Writing about AI feels like writing about oxygen. Or Taylor Swift. The sujet du jour, omnipresent, necessary, transformative, boring. Oxygen is the perfect analogy, actually: AI sucks all the oxygen out of the room, and it feels like nothing else is worth writing about.

If a large object were on a collision course with Earth, we’d write about it endlessly too: is it an asteroid or aliens? Are they friendly? Did Mary Magdalene see anyone buried on it?

The AI asteroid is forcing us to come to grips with the nature of ourselves. We are biological machines. We have certain physical and intellectual capabilities. And we’re building other machines, wildly different in their internal function, yet capable of replicating and exceeding human capabilities in some areas.

But these machines are not Humanity 2.0. They are not a pure upgrade across all dimensions. If two machines differ wildly in their internal structure, it follows that their outputs will differ wildly at least some of the time. The fact that we can get the output to match over some domain of inputs is unremarkable. A fan can blow air, and so can a human (quite a lot of hot air in my case.) A monkey can eat a banana, and so can a human. Claude Code can write a C++ function, and so can a human.

In the limit, the mere existence of a human’s internal structure is itself an output — in the mind of an observer — and it’s one the AI will never be able to replicate. And if it did, it would cease to be AI and start being human. Put that thing out the airlock.

One could argue internal structure matters to us far more than output. Output is commoditized. We value the burial site of a person who once was more than a living person who looks the same, or makes the same jokes. We value an original painting, even badly damaged, more than a technically superior copy of it.

As copies go, we are often more impressed by attempts to replicate not just the output but the internal state. Daniel Day-Lewis is arguably famous because of how he produced the output he did — by carefully and obsessively curating his internal state to match the human he intended to copy — as much as for the output itself.

We assign an almost supernatural quality to internal structure as it relates to human beings. Even though every molecule that constituted her was ordinary, that specific configuration was unique. And if we could somehow reconstruct her exactly, it still wouldn’t be her. Humans extend this supernatural quality to objects too: The Shroud of Turin, Hemingway’s typewriter, Anne Frank’s diary, your ex’s t-shirt.

The supernatural quality is really an experience in the mind of the observer. There is nothing objectively special about these people and their objects except the degree to which other human minds assign value to them. But why do we care about The Shroud of Turin and not my bedsheets, about her and not that person just like her.

Modern marketing already knows what we value: one of a kind, limited supply, never before seen, beachfront, exclusive, rare opportunity.

We value scarcity. Uniqueness is the ultimate scarcity. And humans and their creations are the ultimate uniqueness. You cannot replicate a human being. We’re one of a kind — except twins, and we have a certain fascination with them. And even our machine creations are relatively simple and easy to replicate compared to us. Every copy of a machine is identical for all practical purposes.

AIs operating in a marketplace shared with humans will learn to value scarcity too. So in a world of abundant AI output, human provenance will still be valued, perhaps above all else.

People are already in the habit of assigning human provenance to their work. I wrote this essay by hand, I’ll tell you. Would it matter if an AI wrote it? Would the jokes land differently? Would the evidence stop supporting the conclusions? The output would be identical. But it would lack that supernatural quality.

This is the bedrock differentiator of humanity. It’s the stuff we’re made of, literally. The dystopians on Reddit will tell you that your skills will become value-less. But what if the thing each of us already has, our humanity, will become the only thing of value that’s left?

That’s the kind of cringe-inducing conclusion only an AI could write.

It all started in 1992. Nine-year-old me saw the Mulroney government order replacements for Canada’s aging fleet of CH-124 Sea King helicopters. These national embarrassments needed 30 hours of maintenance for each hour of flight. And the original fleet of 41 had been reduced to as few as 28 through attrition.

So ordering a replacement fleet seemed like a perfectly reasonable thing to do. Some would say the only thing to do.

And then a French Canadian man with a funny accent running for Prime Minister appeared on TV and said “no new helicoptahrz!” And when he won the election, he cancelled the contract as promised. And then his government proceeded to ignore the situation for a decade, thus setting off a thirty year delay in the acquisition of new choppers. (And a 30 year grudge I’ve held against the Liberal Party.)

So that gives you a glimpse into the state of Canada’s military some 33 years ago. Since then, humanity has invented the Internet, the smartphone, reusable rockets, self-driving cars, and we’re on the cusp of autonomous robots at scale. In that time, Canada has modernized its military, creating a force that punches above its weight for a middle power: a three ocean navy, a modern air force, a small but capable army.

Hahahahahahaha, I’m just kidding. We bought four broken subs built in the 80s that the British had already decommissioned. HMCS Chicoutimi caught fire on its delivery voyage across the Atlantic in 2004, killing one sailor. At times, Canada has had zero of the four submarines available to defend the longest coastline in the world.

We retired our four Iroquois-class destroyers. These regularly served as NATO flagships in times past. Without them, the navy is deprived of fleet-level command and control and area air defense. Our remaining frigates have to hide under the umbrella provided by American destroyers.

We tried to replace our FA-18 jets from the 80s. The Harper government selected the F-35 as the replacement in 2010, costs ballooned, there was a scandal, the process was reset, Trudeau promised an “open competition” in 2015, then spent years exploring a stopgap purchase of used Australian F/A-18s, and eventually committed to the F-35 anyway. Except now we’re reconsidering the F-35 purchase because of Trump. Zero jets have been delivered. Sweden is trying to get us to buy the Gripen. The US says if we do, it may be the end of NORAD.

We have like 80 mostly non-operational tanks from… you guessed it… the 80s! We gave 8 to Ukraine. In the photos I saw of their delivery, let’s just say they didn’t look factory fresh. The official diplomatic response from Ukraine was “gee, thanks.”

We have enough ammunition stored to fight an artillery war for perhaps a week, and enough domestic capacity to produce in a month the number of shells Ukraine would need in a single day. Oh, and we only have 33 artillery guns anyway. And guess what, they need to be replaced too.

We also have a shortage of tens of thousands of military personnel, including technicians who require months or years of training. So even where we do have new hardware, like our six brand new arctic patrol vessels, we can only keep one of them operational due to crew shortages.

Not to worry, though. Our arctic patrol vessels can engage the enemy with a tiny mounted gun at a maximum range of a few kilometers. Its Russian counterpart of similar size can engage over the horizon with Kalibr cruise missiles at a range of 1,500 km or more. So it doesn’t really matter whether we field one, six, or one hundred of these bad boys.

But for the first time in generations, there is hope. Prime Minister Carney has committed to spending a staggering 5% of GDP on the military by 2035. We’ve raised salaries. We’re hiring. We have created a new agency that can cut through procurement red tape for large purchases. And consequently, we are in various phases of acquiring a fleet of 15 destroyers, 12 submarines, and 88 jets, as well as new support ships, early warning aircraft, tactical helicopters, self-propelled artillery, and domestic ammunition production.

This would give our navy serious warships with advanced sensors, communications, missiles, and torpedoes. By 2040, if things go as planned, Canada will truly be a sea king, fielding a navy that ranks in the top 10 globally. Nine-year-old me has been waiting 33 years. And maybe this time, it will actually happen.

But for now, we’re more of a sea tuna with our arctic patrol ships and their cute little fins… I mean guns.

On the first day, I was asked to take my shoes off in a cafe. I declined and sheepishly kept wearing them in the single room where it was allowed. But I nonetheless became a regular. And now shoes-on cafes feel alien. Who wears shoes in a cafe?

One night, I heard strange music coming from the main room. I can’t really tell you what it was, so I’ll just call it hippie music. A guitar and some rhythmic chanting was involved. And I felt absolutely compelled, pulled, summoned to dance to it.

Me, dancing alone, with strangers, to unknown music? That’s an analytical nightmare under normal circumstances: am I doing it right? Are people watching me? What is the point of this, anyway? In Bali, none of these thoughts occurred to me, or if they did, I paid no attention.

Not everything here is purely spontaneous. Daily yoga has been a staple. It’s almost undeniably good for you — body and mind — when done consistently. Yet I recoil when I hear dubious concepts like “energy” and “detox” spouted as truths, and unlike worries about shoes, I doubt this concern will fade.

Everything in Bali is like this to a degree: a good premise that doesn’t know when to stop. When I joined a two hour “authentic relating” workshop, I thought the teacher would be our guru leader, slowly indoctrinating us into his cult. Instead, we were led by the most normal and down-to-earth person imaginable. It was we the participants who went too far, veering into trauma dumping, vulnerability olympics, uncalibrated oversharing, and speaking without also listening.

It’s these moments of excess that define this place:

• Eating healthy becomes macaroni and cheese with no macaroni or cheese.

• Spontaneity becomes a feeling of offense when asked what you’re doing next month.

• Lived experience becomes global truth.

• Warmth becomes sunburn.

• Rain becomes flood.

• Two weeks in Bali becomes planning the rest of your life in Bali.

While Bali gives you the disease of excess, the philosophy of yoga also gives you the cure. The body must be nourished but so must the mind. Plans may give way to novelty, but sometimes they must hold steady against chaos.

The fact that I wrote any of the preceding nonsense is evidence I’ve been spending entirely too much time at yoga. Or perhaps it’s evidence that despite my analytical reservations, I have actually learned something here — something difficult to quantify, something not easily published or peer-reviewed. Perhaps spiritual excess is the means through which such truths are found. You travel to the outer limits of experience, and then retreat back to your personal Overton Window, which itself has now shifted.

Contact with the outside world has become mildly strained. I suggested to a friend in venture capital that they evaluate deal flow more based on “vibes,” and I’m not entirely sure I was joking. They seemed tangled in an analytical web. Ordinarily I would jump into the web and enjoy the process of weaving it. Now the web seemed like a trap that was consuming them, not a juicy intellectual problem I was being invited to solve.

The world of analysis awaits me when I leave next month. There are spreadsheets to update, capital allocation decisions to make, houses to sell, and taxes to pay. But the importance of making the numbers go up now seems muted.

My yoga teacher says the last breath we take before we die should be a happy one. That the river of you, with the ego that thinks he’s the biggest river, will finally meet the ocean and realize how small he is. And maybe the lesson here is to realize your size early in life, and find a sense of peace and unity in that.

Or maybe it’s to make an absolute boatload of money. I can’t be sure.

• Apathy: the future is not worth thinking about

• Annihilation: intelligent machines will end civilization

• Abundance: robotics and AI will meet all of your needs

These are dangerous beliefs for the modern human. Not only are they unlikely to come true, but subscribing to any of them will cause you to make terrible plans—or no plans at all.

Apathy

The apathetic fundamentally believe the future doesn’t require preparation. Emblematic of this camp is the elementary school teacher, early in her career, who told me her job was safe from artificial intelligence.

Beliefs like this are common because radical change is so difficult to imagine. Our brains didn’t evolve to simulate the consequences of rapid technological progress. Besides, change is scary; superintelligent machines doubly so.

And in recent memory, apathy was a successful strategy. We have been living in the shallow end of the exponential disruption curve, with slow-moving changes giving us ample time to adapt. But that was the tutorial level; the boss fight begins when the curve turns steep.

This time, those who fail to plan for swift disruption may watch the pillars of their lives—including careers, skills, assets, and relationships—crumble beneath them. Apathy will lead to a form of annihilation—not the kind where the robots kill you, but the kind where your life stops making sense.

But there’s another kind of annihilation, and in this version, the robots really do kill you.

Annihilation

The annihilation camp hasn’t quite decided how we will be annihilated. Perhaps we will be ruled by immortal tech elites or perhaps we will simply be turned into paperclips.

And who dares dissent? If you believe that a global networked self-improving unstoppable machine superintelligence is going to destroy us all, there really is nothing you can do. You swallowed the narrative; now let its poison dissolve your executive function.

If apathy is I don’t need plans, then annihilation is my plans don’t matter. Annihilation breeds apathy, just as apathy leads to annihilation.

There is an evolutionary adaptation at play here: the desire not to spend one’s finite energy on pointless pursuits. If you know the end is near, struggling will only waste your time and amplify your suffering. Don’t play a game you can’t win.

But even a low probability of survival calls for proper planning. After all, disasters always lurk on the horizon. Anything from a heart attack to an asteroid could foil your plans. But that doesn’t mean you should stop making them. If you might live, you ought to plan as if you will.

If we can get annihilation from a certainty down to a probability, then we can quickly restore the paramount importance of planning and agency. And it seems an easy task to chip away at the absolute certainty of just about anything. Could I convince you there’s a 1% chance we avoid annihilation?

Consider that even a malevolent machine intelligence will face opposition. AI safety, reciprocal adaptation, countermeasures, and co-evolution will prevent us from proceeding monotonically to doom. Annihilation is not certain.

Abundance

And now we close the triad with the most optimistic of the three beliefs: abundance. The post-scarcity prophecy in which work becomes obsolete. In this vision, we flourish under a warm blanket of benevolent, pervasive technology. Most goods and services are readily available and ultra-low-cost.

When you hear someone speak of the inevitability of universal basic income or the end of money as a concept, you know they believe in abundance.

The emotional lure here is obvious: all you have to do is wait, and eventually time itself will deliver a cheat code for your life. Who among us could resist such a tempting offer? But by listening to the siren’s call of abundance, we once again sabotage our agency.

We may very well have great abundance in the future. The problem with abundance is that it can at best half-deliver on its promise. Universal post-scarcity is impossible because abundance doesn’t eliminate competition; it shifts it.

Some things will still be scarce, and it is precisely those things that will be in especially high demand. When machine companions are pervasive, true friendship becomes gold. When virtual vacations are free, seeing the ocean in person becomes extraordinary. When AI slop is everywhere, original ideas become priceless.

Abundance breeds a new kind of scarcity. Let’s examine its nature.

The List

The notion of a post-scarcity society always carries an asterisk that reads like a warning label: doesn’t apply to status goods. This is like a label on soap that says can’t be used for cleaning.

Status is a fundamental human need. “People’s subjective well-being, self-esteem, and mental and physical health appear to depend on the level of status they are accorded.” (Anderson et al., 2015)

People strive to appear socially valuable, to command respect, and to wield influence. Status goods aren’t some esoteric afterthought; they’re a fundamental driver of human behavior. And status goods are, by definition, scarce.

When abundance exists in a category of goods, those goods lose their ability to confer status. If we believe in abundance, then this process of devaluation will creep its way through everyday sources of status like homes, cars, jobs, clothes, and leisure time. While these “easy” sources of status will disappear, the desire to obtain it will not. And so when abundance arrives in one area, people will pursue scarcity in another.

To better understand this dynamic, we might imagine a list of resources that are likely to resist technological disruption. But this isn’t any old list. It’s The List: those things that will retain value in the long term. Surely such an input would be critical to anyone’s plans.

No one can know the contents of The List with precision. But we can attempt a rough draft by outlining that which even a superintelligence could not meaningfully create:

• Human connection and attention

  • Friendship and love

  • Access to exclusive clubs and social groups

  • Parasocial relationships (an online following)

  • Power and influence

• Human abilities and appearance

  • Top-tier intellectual, athletic, aesthetic, or creative prowess

  • (The point is not to beat the machines; it’s to rank highly against other humans)

• Physically limited spaces and materials

  • Beachfront property

  • One-of-a-kind cultural or natural sites

  • Rare earth elements and their derivatives

• Originals

  • Ideas

  • Works of art

  • Artifacts

  • Retro technology

  • Some NFTs / cryptocurrencies

Principles of The List

At first glance, this draft of The List seems arbitrary and disconnected. What does skill in, say, basketball have to do with the quantity of neodymium in the Earth’s crust?

But there are unifying principles. And it’s helpful to understand these principles when evaluating whether something belongs on The List. In particular:

• Humans are finite. Battlestar Galactica notwithstanding, machines cannot create more humans. And in a post-scarcity world, we will place greater value on whether a human created a resource, rather than the resource itself. Marxists rejoice; the labor theory of value may finally see its day in the sun.

• Earth is finite. There are no planets even remotely similar to it. Machines cannot increase the surface area of the Earth, nor transmute one element into another in any meaningful quantity.

• History is finite. There were a finite number of things that happened in the time before abundance. Machines cannot go back in time. Artifacts, places, and even people from this time cannot be authentically duplicated.

• Humans rank against humans. Even if machines exceed humans in all areas of competence, our status will derive from our ranking against other humans. Magnus Carlsen is still the number one chess player in the world, even though he carries in his pocket a machine that can easily beat him.

• Math beats machines. Machines cannot break cryptography, and there exists forms of cryptography that even quantum computers cannot break. The finite world of machines can never exceed the infinite world of mathematics.

It’s true that in the long run, with sufficient technology, the first two principles can be eroded. Asteroids can be mined, planets terraformed, humans cloned at scale. But for the purposes of planning, these possibilities seem sufficiently distant.

Planning with urgency

Most of the resources on The List are already very valuable, either in dollars or in qualitative terms. But we are still early. ChatGPT was released just three years ago. Tesla’s Optimus robot is still in development. And many people are not thinking about the future with clarity, or at all.

Once the steep part of the disruption curve hits, competition for The List will increase drastically. Therefore the aim of the successful planner is not merely to make plans, but to execute them with urgency. The earlier you start, the better your future becomes.

But what does that mean in practice? Simple awareness of The List will shape how you allocate your personal capital. You may pursue development of a skill not because it will lead to a career, but because you intend to master it. You may invest in assets secured by location, cryptography, or originality. You may place greater importance on relationships, social capital, and audience.

Planning of this nature is a permanent mindset, not a one-time project. However you proceed, it should be with the knowledge that the game of life will continue in the future. It’s a game that requires your active participation: agency rather than helplessness.