 6 might be running systems so different from each other that comparing their experiences is barely more meaningful than comparing one person’s Layer 1 to another’s Layer 6.

This is why “what AI can do” is becoming a nearly meaningless question.

It’s like asking “what can software do?” The answer depends entirely on which software, configured how, by whom, for what purpose. We don’t ask that question about software because we intuitively understand the space is too vast. We haven’t yet developed that intuition for AI, but we need to, quickly, because the decisions being made in its absence are consequential.

When a CEO tells their board “we’ve tried AI and it’s not that impressive,” what were they running? What layer? What configuration? What instructions? What connections?

Was it stock-standard Copilot?

When a consulting firm says “AI saves our team ten hours a week,” what does their tree look like?

When LinkedIn’s main character of the day dunks on a chatbot for getting a maths problem wrong, which branch were they even on?

The answer matters enormously.

We’re in a moment where organisations are making consequential decisions about AI based on experiences that span a staggeringly wide range of capability, and the people making those decisions usually don’t fully understand the range that exists. They think their experience is the experience. It isn’t. It’s their experience of their configuration - one branch, or leaf, of an enormous tree.

In aviation, mode confusion — where pilots misidentify or lose track of what mode the aircraft automation is in — can lead to automation surprise, where pilots are startled by unexpected automated behaviour.

The AI equivalent is happening right now at an organisational level. People are surprised by AI’s limitations or capabilities because they don’t understand what configuration they’re looking at.

The LinkedIn dunkers are the most visible symptom, but they’re not the real problem. The real problem is the executive who used the default for twenty minutes, formed a confident opinion, and is now shaping their organisation’s AI strategy based on the experience of seeing one fallen twig from a giant tree.

What To Do About It

One caveat before the practical advice: the point of understanding these layers is not that higher is always better. Warren Vanderburgh’s original Children of the Magenta thesis — the one this Substack is named after — makes the strong case that pilots need to know when to reduce automation.

The same applies here. There are moments when you should deliberately drop down a layer or more: when the AI’s context connections are pulling in noise instead of signal, when you need your own unfiltered thinking before the AI shapes it, when a simple prompt to a clean model will give you a better answer than a complex workflow that’s optimising for the wrong question.

Knowing your way around all six layers means knowing which one to reach for in any given moment, including Layer 0 of no AI whatsoever.

If you’ve read this far and suspect you might be operating at a lower layer than you could be, here’s the honest answer: moving up is not trivial. Each layer requires some combination of investment, configuration, and, importantly, a shift in how you think about the tool.

But you can start immediately:

If you’re at Layer 1, the single highest-impact move is upgrading to a paid tier and spending five minutes learning which model to select for which task. That alone will change your experience more than anything else.

If you’re at Layer 2, write proper custom instructions. Not “be concise.” Tell the AI who you are, what you do, how you think, and what you expect. Spend an hour on this. It pays for itself in the first week.

If you’re at Layer 3, start thinking about the tasks you do repeatedly and whether you can structure them into consistent workflows. This requires more thought than the previous layers, but the leverage is big.

If you’re at Layer 4 and beyond, you’re already way ahead of most. The returns from here come from deeper integration; connecting more data sources, adding specialised external tools (many of which, yes, you will need to pay for), and refining your workflows as you discover what actually saves you time versus what’s merely impressive. Keep experimenting. The frontier moves fast, and at this level, you’re well-positioned to see it first.

The point isn’t to make everyone a Layer 6 user. The point is to make sure that when you form an opinion about what AI can and can’t do, you know which branch you’re standing on, and can calibrate accordingly.

If you’re making decisions that affect your organisation, your team, or your career, you should probably explore a few more branches before you decide what you think of the view.

If you’re a senior leader who wants to move beyond Layer 1-2, I run a bespoke GenAI coaching programs designed to get you to the level where AI actually transforms how you work, not just how you talk about it in your exec or board meetings. Book a free initial call here to discuss.

Some disclosures before we begin. First, I know the author. Second, this book sits squarely in my professional domain. Third, the thesis aligns so closely with what this Substack has been exploring that my confirmation bias is probably doing a victory lap. I tell you this not to undermine what follows, but because intellectual honesty requires it, and because the book itself would demand nothing less.

Steven D. Green’s Pilot Competency and Capability is an academic text that takes on one of aviation’s deepest philosophical tensions: the gap between what we say the pilot is — the final authority, the capable and competent aviator — and what the industry has been systematically building the pilot into: a systems operator within a cybernetic control loop.

Green’s argument is built on margins. Not just the margins pilots think about every day like landing distances, fuel reserves, and obstacle clearance, but also on the deeper margins that make safe operations possible at all: the margin of residual attention, the margin between work-as-imagined and work-as-done, the margin between what can be specified in a procedure and what actually keeps an airplane safely on a trajectory through an open, complex, tempestuous environment.

Green argues these margins exist to absorb the unforeseen. They are not there so that operational optimization could find more things to shave time or cost off of. When we shrink them — whether to save fuel, to increase throughput, or to squeeze a few more seconds out of a turnaround — we are spending safety capital that was never ours to spend.

It is, to me anyway, very easy to see how these concepts generalize broadly to our society today, with the exception that the society at large has none of the defenses against that encroachment that the aviation industry has built over decades of literal blood, sweat, and tears.

This is not a new argument in safety literature, but Green grounds it with unusual depth. He traces the concept of prudence to its Latin root, providentia — foresight — and builds a framework around St. Thomas Aquinas’s structure of practical reason: take counsel, judge what you have learned, then execute command. Command, in Green’s telling, is the action that follows judgment. It is what people do with their free will. And it is precisely the thing that gets compressed when we proceduralize everything, because the cybernetic approach to automation treats standard operating procedures as lines of code and the human being as a machine executing them.

One could say I’m a fan of cybernetics, or at the very least drawn to some of its concepts, so when Green takes aim at it, I was on guard. For the cybernetic approach to work, Green argues, the operating environment must be a closed system.

Aviation is no such thing. One could argue very few things operating out there in the real world are. Life is no such thing.

The cybernetic approach to automation, he writes, is a digitization of Herbert Simon’s 1947 premise that two people with the same information will rationally decide upon the same course of action — a premise Green identifies as a definitive statement of determinism, invested in what Sidney Dekker calls naïve Newtonian scientism. The belief that if we can just specify enough variables, control enough inputs, and standardize enough outputs, we will achieve safety.

We won’t.

And Green’s own experience gives this weight. Early in the book, he recounts a 1978 crash in a New Hampshire forest that killed his best friend and very nearly killed him. Over forty years later, he still doesn’t know what caused the airplane to sink into the trees on a calm summer evening.

The hidden axiom of the capable-and-competent pilot is the unstated assumption that the pilot will always see the threat coming.

Tell that to Richard de Crespigny of QF32 or Kevin Sullivan of QF72.

The accident you’ll actually have is the one you won’t see coming, and the capability to manage that uncertainty is exactly what gets eroded when we replace judgment with procedure. Those who see automation as the path to safety should note that the tail of things that can and does go wrong is far longer than any system designer can specify.

Readers of this Substack will recognise the through-line. What Green describes as the margin of residual attention (the capacity that must be protected so the pilot can identify stuck programs, spot unruly trajectories, and shoot them down) is the operational face of what James C. Scott calls métis: practical knowledge that resists codification because the environments in which it operates are too variable for formal procedures. Green’s entire framework of prudence, command, and the ecology of action maps onto Scott’s argument about what gets destroyed when systems demand legibility above all else.

And it connects directly to what I called the grace margin here — the space between what a system prescribes and what a person actually does. Green’s observation that controlling and anticipating workers’ compliance naturally generates pushback from people who desire to retain an identity of craftsmanship is the aviation-specific version of a universal truth: when you proceduralize everything, you don’t get compliance. You get shortcuts, workarounds, and optimizing violations, because the people doing the work understand something the system designers don’t.

The book is dense. It is academic in tone and structure, rich with references to Amalberti, Morin, Weick, Reason, Dekker, and Hollnagel. It covers a lot of territory like ampliative reasoning, self-organized criticality, fat-tailed probability distributions, and cognitive dissonance theory, some or all of which will be unfamiliar to many people. Green doesn’t condescend, or explain to the lowest common denominator. He doesn’t simplify where simplification would distort. This is not a book for a casual audience, and it doesn’t pretend to be.

It is also, inevitably, a book rooted in American aviation regulation. The FAR references, the NTSB case studies, the FAA-centric regulatory framework are appropriate to the subject but will require some translation for readers operating under EASA, CASA, or other regulatory regimes. The underlying arguments transcend jurisdiction, but the direct examples don’t always.

Amidst all that complexity, it can be easy to lose some of the key messages. Thankfully, they’re driven home with more plain language as well, such as this:

If you read one sentence from this review, make it that one.

Green closes by returning to Saint-Exupéry’s elemental divinities (the mountain, the sea, the wind) and arguing that automation, like the sky itself, is an open system. We can, at best, debate on terms of equality with it. We must never assume superiority to it. And we must vest the pilot with the authority and the capability to manage that debate, because the responsibility is towering, moral, and it can reside nowhere else.

• Rating: 4.5 out of 5

• Dog-ear index: 12.3

• Who is it for: Aviation professionals (especially pilots, instructors, safety specialists, and human factors researchers) who want the philosophical and theoretical underpinning for why proceduralization and automation alone will not be sufficient. Also for anyone working in safety-critical systems who suspects that the industry’s faith in procedures and technology has outpaced its understanding of what actually keeps complex operations safe. This is not a casual read. It will reward those who come to it with some background.
  

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

This first chapter of the book is published as Open Access, so it is freely available online at https://library.oapen.org/bitstream/handle/20.500.12657/62009/1/9781003369677_10.1201_9781003369677-1.pdf. For the full book, please support your local bookstore where possible (who are not likely to have it in stock, but should be able to order it for you); Amazon product link for reference: https://www.amazon.com.au/Pilot-Competency-Capability-Responsibilities-Strategy/dp/1032439742/

We teach obedience extraordinarily well. We teach it in homes, in schools, in workplaces, in militaries. The vast majority of the time we reward compliance and punish deviation.

And then, when someone follows orders into a catastrophe — financial fraud, medical harm, atrocities — we act surprised. We ask, “How could they do that? Why did they just go along with it?“

Ira Chaleff’s Intelligent Disobedience wants to address that gap. The title concept is borrowed from guide dog training: a guide dog must learn when to disobey a command that would put the team in danger. When a handler says “forward” at an intersection, but a car is coming, the dog must refuse. Learning not to obey, the trainers say, is a higher order of skill.

Chaleff argues we need to develop the same capacity in humans, and he’s right. But is this book, published over a decade ago, still an applicable guide for getting there?

I should disclose my angle here: I’ve spent a significant chunks of time around the aviation industry, where the problem Chaleff describes was identified, and substantially addressed, decades ago. Crew Resource Management (CRM) emerged in the 1970s and 80s precisely because junior officers were failing to challenge captains making deadly errors.

CRM was the aviation industry’s answer to exactly this problem: structured, trained, practiced authority-challenging within hierarchies. So when Chaleff arrives at this territory, I’m reading with the eyes of someone who has seen the solution work at scale, and who notices where the book falls short of it.

To be fair, the book covers a lot of ground well. The “algorithm of obedience” framework, quoted below, which lays out the conditions under which compliance is appropriate, and the parallel conditions that should trigger resistance is genuinely useful as a teaching tool. There is also a useful training list, with an emphasis on practicing disobedience with a strong voice and commending it when it occurs.

• I am receiving a rule or order from a legitimate source, not from a random direction.

• I understand the rule or order, what its goal is and what is expected of me in achieving that goal.

• The order is good, or at least neutral in terms of the impact it will have.

• Because no serious harm will result from implementing the order and no core value is being violated, I will obey the order.

The Milgram experiments, which form some of the backbone of much of the book, can get tiresome for anyone who’s read anything about this topic. Yes, we know: people obey authority figures even when told to administer electric shocks. If you’ve read any popular psychology in the last forty years, this is well-worn territory, and Chaleff’s extended treatment of the original experiments feels like it’s padding for the familiar.

However, then Chaleff covers some of the lesser-known variations of Milgram’s experiments. The finding that compliance dropped from 65% to 40% when the subject could see the victim, and to just 30% when forced into physical contact with them, tells us something important: distance enables obedience. It works both ways: when the authority figure left the room and gave orders by phone, compliance plummeted to 20%.

This matters enormously right now.

The most frightening variation of Milgram’s experiment isn’t about distance at all, it’s about role. When subjects weren’t the ones administering shocks but were assigned ancillary tasks like reading questions or documenting answers, compliance shot to 90%. They weren’t pulling the lever; they were just doing their bit within a system that happened to cause harm downstream.

They were comfortably distant from the harm.

As Chaleff observes, most of us today occupy the equivalent of the analyst’s role — we’re not directly causing harm, we’re processing data in a contracting office, cleaning up statistical noise in a drug trial, providing affordable school lunches that happen to contribute to juvenile diabetes.

We’re upgrading our phones to the newest user-friendly technology without dwelling on the factory conditions that produced them. The mechanism that neutralises our moral resistance is removal from direct causation. And in an age of drone warfare, algorithmic decision-making, and bureaucratic chains so long that no individual feels the consequences of their actions, the conditions for that kind of compliance have never been better.

That insight alone is worth the read. But it sits alongside material that hasn’t aged well. The case studies of teachers pressuring to cheat on standardised tests, police underreporting crime statistics, football players urged to deliver concussion-causing hits all probably felt illustrative in 2015 when this book was written.

In 2026, after everything we’ve collectively witnessed, they feel quaint and almost inconsequential. The book’s examples operate in a world that seems smaller and simpler than the one we’re navigating now.

There’s also a scope limitation that nagged me throughout. Chaleff is explicit that intelligent disobedience “works within a system rather than challenging the system itself.” He clearly distinguishes it from civil disobedience; intelligent disobedience accepts the legitimacy of the hierarchy while pushing back on specific orders.

I understand the distinction, and it’s useful as far as it goes. But the harder question goes largely unaddressed: what do you do when the system itself is corrupt, when the mission isn’t positive, when resistance within the system will just get you fired, or worse?

The book assumes a fundamentally sound structure with occasionally flawed commands. Today, that’s a highly optimistic frame, and in many of the situations where disobedience matters most, it’s an inadequate one.

The strongest chapters deal with education and development. Chaleff’s argument that we train obedience into children without simultaneously developing the judgment to know when disobedience is warranted is compelling and undersold. The observation that critical thinking programs teach the thinking but not the acting — not the competence to express dissent in the face of positional authority — is sharp.

We teach children to analyse; we don’t teach them to resist.

We should.

If you’re drawn to this topic — and you should be — I’d actually first point you toward Sunita Sah’s Defy: The Power of No in a World that Demands Yes, which I reviewed last year and rated significantly higher. Sah covers fundamentally similar ground but with sharper tools, more contemporary framing, and a five-stage model of defiance that better captures the psychological reality of why we stay silent when we shouldn’t.

That said, Intelligent Disobedience was ahead of its time in 2015, and the core message remains both important and urgent. We need people who can distinguish between legitimate authority and illegitimate orders. We need to teach resistance as deliberately as we teach compliance. And we need to reckon with the fact that in an increasingly automated, bureaucratised, algorithmically mediated world, we are more removed from the consequences of our decisions than ever, which means the conditions for blind obedience have never been better.

The guide dog metaphor is a good one.

But a decade on, we need more than metaphors. We need the training program.

• Rating: 3.5 out of 5

• Dog-ear index: 7.4

• Who is it for: People new to the obedience/authority literature who want an accessible introduction; educators thinking about how to balance discipline with independent judgment; leaders who say they want to hear dissent but haven’t built the structures to support it. If you’ve already read books like Sah’s Defy or have a background in CRM, human factors, or organisational psychology, much of this will be familiar ground.

reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Intelligent-Disobedience-Ira-Chaleff/dp/1626564272

Most philosophy books pretend to be neutral while smuggling in their biases anyway. Michael Huemer’s “Knowledge, Reality, and Value: A Mostly Common Sense Guide to Philosophy“ does something more honest and more interesting: it tells you upfront that it aims to be fair, but not neutral. He’s going to tell you what he thinks, and argue for it, and if you don’t like that, he suggests you go get another book. He even names one.

I respect that. It’s a harder standard to hold yourself to than pretend to be “objective”, and it signals that Huemer trusts his readers to evaluate arguments rather than needing to be guided to the “right” conclusion. Whether he consistently meets that standard is another question, but the disclosure itself earns points.

The book is a wide-ranging introduction to philosophy — epistemology, logic, ethics, free will, philosophy of religion, political philosophy — filtered through Huemer’s consistently libertarian-leaning, intuitionist lens. The subtitle's 'mostly common sense' is doing real rhetorical work here, framing his fairly opinionated positions as something any reasonable person would think if they just thought carefully enough. Sometimes that’s the case; other times it may not be.

There’s a lot of good stuff here on critical thinking. The treatment of fallacies is great; Huemer walks through anecdotal evidence, base rate neglect, whataboutism, cherry picking, and selection effects with the kind of clear, example-driven writing that makes you wish more philosophers wrote this way.

His base rate neglect example involving a rare disease and a 90% accurate test is the sort of thing everyone should encounter at least once in their education. His treatment of whataboutism is sharp too, as he argues that the practice systematically prevents evils from being addressed, because for any wrong in the world, you can always point to some other, worse wrong elsewhere, deflecting attention indefinitely. I’ve seen whataboutism play out far too much in domains ranging from boardrooms to social media my entire adult life.

Huemer also offers one of the better insights I’ve encountered on the relationship between bias and knowledge: that the factors making someone biased about a topic are often the very same factors making them knowledgeable about it. A war veteran discussing war is likely both the most biased and the most informed person in the room. Discounting “non-objective” perspectives therefore risks throwing out the perspectives of the most knowledgeable people. That observation has implications well beyond philosophy, and is relevant to anyone who has ever dismissed an expert’s view because they were “too close to the subject.”

And then he turns the lens on you:

Yet while the vast majority of people are dogmatic, no one thinks that they are. You, reader, are probably dogmatic, but you think you’re not. That’s partly because the word ‘dogmatic’ sounds insulting, and hence it is unpleasant to entertain the hypothesis that one is dogmatic. To make it sound less bad, you can just replace it with the description, “systematically underestimates appropriate belief revision”. You probably systematically underestimate how much you should revise your beliefs when you acquire new information, because the vast majority of people do that, but you probably don’t realize that you do this.

All the greatest hits are here: Pascal’s Wager, the omnipotence paradox (can God create a stone so heavy he can’t lift it?), the tension between omniscience and free will, Nozick’s Experience Machine, deontology versus consequentialism. For readers encountering these for the first time, the coverage is clear and engaging. For those who’ve already wrestled with them, it reads more like a well-organized refresher.

The chapters on epistemology get complicated quickly — Huemer himself acknowledges this transition — and the later material on ethical theory is where his intuitionist framework does the most work and also where a critical reader might push back hardest.

I am probably not be the target audience for this book. Much of the terrain was familiar, and I found myself nodding along more than being challenged. Huemer’s writing style of dry humor worked well for me, so it was fine; pleasant, even. And the pre-existing knowledge is, of course, not a criticism of the book so much as an acknowledgment of audience fit.

For someone coming to these ideas fresh, like a high school or maybe a university student or others who haven’t yet thought systematically about what rationality means or why being irrational might be morally blameworthy, this would be an excellent starting point.

Huemer writes with personality and a directness that is rare in academic philosophy. His style reminded me at times of what I appreciated about Simon McCarthy-Jones’ “Freethinking”; both authors trust their readers enough to take clear positions and defend them openly.

The question I didn’t fully come to a conclusion about is whether “fair but not neutral“ earned its keep or whether this collapses into “persuasive essays with a disclaimer.” On balance, I think Huemer mostly earns it; the libertarian lens does run through so consistently that readers should be aware they’re getting a survey of philosophy from a particular vantage point, which is still kind of presented as the view from nowhere.

• Rating: 3.5 out of 5

• Dog-ear index: 4.3

• Who is it for: Philosophy newcomers who want a clear, opinionated, and engaging introduction to the big questions, and who are comfortable being told what the author thinks the answers are.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Knowledge-Reality-Value-Mostly-Philosophy/dp/B091F5QTDS

I am not, by nature, a history buff, though as a foresight professional we need to be somewhat familiar with history as well. But nearly seventeen years into calling Australia home, Mark McKenna’s “The Shortest History of Australia” landed on my radar and I’m glad it did.

If you live in Australia, there’s a better than 50% chance that either you or one or both of your parents were born overseas. That’s pretty remarkable. For an immigrant like me, someone who chose this country and has been chosen by it in return, I feel we owe it to ourselves to understand the full picture of the place we’ve made our home, not just the brochure version or the one we engage with in our suburb, social circles, or workplace.

The Shortest History of Australia is the perfect tool for this. It will tell you, honestly, what happened, but more than that, it reframes how you see where you are. For a history book, it’s also mercifully short; noteworthy because they often tend…not to be.

McKenna covers an enormous amount of ground in a compact book, and he does it with an honesty that never tips into either jingoism or self-flagellation. The narrative begins not in Sydney Cove but in the north, with tens of thousands of years of Indigenous history; a corrective to the default European-arrival starting point that still dominates popular understanding.

From there, we get the colonial project with its Christian underpinnings (Cook’s claims were also theological, which was news to me), the shaping of national identity through war and migration, and the slow, painful, still-incomplete reckoning with what was done to First Nations peoples.

There are ugly details to be found not just among the First Nations affairs; Australia is the world’s worst country for mammal extinction. Only Australians could remember the abysmal defeat of Gallipoli as a victory. The learning curve on water and fire management - skills Indigenous Australians had refined over millennia - was, and in some ways still is, embarrassingly slow.

McKenna doesn’t flinch from the uncomfortable facts, and the book is better for it.

Unsurprisingly, the sections on migration and multiculturalism resonated deeply. McKenna captures a striking phenomenon: how migrant views of their home countries tend to fossilise at the point of departure, frozen in time while the country itself moves on. I recognise that pattern; even the language of those who migrated here half a century ago feel like live historical archives. I can only imagine how stupid I’ll sound in Finnish in another decade or two.

He also traces how Australia went from enshrining racial purity at the start of the twentieth century to embracing racial equality by its end, a transformation so dramatic it’s easy to underestimate. Whitlam’s decision to make multiculturalism official government policy in 1973 was pivotal, and the book makes a compelling case that Australia only truly began to feel like an independent country after his election in 1972. The sheer volume of progressive legislation his government passed in just three years is staggering.

Yet McKenna is equally clear-eyed about what remains unresolved. Australia has become one of the world’s most diverse, multicultural liberal democracies, but it is also one of the most property-obsessed nations on Earth, increasingly politically polarised, and, another one of those places where the belief that life will be better for future generations is fading.

One’s postcode remains one of the strongest predictors of opportunity, and levels of sexism, misogyny, and domestic violence remain stubbornly high. The suburbs, McKenna shows, are simultaneously where the miracle of multicultural coexistence is lived out daily and where inequality quietly deepens.

And then there’s the unfinished business. As Australia approaches the 250th anniversary of British arrival, it has yet to recognise First Nations Australians in a just and substantive way. An indifference remains, McKenna writes: a failure to imagine what history looks and feels like for those on the other side. While my broadly multicultural circle of friends I have here is one of the best things in my life, I don’t have many Aboriginal friends; I should. That’s my own small version of the larger pattern he describes.

This was fascinating, authentic, honest and, true to the title, short enough. McKenna manages to compress millennia into a book you can finish in a few sittings without it feeling rushed or superficial.For anyone living in Australia, whether born here or arrived yesterday, this should be required reading. It might help you understand it better, and understanding is where everything worth doing starts.

• Rating: 4.5 out of 5

• Dog-ear index: 7.3

• Who is it for: Anyone living in Australia who wants to begin to understand the place beyond the surface; especially fellow immigrants who, like me, owe it to themselves and their adopted home to know the full story.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Shortest-History-Australia-Mark-McKenna/dp/1760643599

The TL;DR version: AI now produces genuinely good foresight analysis. Not “good for an AI”, just good, period. The profession’s value is shifting from analytical work to holding the arc of an engagement: keeping people in the room, literally and figuratively, when the futures get uncomfortable, framing the right questions, facilitating the conversations that matter, and shepherding insight into action.

I’m calling this the dramaturg model. The analytical middle of foresight can increasingly be done with AI. The need for human work is stronger at the ends; the room where people confront uncomfortable truths, and the organisational reality where change actually happens. That’s where practitioners need to invest.

This shift is underway now, and it will keep moving.

I’ve spent considerable time over the past few years using generative AI for strategic foresight work, constantly trying to keep my finger on the pulse of what these tools are capable of in my domain(s). Running full engagement cycles: signal scanning, driver analysis, forecasting, scenario development, consequence mapping, cross-impact analysis, resilience assessment, strategic transition frameworks, future personas, tangible artifacts.

The kind of work that typically involves a team, a timeline measured in weeks or months, and a substantial budget.

Recently, something has shifted, and I claim that the tools have silently crossed the ‘good enough’ threshold for so many things that we foresight professionals need to recalibrate what we are here for.

I’m not alone in noticing. A 2025 survey by the OECD and World Economic Forum found that two-thirds of foresight practitioners are already using AI in their work, with most reporting significant time savings and expanded analytical capacity. The shift is underway.

In short, the output is now good. Scenarios that are genuinely distinct. Cross-impact matrices that reveal systemic patterns I hadn’t consciously assembled. Personas that are emotionally compelling. Strategic recommendations that are specific, actionable, and internally consistent across a body of work that grows more interconnected with each exercise, with a coverage and structure that feels almost suspiciously complete.

What does this mean for us?

I am a foresight professional. This is a significant part of my professional identity and livelihood. And I have watched a machine do a version of my work - not all of it, but a substantial and valuable portion - at a speed and consistency I could never match alone. That excited me, but it also scared me a little. I think the fact that it scared me is important.

(In an earlier draft of the above paragraph, I also had the words ‘not the hardest parts’ in what parts of my work AI cannot do, but I’m conscious that may have been a comforting lie, a psychological crutch to still feel special. Maybe AI is doing the hardest parts.)

If you’re a foresight practitioner reading this, you’re likely in one of several positions. You may have experimented with AI tools for foresight work already and formed a view, in which case I’d urge you to revisit that view, because the capabilities are shifting faster than most assessment cycles can track. And when you do, check yourself for motivated reasoning — are you exploring honestly, or looking for confirmation of a conclusion you’ve already reached?

Specifically: the jump from Anthropic’s Claude Opus 4.5 to Opus 4.6 has been, in my experience, a deceptively large shift, and in my experience the single most significant threshold-crossing upgrade in AI capabilities for foresight work I’ve encountered. If you tried this even six months ago and found it wanting, your conclusions have expired. Try again.

You may not have experimented yet but hold a position on whether AI is relevant to your practice. If so, I’d ask you to hold that position lightly for the length of this article. This is a domain where views need reconsidering at a cadence most of us aren’t comfortable with, because like it or not, things are changing fast.

Foresight professionals of all people should know this. How often do we apply that knowledge to ourselves?

You may also have principled objections to AI; ethical concerns about training data, labour displacement, environmental cost, concentration of power. I respect that position. Those concerns are legitimate and they deserve serious engagement. Nothing in this article argues that you’re wrong to hold them.

What I would say is that the capabilities are arriving regardless of whether any individual practitioner adopts them, and that understanding what AI can do in your domain is valuable even, perhaps especially, if you choose not to use it.

Foresight has always been about understanding even the futures you might not want.

Or you may have already composed a mental list of reasons this doesn’t apply to you on the assumption that your work is somehow exempt. If so, I want to gently suggest that the ability to construct reassuring narratives about why uncomfortable futures won’t materialise is a failure mode we’re supposed to help our clients avoid.

We’re not exempt

There is some level of professional illusory exceptionalism that can be found from many, if not most, domains. In foresight work it goes: “Yes, AI will transform accounting, law, journalism, and customer service — but our work is too creative, too human, too nuanced for machines. This is the one domain where human insight is critically important.”

This story is seductive precisely because it contains a kernel of truth.

Parts of foresight work are deeply human. But the kernel is not the whole crop.

Henrik Skaug Sætra, in a recent preprint, argues that dismissing AI as “stochastic parrots” has become a professional comfort blanket, a way to avoid engaging with how these systems are actually reshaping work.

The critique that gave the phrase stochastic parrots its power was important and necessary. But as a default stance, Sætra argues, it risks becoming a way to not see what’s happening. I think foresight has its own version of this comfort blanket: the belief that our work is too creative, too contextual, too human for AI to meaningfully contribute.

It’s a comforting story.

It’s also increasingly wrong.

Nick Foster, in Could Should Might Don’t, critiques the shiny, consequence-free variety of futures work he calls “Could” futurism. His call for “mundane futures” — more time in the messy middle where people actually live — resonates deeply.

I want to make a parallel call: mundane practice. Less grand theorising about AI’s role in foresight. More honest accounting of what happens when you sit down and do the work with it.

Let me be specific. If you write about megatrends, macro forces, or driving forces of change for a living — AI already does this at least as well as you do for most contexts, and in many cases better. It synthesises across wider source material, maintains consistency across frameworks, and knows the client context better than any random foresight professional. This isn’t capability that’s coming soon. It’s here.

The barbell of human foresight engagement

What I discovered in running full foresight cycles with GenAI is that the human contribution isn’t evenly distributed. It follows a sort of a barbell, or U-pattern: heavy human involvement at both ends, AI-dense in the middle.

The opening end is orientation and framing. This is the work of getting a room of executives to admit what they’re actually afraid of, and for all of them to gain honest situational awareness. It’s about reading the room; noticing when the CFO checked out or the CEO leaned in, and choosing which provocation to deploy based on who flinched.

Some of the most powerful foresight tools work precisely because they happen in a physical room with actual humans. An orientation exercise can transform a group’s situational awareness in ways that no document, no report, and no AI-generated scenario can replicate.

When you watch a leadership team physically map out the forces shaping their world and their views of them, it becomes collective sense-making. People read each other’s body language, pick up on hesitations, build shared understanding through the friction of disagreement. The visceral discovery when you see your leadership team doesn’dest, in fact, agree on something everyone thought was a shared truth, or the quiet executive who suddenly speaks up and reframes the entire conversation are moments that don’t happen in a chat interface.

Physical presence still has power, and rooms can be efficient vehicles for mindset shifts. I’ve watched a team walk in thinking they were there to discuss a five-year technology plan and walk out realising they had a workforce crisis they hadn’t named. That shift is possible because someone was in the room facilitating, probing, creating the conditions for honest conversation.

AI can generate the analytical inputs that make those conversations richer. It cannot be the person standing at the whiteboard when the room goes quiet and someone finally says the quiet thing out loud.

Will AI-mediated facilitation eventually enter this space? Almost certainly. Immersive simulation environments are already a thing. Robots are coming. But I doubt we’ll be comfortable having machines run deeply human exercises like collective orientation and values negotiation for some time.

The technology may arrive before the trust does, and in facilitation, trust is the technology.

The middle is analysis, synthesis, and generation. Signals through scenarios through consequences through strategy.

From an effort perspective, this used to be the bulk of many foresight projects — and this is where AI excels. GenAI models can now maintain coherence across a massive body of interconnected analysis, a dozen-plus exercises in consistent relationship with each other.

The quality I could match, potentially even exceed, given enough time. What I couldn’t match was the quality at that volume and speed, sustained across every exercise in the cycle, with a consistency no team could replicate in the same timeframe.

And also, when is there ever ‘enough time’?

The closing end is action and implementation, where things get translated back into the physical reality of the organization. Building roadmaps. Catalysing change.

All this requires what the Greeks called metis — practical wisdom, the knowledge of how things actually work in this specific organisation with these specific humans and their specific histories, grudges, and ambitions. Who needs to be in the room to make this decision stick? Who will quietly sabotage it if they weren’t consulted? Which board member needs to feel like this was their idea? That knowledge lives in relationships and corridor conversations and decades of institutional memory.

No AI has access to all that, and providing it with enough context to be genuinely helpful here is itself a substantial act of judgment.

An important caveat: I don’t think the barbell is a permanent shape. It’s the shape of the work right now. The middle is already highly AI-capable; the ends are still more firmly human. But the boundary will keep moving. AI will get better at facilitating structured conversations. It will gain access to more organisational context. The human ends of the barbell will compress over time — not disappear, but narrow. Recognising that the barbell is a snapshot, not an equilibrium, is itself a foresight act. Plan for today’s shape; don’t assume it’s tomorrow’s.

Points of disagreement

I’m not the first or the only one thinking about the changing nature of foresight work, quite obviously. The growing body of work on AI in foresight is encouraging and, I think, incomplete in some important ways.

A recent case study from Queensland University of Technology, working with the Queensland Government, documented a 60-70% reduction in staff hours through AI-augmented scenario development. They doubled their scenario output. They completed analysis in weeks rather than months. These are real gains and they match my experience. Where I depart from the efficiency framing is this: the most important thing wasn’t going faster. It was going deeper. More exercises than time constraints typically allow. Branches I wouldn’t have explored. Connections I wouldn’t have seen.

Speed is the obvious benefit. Depth is the transformative one.

The OECD/WEF survey identifies three maturity levels for AI in foresight: basic analysis augmentation, creative sparring partner, and fully integrated systems. What strikes me is that this framework describes integration depth but not role transformation. You can be at the highest maturity level and still think of yourself as an analyst who uses AI tools. That’s not the shift I’m describing.

What I’m describing is something more fundamental: a shift in what the foresight professional is.

There’s also a strand of literature that frames AI’s contribution in expansive, almost breathless terms like “analytical supremacy,” “epistemological pluralism,” AI that dynamically models complex interdependencies and simulates emergent cascading effects. Some of this is real. But these accounts describe AI as upgrading everything more or less uniformly. My experience says the opposite. The upgrade is profoundly uneven; the jagged edge is real even in foresight work. The analytical middle leaps forward. The human ends remain stubbornly, beautifully resistant.

That unevenness is the most important finding, and acknowledging that it too will shift over time is part of honest practice.

A finding from the QUT study deserves particular attention: AI hallucinations occasionally served as creative provocations in scenario work. I’ve seen this too; in divergent ideation, we like telling people there are no bad ideas, but do we extend the same charitable principle to our AI tools? This only works if the practitioner has enough independent expertise to distinguish a genuinely interesting provocation from confident nonsense.

Sætra draws on Stephen Barley’s sociology of work to make a distinction that matters here: between substitutional and infrastructural technological change. Substitution is when a new tool replaces an old one in roughly the same role.

Infrastructure rewires how the work is organised.

Most commentary about AI in foresight treats it as substitutional; a faster way to do scenario analysis, a better research assistant. My experience suggests it’s infrastructural. It doesn’t just speed up the middle of the barbell; it changes the relationship between the parts. When the analytical work that used to take weeks happens in hours, everything upstream and downstream reorganises around that new reality: client expectations shift, engagement design changes, the skills that matter are different, the bottleneck shifts.

Sætra identifies a historical pattern that should concern us: when a bottleneck moves from human to machine, the work doesn’t vanish, but the power embedded in that work does. The typesetter didn’t disappear overnight when the Linotype arrived. But the bottleneck moved, and with it went the leverage.

If foresight’s analytical middle follows this pattern — and I think it’s beginning to — then the question isn’t whether the work gets done. It’s who controls how it gets done, and what happens to the practitioners whose expertise was built on doing it. (Only half-facetiously I would suggest one place the bottleneck shifts to is calendar-wrangling of the stakeholders)

This connects to the deskilling concern that bothers me more generally.

If AI handles the analytical middle, where do junior practitioners develop the pattern-recognition skills that make senior expertise valuable? It’s the same tension aviation discovered decades ago — hence the name of this newsletter — when autopilot made flying safer but slowly eroded the manual skills pilots needed when the automation failed.

I don’t have a clean answer. But naming it honestly is more useful than pretending it doesn’t exist.

And I think the honest next question is: could we stop this even if we wanted to?

Almost certainly not. Foresight has always been a tough sell. It’s been seen as expensive, slow, its value diffuse and hard to measure. No client is going to voluntarily pay for the slow human-only version of analytical work to preserve our apprenticeship pipeline.

That’s asking the market to subsidise practitioner skill development out of kindness. Markets don’t do that.

There’s a deeper reason too. The acceleration isn’t just a threat to practitioner skills; there’s an argument to be made that it’s a genuine necessity.

A year-long foresight project on AI-adjacent topics is already obsolete by the time it ships. The world, in at least some domains, is moving faster than traditional foresight cycles can track. The profession had a speed problem before AI arrived as a tool, and now the same technology causing the acceleration is also the only plausible solution to it. If we can’t keep pace with the rate of change, our work is just wrong.

And there’s a genuine good buried in the disruption. Foresight has been a luxury good, available primarily to large organisations with large budgets. If AI makes meaningful futures work accessible to a small nonprofit, a local government, or a startup that could never have afforded a team engagement, that’s not a downside with a silver lining.

The counterweight is that sense-making has a biological clock. You can generate a cross-impact matrix in minutes, but the leadership team still needs time to sit with what it reveals. The uncomfortable finding that your core business model is a first-curve asset heading for decline needs to marinate.

The conversation where someone finally names the thing everyone was avoiding can’t be scheduled for 2:15 PM between the scenario generation and the action planning. Human processing time isn’t a bottleneck to be optimised. It’s where the value is actually created. Speed the analysis, yes. But don’t mistake faster generation for faster understanding.

The Dramaturg

All of this points to a specific evolution in the foresight professional’s role. I’ve been reaching for the right word for it. Conductor, curator, orchestrator all come to mind, but I landed on dramaturg as the best option. The role has elements of all the three before-mentioned roles, but dramaturg captures something more.

In theatre, the dramaturg holds the arc of the production. They don’t write every scene, but they know which scene needs to land. They don’t perform, but they understand what the audience needs to feel and when. They’re the bridge between raw creative material and its impact on the people who experience it.

That’s what I was doing in my AI-assisted foresight work. Making curatorial decisions constantly. Which exercise next? Do we have enough pre-work for this one? Is the body of divergent work sufficient to make the convergent exercise meaningful? Those decisions drew on years of practice; knowing what good foresight output feels like, and knowing when something’s missing before I could fully articulate why.

The dramaturg knows that the cross-impact matrix is analytically elegant but the persona of a woman managing her husband’s heart failure from a trailer in rural Mississippi is what will actually change behaviour in the boardroom. They design the emotional journey from comfortable assumptions through productive discomfort to committed action.

That design sense doesn’t quite automate. Yet.

This is just a snapshot

I’m wary of making confident predictions with specific timelines, that would be the kind of shiny futurism this piece is arguing against. But I can describe the direction.

The analytical middle of foresight is moving rapidly toward AI competence. Signal scanning, driver analysis, scenario generation, consequence mapping, cross-impact analysis, these are already being done well, and the capabilities are improving visibly.

Two-thirds of practitioners already use AI for this work, and even within cautious government environments, the productivity gains are too significant to ignore.

The creative and communicative layer is shifting next. Personas, artifacts from the future, the narratives that make abstract foresight visceral, these were among the strongest outputs in my work. As more practitioners discover this, the portion of an engagement that requires direct human generation will shrink further.

What will remain human longest is the relational work at both ends: the initial framing and the closing action, because these depend on physical presence, institutional knowledge, and trust between people who share stakes.

The profession may bifurcate. Practitioners who can articulate what they add beyond analysis, the dramaturgs, will find their work more valued, not less. Practitioners whose value proposition was primarily analytical rigour will face increasing competition from clients who can do that themselves.

This is the same structural shift that has reshaped graphic design, copywriting, coding, and data analysis. Foresight isn’t exempt.

I don’t know exactly how far this goes. I’ve run full engagement cycles in single sessions that produced output that was genuinely suitable for real client work. That would have been inconceivable two years ago.

What’s inconceivable today may be routine in another two.

Foresight professionals, of all people, should know better than to assume the present pace is the permanent pace.

So what?

Stop treating AI as either saviour or threat. Start treating it as a tool you need to learn to use well, the way you learned to use scenario frameworks and driver analysis and futures wheels and cones.

Even if your ethical stance is not to use any AI tooling, understanding what it can do in your domain is prudent due diligence.

Run the tools with it. Notice where it surprises you. Notice where it falls flat. Build your own sense of where the barbell sits for your practice, your clients, your domains.

Invest in the parts of your practice that live at the ends: reading rooms, framing questions, designing emotional arcs, shepherding action through institutional resistance.

Get better at being a dramaturg.

Engage with the discomfort rather than narrating around it. If you’re a foresight professional and the idea of AI doing your analytical work well doesn’t make you at least a little uneasy, you either haven’t tried it or you’re not being honest with yourself.

That unease is a weak signal. Use it the way you’d use any other weak signal: as evidence that something is changing and your current model may need updating.

It hits different when the weak signal is about your work than the client’s business, doesn’t it?

Get used to it. You if anyone have the tools to deal with that discomfort.

And acknowledge that even the barbell model I’ve described here is a snapshot, not a destination. The human ends may narrow. The middle may expand. New capabilities may emerge that don’t fit neatly into any current framework.

The shape of foresight practice in five years will surprise us, and our ability to be surprised by it, and adapt, is probably the most durable professional skill we have.

There’s a question I haven’t addressed here: what happens when clients figure out this capability shift? When they arrive having already run the scenarios, already generated the personas, and they say they only need you for two days in the room instead of two months of engagement? That’s a different piece, but if the barbell model is right, it’s not the threat it first appears to be. More on that later.

None of us trained for this version of the profession. But then, nobody trains for the exact futures they get.

That’s always been the point.

nb. I am using the terms AI and GenAI somewhat interchangeably here for brevity. I am quite aware of the wildly different flavors that get clumped under the AI umbrella, but this isn’t the essay to sort that out.

PS. A necessary caveat: the results I describe didn’t come from lazy prompting. They came from sustained, systematic work, including carefully constructed Claude Skills, lengthy custom instructions, and iterative refinement. The tools are capable of genuinely good foresight work, but they don’t produce it by default. How you use them matters enormously.

Nick Foster’s Could Should Might Don’t is published by Particular Books (Penguin). My review is here.

Henrik Skaug Sætra’s “The Tyranny of the Stochastic Parrot: How AI Critique Became a Way to Not See What’s Happening” is a 2026 preprint from the University of Oslo. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6249318

World Economic Forum/OECD (2025). AI in Strategic Foresight: Reshaping Anticipatory Governance. https://www.oecd.org/en/publications/ai-in-strategic-foresight_aa573076-en.html

Picavet, E. et al. (2025). “Human–Machine Collaboration for Strategy Foresight: The Case of Generative AI.” Public Administration Review. Queensland University of Technology. https://eprints.qut.edu.au/261336/

Soeiro de Carvalho, P. (2024). “How GenAI Will Transform Strategic Foresight.” IF Insight & Foresight / Hong Kong Institute of Futurologists & Foresight Analysts. https://hkifoa.com/wp-content/uploads/2024/12/how-genai-transform-strategic-foresight.pdf

Every organisation you’ve ever worked for has tried to simplify you: your performance has been reduced to a number; your skills to a list of keywords that HR has built company-wide profiles on (usually failing to accomplish anything meaningful in the process); your potential to a box on a nine-grid; and so on.

If you’ve ever felt that the version of you that exists in a system bears only a passing resemblance to the actual you, if that, James C. Scott has 445 pages explaining why — and what happens when that impulse to simplify is backed by the full coercive power of a state.

“Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed“ is an ambitious, dense, and frequently brilliant examination of what happens when states try to make the messy complexity of real life legible - reducible to categories they can then measure, manage, and control. First published in 1998, it remains remarkably relevant, even as some of its examples have aged a bit.

Scott opens with an example that seems deceptively simple: forestry. When eighteenth-century German states looked at a forest, they didn’t see an ecosystem. They saw - surprise! - revenue. Specifically, they saw how much timber could be extracted annually.

Everything else about the forest - the berries, the fungi, the medicinal plants, the grazing, the habitat, the spiritual significance - was invisible to the fiscal lens. The forest-as-habitat disappeared and was replaced by the forest-as-economic-resource.

This, Scott argues, is what states do to everything they touch: they strip away the particular, the local, and the contextual in favour of standardised categories that can be counted, taxed, and administered.

From forests, Scott expands to measurements (did you know that in pre-modern Malaysia, the answer to “how far is the next village?” might be “three rice-cookings”? - a unit that conveys what the traveller actually needs to know far better than any number of kilometres), to surnames (which are a surprisingly recent invention, imposed by states for the purpose of making populations legible), to language standardisation, to the redesign of cities, and eventually to the grand disasters of high modernism: Soviet collectivisation, compulsory villagisation in Tanzania, and the architectural hubris of Le Corbusier’s planned cities.

The book’s argument rests on four elements that, Scott contends, must combine to produce truly tragic state-initiated social engineering: administrative ordering of nature and society; a high-modernist ideology with uncritical confidence in science and technology; an authoritarian state willing to use coercive power; and a prostrate civil society that lacks the capacity to resist.

It’s hard to read this list in 2026 without a chill of recognition, except that today’s high modernists have replaced uncritical faith in science with uncritical faith in a blinkered, biased, and demonstrably false worldview.

I found the book to be at its best when discussing cities, drawing heavily on the work of Jane Jacobs. The contrast between Le Corbusier’s sterile modernist vision and Jacobs’ celebration of messy, organic urban life is one of the book’s most compelling sections, but YMMV of course.

The concept of being on “sidewalk terms” with people in your neighbourhood — not friends, but acquaintances who recognise one another and who collectively form an intricate web of informal social order — resonated deeply with me, and also ties into points made in many other books, such as Joe Keohane’s The Power of Strangers. Jacobs saw that the busiest room in a house is the kitchen, and the busiest street in a neighbourhood functions for the same reason: it is the most versatile setting, a place of socialisation and exchange. Understanding this, Scott argues, is no more difficult than understanding why the kitchen draws people; and yet urban planners have spent a century trying to engineer it away.

The analogy Scott draws between city development and language development is particularly elegant. Both are the unplanned creation of millions of people over time. Both resist central planning; Esperanto and Brasília share the same fundamental problem.

And both, when allowed to develop organically, tend towards a rich, multivalent complexity that planned alternatives simply cannot match.

The book’s final chapters, on what Scott calls métis (a Greek term denoting practical knowledge that can only come from experience) are where I found the deepest current value.

Métis is the firefighter who reads a situation; the sea captain who senses a change in weather through the roll of the ship; the farmer whose knowledge of this particular piece of land cannot be reduced to generic agricultural principles. It is knowledge that resists codification precisely because the environments in which it operates are too complex and variable for formal procedures.

You can master every principle and still fail at the craft, because knowing how and when to apply rules in a concrete situation is métis - and it is precisely what our current AI models lack. They can ingest every principle ever written down; what they cannot do is develop the practitioner’s feel for when the principles don’t apply.

The concept has stayed with me. In fact, it became one of the triggers for a recent essay I wrote on what I call “the grace margin“ - the space between what a system prescribes and what a person actually does, where someone can look at the rules, look at the situation, and choose judgment over procedure.

The grace margin lives in métis; in the practical, informal knowledge that no system can fully capture and no algorithm can replicate. Scott showed me where that concept has its roots, and why its erosion matters far more than most people realise.

The observation that formal order is always and to some considerable degree parasitic on informal processes is one of those insights that, once seen, cannot be unseen. Scott illustrates this brilliantly with the work-to-rule strike, aka malicious compliance: when workers follow every rule and procedure to the letter, production grinds to a halt, because actual work depends on a host of informal practices and improvisations that could never be codified.

The formal scheme cannot create or maintain these processes; it can only exist because of them.

I do have some reservations. The book is not a quick read. The concepts are heavy, the language is heavy, and Scott’s academic thoroughness sometimes comes at the expense of pace. Some readers will find the extended case studies, particularly the chapters on Soviet collectivisation and Tanzanian villagisation, more exhaustive and potentially exhausting than illuminating, though I appreciated having a case study I’d never encountered before in the Tanzania material.

More significantly, there is a datedness issue that becomes harder to ignore as the book progresses. Scott’s framework is built almost entirely from twentieth-century examples, and some of the most significant mechanisms of legibility and control in our current era — social media, the internet, algorithmic governance, platform economies — simply don’t register.

In many ways, what has been done in agriculture alone in the past twenty-five years represents another layer of extraordinary change, much of it not for the better, but some also for the better (e.g. permaculture, regenerative agriculture).

The core arguments remain powerful, but the book would benefit from a contemporary companion that extends Scott’s framework into the digital age, where legibility is no longer imposed by census-takers and cartographers, but extracted by algorithms processing our every click, purchase, and movement.

The state still sees; but so, now, does everyone else with sufficient data and computing power, and potentially in much higher detail than the state does.

Despite this, “Seeing Like a State“ is a book whose central insights have only become more urgent. The impulse to simplify, to make legible, to eliminate the variance where human judgment lives — this impulse hasn’t weakened since 1998. If anything, it has been turbocharged, because we can now have so many data points it feels like we could finally quantify everything and make everything legible. I’m highly skeptical of that view.

Scott’s warning that high-modernist designs tend to diminish the skills, agility, initiative, and morale of their intended beneficiaries feels less like historical analysis and more like a forecast of what algorithmic management is doing to workers right now.

And his closing plea for métis-friendly institutions — organisations that are multifunctional, plastic, diverse, and adaptable — feels almost radical in an era that worships efficiency and consistency above all else.

We need more flexibility and we are, as Scott feared, designing it away.

• Rating: 4 out of 5

• Dog-ear index: 6.6

• Who is it for: Anyone who has felt ground down by bureaucracy and wanted to understand why; people interested in urban planning, governance, or the tension between efficiency and humanity; and anyone building or deploying systems — algorithmic or otherwise — who wants to understand what gets lost when you optimise for legibility.

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Seeing-Like-State-Certain-Condition/dp/0300246757

This review comes with a disclosure: foresight is a significant chunk of my professional identity. Reading a book that critiques how we think about the future was simultaneously validating and uncomfortable - like having someone narrate your therapy session while occasionally getting the details wrong, prompting the occasional “Hey wait that’s not how it went down!”

Nick Foster’s Could Should Might Don’t gives us, in its title, a simple taxonomy for understanding the different flavors of futures work.

“Could“ futurism is the shiny, techno-optimistic variety that dominates TED stages, conferences, and often corporate innovation labs. “Should“ futurism brings in values and ideology. “Might“ is a more rigorous strategic foresight territory of scenarios and probabilities, but with its own problems. And “Don’t“ - the hardest and usually, at least from an action-point of view the most neglected, branch - forces us to confront futures we must actively prevent.

The framework is, to my moderate surprise, genuinely useful, and Foster wields it effectively against the futures-industrial complex. Many of his views align with mine, which means arguably provocative descriptions like trade shows as “cacophonous orgies of energetic capitalism“ made me chuckle at the accuracy rather than take offense. Foster’s takedown of commercial Could futurism - all gleaming surfaces and conveniently absent consequences – also clearly articulates a lot of the frustrations I’ve felt for years.

The observation that “nothing on earth works the way they pretend“ in these polished visions cuts to the heart of why so much futures work fail to connect with reality – and, frankly, why so many products arising out of that style of futures work fails miserably when it comes to contact with reality.

Foster makes several important points. A few examples include that newness as an inherently positive attribute is a modernist mindset shift we rarely question; that we have an insatiable and largely unwarranted desire for predictions; that identifying trends is far more questionable than trend-spotters would have you believe; and that the most transformative technologies eventually become mundane, embedded invisibly into everyday life. The concept of “Endineering” - thinking about endings and full lifecycles rather than just shiny beginnings - deserves wider adoption.

The book’s call for “mundane futures” also resonates deeply. Most of us will live in the statistically swollen middle, not in gleaming smart cities or dystopian wastelands. So much of futures work either fundamentally misunderstands or willfully ignores such basics as how cities come into being in the first place. The characters in our future, as Foster notes, “will not necessarily need to save the world at every turn - most of them will simply live in it, quietly getting through each day.”

This is where futures work should spend more time, yet rarely does.

And yet.

Reading as an insider of sorts, I found myself in a recurring “yes, but...“ pattern. The critique Foster lays out sometimes feels like it hasn’t fully engaged with the better work being done in the field - the rigorous scenario work at places like IFTF, for instance – curiously, IFTF is not so much as mentioned even though RAND Corporation, its parent organization, is (full disclosure: I am a Senior Research Affiliate at IFTF).

While Nick to his credit fully acknowledges his own biases and blind spots, occasionally they raise their heads in a somewhat ironic manner: his anecdote about finding time for reflection on long flights immediately after discussing how futures thinking skews toward the privileged is a perfect, probably unintentional, demonstration.

More frustratingly, I finished the book asking “now what?“ The diagnosis is comprehensive and often brilliant, but the prescription feels thin.

We need to do better, demand better, be more critical consumers of futures work - yes, agreed, violently so. But for practitioners seeking concrete guidance on how… There’s not all that much there.

Which, to be honest, is not what the title or subtitle promised either. Offering a prescription was never part of the plan, probably. I get that. It’s still something I as a practitioner would want effective ideas or options for.

My perspective is skewed by proximity. For readers who are not foresight professionals - which is most people - this book offers something very valuable. Executives, policymakers, and people who consume futures work (this is pretty much everyone, if you pay attention to marketing) would benefit enormously from Foster’s framework for identifying when they’re being sold seductive nonsense.

The billions wasted on manifestly bonkers projects like The Line - that mirage monument to Could futurism that treats the desert as a blank slate for techno-utopian fantasy and, let’s face it, is never going to get built - suggest we desperately need more critical consumers of futures visions. This book helps create them.

For those of us already living in Might and Don’t territories, “Could Should Might Don’t” offers the bittersweet comfort of having our professional frustrations eloquently articulated - along with the sobering reminder that this more rigorous approach is harder to sell. It’s validation, but not transformation.

• Rating: 4 out of 5 (closer to 4.5 for non-foresight readers)

• Dog-ear index: 11.8

• Who is it for: Anyone who consumes futures work and wants to become a more critical reader of it; executives and policymakers who’ve been dazzled by shiny visions; foresight professionals seeking articulate validation of their frustrations (though perhaps not new tools to address them).
  
  Oh, and the Could-futurists who need a good hard look in the mirror!

[reminder: I highlight important parts of the books I read, and dog-ear the really important pages. The dog-ear index is simply the average number of dog-eared pages per 100 pages]

Product link for reference only; please support your local bookstore where possible: https://www.amazon.com.au/Could-Should-Might-Dont-Future/dp/1837263833