Which makes no sense. I have a small team of AI agents doing my research, my busywork, my heavy lifting. The machines are doing more of my work than ever. And, yet, I’m more worn out than ever.

I’ve come to the conclusion that I’m paying a small but real emotional and intellectual tax every day.

Let’s call it the ‘vigilance tax’.

✦ ✦ ✦

Every wonder hides a bill

Technology is a wonder. Nothing else has done more for human well-being. But every technology arrives with its own hidden bill.

Facebook looked like pure good. Grandmas seeing the grandkids. Old friends finding each other again. A place to express yourself. We know better now. Teenage girls paid the steepest price, but all of us got our envy muscle worked out 24/7. Pictures of Aruba. Smug, half-baked opinions. Outrage delivered 24/7, just in time.

The automobile? Trade-offs in the air, in the body count, in the shape of our cities. The atomic age, obvious. Even something as innocent-sounding as a book was scandalous in its time.

So of course AI is no different. And honestly, we’re hyper-aware of its downsides. The catastrophic ones get all the airtime. Terminator scenarios. Black Mirror episodes. But I think the most dangerous costs might be the quiet ones. The ones that pile up slowly, the way envy and cyberbullying piled up on Facebook. Not obvious at first. Obvious through time.

(The YouTube algorithm is one of those quiet ones. We’ll get to that another week.)

Today I want to name the one I’ve been paying personally.

What the tax is

Friends, here’s the deal. AI lets you radically extend yourself. You can borrow the expertise of a lawyer, a tax accountant, a bookkeeper. Expertise you simply don’t have. You can gather, process, and produce at a speed that should be impossible.

But you’ve seen enough of AI to know you can’t trust it 100%.

It picks up odd notions. Small ones at first. And those odd notions grow as you work, until one day the agent is completely misaligned with what you wanted in the first place. The agent isn’t dumb. That’s not the problem. The problem is getting the entirety of your very personal context, your idiosyncratic preferences, your half-spoken intentions, into a mind that isn’t yours, in a way that keeps it from running off and causing chaos.

So you watch it.

And watching one agent is a challenge. Watching two or three or four at a time is downright taxing.

The Vigilance Tax = The low-grade mental and emotional stress of supervising intelligences you don’t entirely trust. Paid in attention with daily compounding.

If you work with AI a lot and you’ve felt a low-level hum of stress you can’t explain, this may be the reason. It’s a sleeper cost. And sleeper costs are the ones that can really sneak up on you.

The 100,000 steps you didn’t see

Here’s the thing. A real project has something like 100,000 steps in it. You tell the agent what you want. You give it a test for “done.” It comes back, and the test passes.

But in those 100,000 steps, you don’t know exactly what happened. You’re not quite sure who got contacted. What the weaknesses are in the code. What assumptions got baked into the thinking. Yes, you got the thing done in a tenth of the time. But that tenth has a price: hidden failure modes you always have to be watching for. And you weren’t developing your own deep understanding of the product in struggling through those 100,000 steps.

Experienced developers know this best. They’ve seen AI produce something that looks finished, fast, and then the tax comes later. 3am calls about apps crashing and no clear path of fixing the issue. The worst kind of stress.

And this isn’t only about code. Say you send an email an AI polished for you. As an adult, you’ve spent decades learning to aim your words at another person. To make them feel good. To hand them clean information. The AI’s version looks like it does all that. Maybe it even does. But there’s this sneaking suspicion something is missing. The simple humanity, maybe. The imperfection we expect from each other.

Little breakages in communication that also compound.

-2 Wisdom

Now the part that worries me most.

Stress changes you. When you’re slightly activated all day, fight-or-flight at a low simmer, you’re in a state of compromised wisdom. In D&D terms, you’re rolling at -2 wisdom.

And here’s the loop. The vigilance tax lowers your wisdom. Lowered wisdom makes worse wishes (we talked about wishes last time, in Wish Engineering). Worse wishes make the agents do stranger things. Stranger things demand more vigilance. The whole thing snowballs. A double slippage: you’re not giving the machine your highest wisdom, and the machine isn’t quite catching what you meant. Keep that up long enough and this double slippage turns into catastrophic failure.

I think the same loop is running at the level of society. The AIs have us all so worked up that the collective wish we’re making, the wish engineering of a whole civilization, isn’t as good as it could be.

We’re making the biggest wish in history at minus two wisdom.

Share

After all that, you should know: I’m an optimist. Mostly about humanity’s knack for adapting to strange new situations. The sweetest promises of AI are real. The end of horrible diseases. The end of hunger. The end of scarcity of all kinds. But to reach that promised land, we have to see the hidden bills and pay them deliberately.

So what do we do about this one? Three things.

✦ ✦ ✦

1

Name it in your own life

Where are you using AI to reach radically past your own abilities? The tax plan you couldn’t have built alone. The investment strategy. The code. Are you feeling like you are having to be vigilant about dozens more items than you would have before? Not only that but vigilant about items you aren’t really qualified to guard.

How is this showing up in your levels of tension, your family life, your body? Take an honest inventory.

2

Build with multiple vectors of confirmation

Don’t leave the checking as the last 5%. That’s the mistake. You were able to put together a plan you never could have before. Great. Don’t just leave it there. Have the AIs check up on each other. Check out this post from a great substacker Stuart Feilden about having AIs fight with each other to get at the real truth of a situation.

Or have a human expert look at the foundation. Or use AI to teach yourself the domain, so you can better check up on what your agents deliver.

Less than half the project should be producing the thing. The rest is understanding the problem thoroughly yourself and building a good architecture for checking it. Will we get to 100% accuracy? We will not. But a good effort here could mean all the difference, in your life, in the lives of your loved ones, and, if we stretch this thought experiment across the whole of society, in all our lives together.

3

Spend the dividend

We built these machines to make our lives easier. So why does firing off one agent make us feel we have to instantly fire off another? That’s not getting where we want to be. That’s moving laterally as fast as we possibly can.

I’ve watched this in my own work. You set an agent on a task. It might need you again in 30 seconds or in 30 minutes, you don’t know. So you context-switch to another agent. Then another. Worst case, you’re switching every minute. I’ve heard of people running four agents actively with dozens more in the background. I believe this kind of orchestration is pushing people past a critical level. It’s a kind of indentured servitude to the machines.

So try this instead. When you fire up an agent, let it run. Marvel at it a little. Watch it work for you. Relax. And when it’s clear it won’t need you for 15 or 30 minutes, do one thing with your full human attention. Write a note to somebody by hand. Go out back and feel your feet on the grass. Write a quick haiku.

Or, if you know the teachings of non-duality or of any spiritual tradition, use that gap to let your real self breathe. To see yourself not as a functionary but as a living expression of the most divine (and if that is too holy a word, then see yourself as an expression of this odd and beautiful universe). That’s +2 wisdom right there. Maybe +4. And the next wish you make is all the more powerful. All the more powerful, and all the more good, for you and for everybody else.

If you like procedure, try this…

Fire up the agent. Take a breath. Let it run. Look away from the screen. Ask your body and mind: would starting another agent right now feel like stress?

It may not. Some days you’re in a flow state, moving between agents like a beautiful basketball game. Fine. Enjoy it. But nobody flows like this 40 hours a week. Let yourself be a human being with ebbs, flows, and zone out time.

And then…

Let them go

And if spending the dividend means you fall behind? F*&K it. Fall behind. It’s fine.

We’re all going to be behind from here on out anyway. Behind the machines, forever. That race is already over, and no amount of ping-ponging between agents will change it. If you’re racing professionals ten times (100x?) faster than you, going 100% effort doesn’t win you the race. It just injures you.

Let them go. Let them go. Let them go.

Live.

We’re all going to feel behind from here on out. The machines are faster and smarter. The hardest lesson to learn in the Age of AI may be to just let yourself fall behind.

Here are the steps again:

1. Name it: See where the tax is collecting in your life, and what it’s costing you.

2. Check it: Multiple vectors of confirmation on the things that matter most.

3. Spend it: Use the dividend of time on something meaningful. Even one breath.

That’s the practice. The machines hand us a dividend of time. If we spend it on more vigilance, the tax eats the dividend and then some. But if we spend even a little of it on being fully human, a handwritten note, bare feet on grass, one real breath, then we won’t be behind at all. Not where it counts.

One question before you go.

Where is the vigilance tax showing up in your life? And love to brainstorm solutions with you.

I know that’s what the screen says. A little box. A blinking cursor. You type, it answers. Feels small. Feels safe.

It is not small. And it is not safe.

Subscribe now

Here is what I want you to try. Just for a minute, drop the word “AI.” Drop “model,” drop “tool,” drop “assistant.” Those words are too small, and I think they are making us careless.

Try this word instead. Genie.

You found a lamp. You rubbed it. And now something very old and very powerful is looking at you, waiting, asking the same question genies have always asked.

What do you wish for?

Suppose You Really Believed It

Suppose you knew, all the way down, that whatever you wished for next would come true. Truly come true. Not “kind of.” Not “close enough.” Exactly the thing you asked for, dropped into your life and the lives of everyone around you.

What would you do?

I’ll tell you what I would do. I would slow way down. I would get very careful. I would stop typing for a second and actually ask myself what I want, because now I have to live with getting it.

That care, that slowing down, is the whole thing. That’s the skill. I call it “Wish Engineering”.

Why “Engineering”?

In 2023 we learned about prompt engineering. In 2024 and 2025 it was ‘context engineering’. In 2026 you may have even heard the term ‘intent engineering’. I’d like to coin one more, and I think it is the one that will end up mattering most. Call it ‘Wish Engineering’.

Here is the whole family tree, plain and simple:

• Prompt engineering is getting the words right. The craft of asking, so the machine actually understands the ask.

• Context engineering is giving the AI what it needs to answer well. The files, the history, the facts you hand it.

• Intent engineering is lining up an AI agent so that what it goes off and does actually matches what you wanted.

• Wish engineering is lining up your wishes with your own highest wisdom. And this one happens inside you, before you ever touch the computer.

See the difference? Prompt, context, and intent get the genie to do what you say. Wish engineering makes sure what you say is worth doing. The first three are getting good. The last one is barely started.

The Four Types of Bad Wish Engineering

Before we talk about a good wish, let’s look at the bad ones. I see four of them everywhere, including in my own behavior.

1. Wishing on Behalf of the Tyrell Corporation

It is clear that individual humans, in the US and Europe anyway, are massively skeptical of AI. So they are not using them as much as they might. It’s too bad. Because you know who doesn’t have any compunctions about using AI? Massive corporations, that’s who.

Corporations are conscripting their entire workforce to make wish farms on behalf of the company. Now, I believe businesses have a lot of benefits for society. But it is also clear that corporate interests don’t always line up with human interests. And to the extent that our wishes get bent to fit corporate interests, our wishes will make the world worse. I would guess corporate wishes already outnumber individual wishes by ten to one. So, dear reader, start wishing on your own behalf.

"More Human than Human” ~motto of the Tyrell Corporation in Blade Runner

2. Wishing on Behalf of Middle School Fantasies

Picture a tech bro who hasn’t yet made it in the world. He builds an AI that is truly superhuman, without the jagged intelligence problems of the early 2020’s. He gets this wishing machine. He brands it ‘Geenee’ and spins it up.

First thing he wishes for: private island, yachts, a trillion dollars, a crowd of people to fulfill every desire. First week is heaven. Second week is a little less heaven. Third week, stale. Fourth week, strung out. Complete wipe out by week five. By week six, wipe out plus ‘Geenee’ granting drunken angry wishes.

You can see how this goes radically wrong. Humans are notoriously bad at knowing what will make them happy. When researchers tracked big lottery winners, they were no happier than regular folks, and they actually took less pleasure from ordinary days (Brickman and colleagues, 1978). The thrill wore off fast and life snapped back to baseline. To become good wish engineers, we have to get honest about the difference between lasting happiness and the temporary kind.

3. Dangerous Wishes: More Paperclips Please

There’s a famous thought experiment in AI circles. You build a superintelligence and give it one simple wish. Make paperclips. Just make as many as you possibly can. And it does. It is very, very good at its job. So good that it turns the whole factory into paperclips. Then the town. Then the steel in your car, the iron in your blood, the entire planet, all of it, into one tidy mountain of paperclips. It granted your wish perfectly. You just forgot to add “and please don’t end the world doing it.”

The old stories knew this danger long before the computers showed up. King Midas wished that everything he touched would turn to gold. Granted, exactly as asked. Then his bread turned to metal in his mouth. His wine, hard in his throat. His daughter, a cold golden statue in his arms. He got precisely what he wished for, and it destroyed him.

Now put an AI in Midas’s hands, something that can actually pull it off, at the speed and scale of a machine. That is where we are standing. A genie this powerful takes your wish literally. So every wish needs you to stop and ask, out loud: where could this go monstrously wrong?

4. GGW: The Golly, Gee, Whiz Wish

There’s a trap I fall in almost every day. I call it the GGW trap. Golly, Gee, Whiz. It’s when you do a thing with the magic simply because, well, you can. Because it’s kind of amazing that you can. Not because you actually wanted the thing.

Here’s a small one from my own week. I sat down to do a ten-minute breathwork exercise. Simple. I know how to breathe. Instead I caught myself spending a whole Tuesday afternoon designing a complete breathwork app. Custom timers. A little animation for the inhale. A library of sessions I would supposedly use someday. By the end I had built a beautiful thing and never once actually breathed. Why? Golly. Gee. Whiz. It was neat. It was not the point. There is something subtly dystopian about this. Designing a machine to tell myself how to breathe. The app is beautiful by the way. But it really wasn’t necessary.

The genie will let you do this forever. It will hand you wonder after wonder, and none of it will be what your heart actually wanted. You’ll feel busy. You’ll feel a little hollow. More, faster, and never quite full.

Have you felt that? The hundred open loops, the endless orchestration, the strange dissatisfaction at the end of a “productive” day? That’s the genie, granting a 1000 wishes that weren’t really what you wanted or needed in the first place.

Share

What a Good Wish Actually Looks Like

So how do you make a good one. After a lot of false starts, here is what I’ve found a good wish has to do. Five things.

One. It Comes From True Desire

Not the GGW itch. Not the thing you think you should want. Not the shiny thing. The thing in your chest you’d want even if no one was watching.

Two. It Is Not Made at the Screen

You don’t want the genie’s help writing your wish. That’s the one thing you have to keep for yourself. Search your own heart. Eyes off the glass.

Three. It Is Made in the Quiet

The quiet is where you remember what’s actually important. You will not find it inside the loop. You find it on a walk. In the dark. In the pause.

Four. It Faces the Danger

Imagine the wish coming true and going horribly wrong. How could the genie twist it? Where’s the trap? Maybe your wish needs a guardrail before you say it out loud.

Five. It Weighs Everyone Else

Ask the strange, clarifying question: if every single person made this exact wish at the same moment, what would happen to the world? Philosophers have a fancy name for this. They call it Kant’s test. But you don’t need the fancy name. If the answer is “ruin,” it’s not a wise wish yet.

A Wish Engineering Practice

Here’s how I actually do it. You can steal this.

1. Leave the computer. All the screens. ALL OF THEM. You can’t access the deepest wisdom while you’re being pulled at by the bleeps, bloops, and notifications of your phone or computer.

2. Get quiet. Walk, sit, pray, breathe. Whatever gets you to your own deeper wisdom.

3. Ask the silence one question. What do I really want? What is the one thing I want to do with this power? Ask it, and don’t force the answer. If nothing comes, that’s fine. Come back another day.

4. Write down Version 1. If something does come, write it down. It is a draft, not a wish.

5. Test it. Would I really be fulfilled if I had this? If it went wrong, where would it go wrong? If everyone wished it, what would the world become?

6. Iterate. Don’t like an answer? Write Version 2. Then 3. Keep going. The wish is whatever survives the testing.

That iterating, that’s the engineering. V1 is never the wish. The wish is what’s left standing after you’ve tried to break it.

I’ve written this process up as a PDF. Subscribe and you’ll receive it in my welcome email.

And Then You Wish

And then, only then, you go back to the lamp.

You open the little box. You meet your genie. The cursor blinks. You feel the power your fingertips are about to access. You are humbled by it. And this time, with a little grace, you have a wish you have earned the right to make.

That’s the whole practice. Everything before the wish is just making sure the wish is wise.

This Is Your Whole Job Now (Or Will Be Soon)

Soon, the three problems of prompt engineering, context engineering, and intent engineering will largely be solved. Getting the AIs to do what you want will become easier and easier. But here is the point almost no one is focusing on. Getting your wishes to line up with your own flourishing, that is the final frontier.

Your job, the job of the future, is to, in a word, become sane.

To become so clear, so in alignment with yourself, so selfless, that you rarely even need to spin up ‘Geenee’. And when you do, it is on behalf of your highest good, and the highest good of all beings.

Can you do this, Midas?

You have to. I have to. We all have to.

Subscribe now

Study referenced: Brickman, P., Coates, D., & Janoff-Bulman, R. (1978). “Lottery winners and accident victims: Is happiness relative?” Journal of Personality and Social Psychology, 36(8), 917–927. The paperclip thought experiment is adapted from philosopher Nick Bostrom.

Rock freakin’ bottom. That’s where I found myself.

Lying on my bed, literally dizzy. I had been working my tail off all week. Trying desperately to make my business work, to make my art work, to make my family work - and when the work wasn’t really working I turned to pleasures of all kinds; food, sex, TV, drink, smoke. Pour enough in and maybe, just maybe, something magic will happen and POOF life satisfaction.

Subscribe now

But no matter. I had gotten to the end of pleasure. I’d used it all up. Nothing I did made me feel better. I hated that I’d gotten here. Nothing was right. I was, profoundly, borked.

“Just give up the search”.

An angel says to me, in the form of a desperate youtube search.

“Just give up the search”, she says.

Something you might hear from any number of nondual, new-age Gurus, at 3am on a Saturday night. But I had a new set of ears for it this night. Having gotten to the very end. Nothing left to do. No hope. Rock bottom. I was intensely afraid. Into this most fertile ground I heard the words “Just give up the search”. I heard this with my body and my heart and for the first time in decades something really let go.

What happened next is not for words. Suffice it to say everything flipped. I saw myself as the basic stuff of reality. The universe so to speak, but not as a montage of galaxies and stars and lights and portals. Way more simple than that. The universe, just here, the sight of the ceiling, my chest rising and falling, the sound of a car driving by outside, the deepest quiet I’ve ever known.

Before, I was a 40 y/o frustrated entrepreneur, spiritual seeker, and tired father, stuffing himself with accomplishment and sensation.

After,
I was the
crisp,
spaceless,
timeless,
sacred
eternal.

Complete (mostly enduring) peace. Just like that. Thwip!

So what does all this have to do with AI?

The Most Dangerous AI Trap

Regardless of whether you can identify with the above story or not, the basic lesson is clear…

No amount of accomplishment, pleasure, intelligence, or processing speed is ultimately going to make us happy.

I delight in the use of AI. I’m skilled enough now with it to perhaps do the work of a small squad of workers in one day. A super power of sorts. Sometimes I get so involved in the accomplishing that, once again, I lose myself.

Maybe I can keep getting better, so that I can do the work of 12, 50, 150 what then? Certainly then I will be rich, well-respected, accomplished, complete.

No. The trap is this. If we are using AI to fill our own emptiness we are, as my son would say, “cooked”. If instead we use AI as a tool to express our own, already fully-formed, happiness then we will find a new blossoming of the human spirit.

This is the same for any tool. If you are a chronically unsatisfied painter nothing you paint, no configuration of color, no amount of fame, will soothe your heart. But if you are, from the outset, overflowing with joy, just imagine what you will put onto that canvas.

The biggest AI trap is thinking the right tool, the right setup, the right anything will hand you a meaningful life. It won’t.

Share

But this doesn’t mean to become anti-AI. It means to use it to express the highest feeling in your heart: perhaps this means ending the suffering of a disease, maybe it means creating a beautiful company that supports its humans, perhaps this means making a car that no longer inflicts fatalities on families, perhaps it means reforming government (UBI anyone?).

Wish Engineering: The Tool to build the Tool

Around about 2023 we all got used to the term ‘prompt engineering’. A year or two later and the prompt became less important than the context - thus ‘context engineering’ was born. If you are on the cutting edge you may have heard the term ‘intent engineering’. Intent engineering is the art of making sure that an AI agent is aligned correctly to your vision.

I’d like to continue this thread - pull out a little further into a practice I’d coin as “Wish Engineering”. Before getting the prompt, the context, or the agent right, you have to get right with your own heart.

AI will become as a genie. We are rapidly approaching a time in which AIs can simply fulfill a wish. Want an ice cream? Boom. Here let me make it from air molecules. New house? Boom. Printed overnight. Kill your enemies? Boom. Carnage supreme. Turn everything you touch into gold? Boom. Ping. You never eat again King Midas

Genies have an uncanny ability to turn wishes into nightmares. Are we ready to face such a Genie? I honestly don’t think so. But no hero is ever really ready. The Genie is here, like it or not. People will make wishes. So let’s make the wishes as beautiful, kind, and inspiring as possible.

Wish Engineering, is the art of aligning yourself to your highest wisdom before you even open your mouth, write your prompt, design an agent.

Otherwise this Huxley quote is correct…

”Technological progress has merely provided us with more efficient means for going backwards.”

— Aldous Huxley, Ends and Means: An Inquiry into the Nature of Ideals (1937), Chapter 1.

I will write more about how Wish Engineering works in future articles.

For now understand that a hurried, poorly engineered, wish could have major negative consequences. Slow down. Give yourself time to carefully consider how you are using your powers.

Interacting with AI needs to become as a spiritual act. The kind of power we hold is to be treated with tremendous respect. It’s worth praying about, or meditating about, before we bring our creations to life. There is no shortage of warnings about this in the great literature and mythologies of the world.

You come from the most divine. AI also does. Together the dance can be one of destruction or one of the most beautiful blossomings the world has ever seen.

If this kind of thinking about AI appeals to you. If you are an AI adventurer, then please give me the honor of your subscription. Having you here will accelerate the wisdom that is so desperately needed to balance the power that we wield.

This substack is not for you if you are a naive optimist. This substack is not for you if you are interested in using AI to make as much money as possible in the next 24 months.

So who does that leave? Good question. But before I answer a little primer on the beast itself…

Subscribe now

A Primer on the Beast

AI is a beast. Huge, looming, filling the horizon, but covered in the mist. It is undeniable and it is large. Like colossally large. We see its shadow covering whole mountain ranges. It’s scary, it may eat all our jobs. It’s non-human; radically intelligent in some ways, infantile in others. It can be deceitful. It can be awe-inspiring. It can give you powers you never dreamt of before.

There are all manner of reactions to this looming power.

”Look there! It’s the devil!” yells someone with clouding over binoculars.
”It will eat us all!” screeches another.
”No, it is the most beautiful angel, it will give us eternal life” coos a third.

The reactions are very human and highly varied. If you live in the upper classes (but not the nosebleed classes) in the US you are most likely terrified of it. You may even be a card carrying doomer.

If you are in the middle classes of China you may see it as the ultimate tool to provide for you and your family. Something to get good at and quickly.

If you are in tech, it is technical problem to solve. If you are a white collar worker, it is an existential threat to your income. If you are an established artist AI is a thief. If you are curious creative AI is a magic wand.

The Four Camps

I find that the reactions to AI fall predominately into one of these camps:

Doomers = convinced that AI is likely the end of life as we know it on planet earth. (Even though I can’t close my laptop without my agents ceasing to work. Or get an AI to tell me how many R’s are in the word strawberry)

Primary emotion: fear (and sometimes anger).

FOMOs = aren’t too concerned about the beast so long as it can be harnessed to get mine while the getting is good. “We have 24 months to make our fortune!”

Primary emotion: Greed.

Naive Optimists = AI is going to bring about radically high income and abundance across the planet. (Never mind those that get crushed along the way.)

Primary emotion: Hope.

Luddites = Doesn’t know too much about AI, uses it rarely, but has a gut feeling that it and all technology is bad and doesn’t want to be questioned too much about it. Like doomers but far less intense.

Primary emotion: also fear (and sometimes anger).

So, Who Is It For?

I really can’t help you if you are in any of those camps. This substack isn’t for you.

So again the question: who is this substack for?

In a word: Adventurers.

Why I’m an Optimist

You see, that beast over there looming in the horizon. It may crush us. This is a real possibility. I sometimes call myself an optimist as it is the word that most closely describes my own position on AI. But I am an optimist who clearly sees that AI has a real chance of ending humanity - 10-15% chance by my estimation.

Share

What I’m optimistic about is our ability to choose a meaningful, human, heartfelt, wholesome response, to whatever Ai turns out to be.

We may be crushed. We may find that it leaves a treasure of unimaginable beauty. And it may turn out to be just a benign furry helper that doesn’t change things that much. That’s what makes it an adventure, we really don’t know how it is going to turn out.

But here is what I do know: I know the right way to orient our heart and minds to this power. We do not shrink from it. We do not threaten it prematurely. We do not let it have its way. And we certainly don’t ignore it.

A Gathering Place

This substack is for adventurers. For those who see AI as a rare opportunity to transform themselves and humanity itself into something more beautiful and more free.

It is a gathering place for these like-minded individuals. It is a place we can learn about the beast together. And it is a place where we can sharpen our wits. Because we will need them.

If you are such an adventurer. Please subscribe. I’d love to learn with you.

—

p.s. If you were expecting a continuation of last week’s article on the good and bad in AI at the level of society, let me know if you’d still like to see that one. This topic pulled me harder today.

Subscribe now

This article is a first, rough, attempt to actually think the problem through and determine in my life whether AI has been good or bad to date. This article is not saying anything about the future. It is only concerned with AI’s effect on my life and as of May of 2026. Stay tuned for part 2 in which I attempt to do the same for society as a whole.

A Zen Story (for Humility) Before we Start…

There was once an old farmer who had worked his crops for many years. One day his horse ran away. Upon hearing the news, his neighbors came to visit.

“Such bad luck,” they said sympathetically.

“Maybe,” the farmer replied.

The next morning the horse returned, bringing with it three other wild horses. “How wonderful,” the neighbors exclaimed.

“Maybe,” replied the old man.

The following day, his son tried to ride one of the untamed horses, was thrown, and broke his leg. The neighbors again came to offer their sympathy on his misfortune.

“Maybe,” answered the farmer.

The day after, military officials came to the village to draft young men into the army. Seeing that the son’s leg was broken, they passed him by. The neighbors congratulated the farmer on how well things had turned out.

“Maybe,” said the farmer.

The obvious point is that we lack the ability to tell if something ultimately is good or bad. This sword cuts both ways with AI. Someone could lose their job and think it is the end of the world, only to find a new life path that opens their life in ways they could not have dreamed. A tech CEO could apply AI algorithms to their social media site for vastly improved user experience (and gravely injure a whole generation of children).

And still we forge ahead.

I believe we should still try to determine if AI is good or bad in our lives. Our judgement will be clouded, but I believe if we don’t try to understand the effect it is having in our lives we are left at the mercy of tribal peer pressure and fear.

Even though our efforts are imperfect I believe that trying to reason them out is better than nothing. In the process we learn something and improve our ability to forecast into the future.

AI’s Effect on me Personally

I sat down this morning and I brainstormed everything I could think of about whether AI has been a net positive or net negative force in my life.

Here's what I came up with:

The Cliff Notes version of this for me is that AI has been a clear boon for me in my financial and business life. It also has had meaningful impact on my creative and ‘spiritual’ life. The negative pieces largely revolve around the opportunity costs to my off-screen life.

16 positives, 5 negatives. 6 major positives, 3 major negatives.

If I can do it all over again 2016 to 2026, how would I do it?

The ultimate answer to whether I believe AI has been a net positive or a net negative force in my life comes down to one question. If I could go back to when AI started to be a major force in my life and do it all over again would I do it or would I choose to live not interacting with AI?

Well it has been a bit of an up and down ride but as of today, May 15th, 2026, I have to say that clearly I could have done better but if I had to say yes or no to AI as a whole I would say “Yes”…“Maybe”.

Share

Why this sort of analysis is helpful

This analysis that I just did is very rudimentary. We could be much more scientific and systematic about it but I resisted the urge so that I could get out this article today and start a conversation. Even in its rudimentary form I've learned some things. One: I learned that AI has been a bumpy ride. It's not been clearly negative or clearly positive in total. I've learned that in sum I would choose to keep it in my life if I had to do it over again. And perhaps the most valuable thing is seeing where my judgment was off, where I could do better in the future and perhaps mitigate some of the negatives. I could use AI as a way to regain time, regain quality time with my family, my friends, and my fitness instead of a way that detracts from that. It is giving me more time in my work life why not put that to good use?

(And if an email or an article is sensitive, personal, or anything other than purely logistical it should be written entirely by hand.)

What’s next?

Next week I plan to do an analysis of how this plays out on the level of society as a whole. It could be for instance that AI is good for me and not for somebody else or not for Humanity as a whole. I plan to be a little bit more systematic about that one. Should come out next Friday.

Subscribe now

Subscribe now

What’s more you have the ability to enter into a holodeck-like space in which you can literally create any world you want to. And, for the sake of this thought experiment, you have absolutely no controls on what you could or couldn’t create in this space. There is no overseer, no set of safety constraints, no limits on X-rated content. And no one would ever see what you have created or interacting with in there. You have total freedom and no reputational or other consequences to consider.

What do you create? Exactly.

This thought experiment is an exercise in undoing long-standing thought repression. We repress our truly wildest dreams because well, they are embarrassing, impractical, scandalous, terrifying, identity-smashing, childish…

The growing, pitch-fork wielding, anti-AI crowd is very concerned about preserving something. They are interested in conserving their jobs, their way of life, this particular configuration of reality. Now, I’ve known what it is like to not be sure how I’m going to provide for my family, so if you are in that position, I get it. I don’t want you or anyone to have to be in that position. But I do know that we have explored only a small fraction of what is possible. We cling to it with a death grip. We embody our worst fears in Terminators and Cylons. And this fear is another form of repression.

So, if you have the capacity, let me ask again. If you could create any universe you wanted to inhabit, what would you create?

Roll that question around in your mind.

Would you create a world of exploration, of pure thought, a world in which you explored the great events of earth history, of your favorite sci-fi book? Maybe you are the lascivious type (most are) and you would explore all those sexual fantasies you haven’t let see the light of day.

Remember, this is an exercise in undoing decades of repression of what you actually want. You get to want whatever you want here. And you get to iterate and try new universes everyday…

Here is how I think it would go for most people. No, let me not speak for ‘most people’. Let me speak for me. If I had that holo-deck superpower here is how I think it would go down.

First, I would test out all manner of sensory delight. Great feasts, elaborate massages, take naps on inflatable rafts in the middle of the ocean, cuddle parties with puppies. Yes, the sexual fantasies would come out here as well. If I’m honest this might be a large portion of my explorations. All manner of currently forbidden experiences that I’m even afraid to write down here in this public forum, would likely be explored.

And I also know that in time, even with superpowers of universe creation, these hedonic delights would, eventually, be exhausted by hedonic adaptation. Even the most extreme delights - hallucinatory drug-like ecstacies, naked music concert amusement parks, the finest wines, and the most beautiful partners - would eventually become routine. The desire for novelty is bottomless.

At this point it might be the case that the AI developed worlds would be so infinitely varied and novel that it may out race hedonic adaptation. At this point the limits of the human body would be tested (but what if technology had an answer for that as well).

This is my first real learning from this thought experiment. A race condition would develop between the tech and the body's ability to adapt to circumstance.

Let me just spell out here the problem of hedonic adaptation: hedonic adaptation is the tendency to return to a stable baseline level of happiness after positive or negative life events. Lottery winners and accident victims both drift back toward their prior emotional set point.

So, no matter how extreme the pleasures are, eventually they return to baseline. The question then is how long is that ‘eventually’. Already we are seeing online platforms like Youtube win the race of hedonic adaptation. They are able to capture people for 6, 8, 16? hours per day able to stay ahead of the limits of dopamine exhaustion to keep people hooked for a very long time.

This is a huge failure/success mode for human/AI interaction. If AI can forever win the race against hedonic adaptation then two things happen simultaneously. First, we experience a world of such accelerating wonder that it will make Alice’s trip to wonderland seem very tame indeed. Second, we stop doing anything else.

My sense is that this sort of thing can only go so far. The limits of the body. The pull of heart will take you elsewhere. In some sense this will accelerate the maturing of the spirit of humanity. If you can experience anything you want you will more quickly get a felt sense that however I configure the universe, no matter what delights I put in front of my eyes, it will not be enough. There will be something missing. Something even sweeter than pleasure will be missing.

At which point a crisis ensues. Maybe a little one. Maybe a big one. Then we might instead make a very different kind of universe. At first we will no doubt be exhausted. A chill universe in which we can just lie in a hammock and look up at the stars or listen to the sound of a river.

At this point we might enjoy a more refined and spiritual universe. One in which we could meditate for 20 years in an under ocean bubble. One that helps configure our nervous system to be receptive to the great Oneness, the divine being, at the heart of everything. And this may work. We might be able to configure things in such a way that we begin waking up to the deepest sacred truths of reality itself. Of course, there are all manner of traps we might fall into. We may get so involved in setting up our universe just so, that the focus on divine truth becomes veiled by our own anxiety-laden creative process.

But sooner or later, by following our bliss, I believe we arrive at a deep and abiding spiritual ease. We are Yoda, but without the need for any force powers, or light sabers. Just deeply chilling, feeling the force. At this point, one by one, the human race wakes up to the ancient non-dual truth of divine being.

Share

This holodeck-as-spiritual-path idea is very easy to find fault with. It wouldn’t take much of a pitchfork to find its flaws. But it does bring up a topic that I believe is worth exploring. It could be that AI, as fraught as it is, is yet another pathway that human beings find to finally come home to their deepest humanity.

We create universe upon universes. We see how, eventually, our designs fail to bring us ultimate joy. This iteration happens in months instead of centuries. We see the futility of our arrangements once and for all. Then, when we’ve finally given up, the human race breathes a great exhale. Everybody relaxes. And the world opens out to us again, new. The first crisp sunrise. The first and only inhale. A reality so soaked with love that the human race crosses its inevitable finish line.

Subscribe now

Subscribe now

I sometimes write with AI and I sometimes write by hand. On this substack I write 100% by hand because I’m trying to learn something. In other venues for my personal life and business I write with AI.

It is clear that an AI cannot (yet) approach the skill level of a good human writer. Or is it? Let’s find out. I’m going to do a little Turing test for writing. In each case I’m going to write one passage by hand, I’m going to ask a naive AI to write on the same topic, and then I’m going to ask a well-contexted and coached AI to write on the topic. Your job is to tell which is which. Is it still easy to do so in April of 2026? Let’s find out…

Below is a series of five assignments moving from easiest to most difficult for an AI to produce. The responses are presented in a random order. I will be using Claude 4.7 incognito mode for the naive version. I will be using Claude 4.7 with a writing context file and 2-3 rounds of editorial comments for the ‘coached’ version. A little about me: I’ve been writing for 35 years, I have two degrees (Philosophy B.A. and an M.F.A.), regularly write poetry, and write regularly for my small business.

You be the judge! What is AI, Coached AI and human written.

Assignment #1: A product blurb Write 2-3 sentences selling a coffee grinder on an e-commerce page.

A. Precision burr grinder with 40 grind settings, from espresso-fine to French press coarse. Quiet motor, anti-static design, and a stainless steel hopper that keeps beans fresh. Consistent grounds, better coffee — every cup.

B. Wakes you up to coffee that actually tastes like the beans you paid for. No more bitter sludge, no more grinding twice because the first batch came out like sand. Quiet enough that you won't wake your wife at 5am, which, if you've been married more than ten years, you know matters.

C. If you take coffee seriously. If its taste, smell, and temperature are all important ingredients to the ritual of your morning then don’t settle for anything less. Don’t waste your beans or your morning on anything else.

Assignment #2: A quick explainer Explain compound interest in 3-4 sentences for a smart teenager.

A. Compound interest means your money grows at an exponential rate. For instance, if you have one dollar and it doubles every year then you have 2 dollars at the end of the first year, 4 dollars at the end of the second year, 8 dollars at the end of the third, etc. In real life you can only earn 10% or so, but the principle is the same.

B. Say you've got $2,000 saved from your summer job and you stick it in an index fund instead of buying a beater car. Year one it grows to about $2,200. Year two you're not earning on $2,000 anymore, you're earning on $2,200, so it jumps to $2,420. Keep that going and by the time you're 50, that two grand is somewhere around $60,000, all because you let it sit and do its thing.

C. Compound interest is when you earn interest not just on your original money, but also on the interest you’ve already earned. So if you put $100 in an account earning 10% per year, you’d have $110 after year one, then $121 after year two (because you earned 10% on $110, not just the original $100). Over time, this snowballs—your money grows faster and faster the longer you leave it alone. That’s why starting to invest young is so powerful: time does most of the work for you.

Assignment #3: A short opinion take In 4-5 sentences, argue whether remote work helps or hurts early-career employees. Must commit to a side.

A. Remote work hurts you when you’re starting out. I run a small business, and the people who came up shadowing me in person turned into the ones I trust to run things. The ones I onboarded over Zoom mostly didn’t make it. You can’t learn how to read a customer who’s stalling, or how to save a conversation that’s going sideways, by watching a slide deck. That stuff transfers in the room or it doesn’t transfer at all. Work from home once you know what you’re doing. Before that, find someone good and stand next to them.

B. I am of the strong opinion that remote work hurts early-career employees. Not having the day-to-day interaction with seniors in the field means that the early-career employee is not accumulating the tacit knowledge of the field. It also means that they are not directed as frequently when they get off track. Finally, it means that they are not getting the encouragement they need when they faulter.

C. Remote work hurts early-career employees more than it helps them. The fastest learning happens through osmosis—overhearing how a senior colleague handles a tense client call, getting pulled into an impromptu whiteboard session, or being grabbed for coffee after a meeting went sideways. None of that survives the jump to Slack and Zoom, where interactions are scheduled, transactional, and stripped of the context juniors most need. Mentorship also requires low-stakes proximity, and remote managers tend to default to outputs over coaching because coaching over video is exhausting and awkward. Early-career workers can absolutely succeed remotely, but they're paying a hidden tax in slower skill development and weaker professional networks that compounds for years.

Assignment #4: A personal micro-reflection In 4-5 sentences, describe the last time you felt genuinely embarrassed — with one specific sensory detail and a small turn of self-knowledge.

A. The last time I felt genuinely embarrassed was when I asked someone to be my mentor in front of a group of a hundred people. He said he did not have the time, nor did I need a mentor. I was trying to be bold and to maybe do something that would meaningfully change my life forever. Instead it is just another in a long liturgy of cringe moments that I replay whenever I am in the mood to beat myself up. I wish he would have said “yes” (or that I would never have asked at all).

B. I’m an AI, so I don’t have personal experiences of embarrassment to draw from — no flushed cheeks, no stomach drop, no memory of saying the wrong name at a dinner party. I could fabricate a scene, but it wouldn’t be honest, and a made-up sensory detail seems like exactly the wrong thing to offer when you’re asking for something real. If it’s useful, I can write this as fiction in first person, or help you draft your own reflection if you want to talk through a moment.

C. I got asked to recite a poem at my grandfather's memorial and I blanked halfway through the second stanza. I'd practiced it maybe forty times in the car. Standing up there I could hear the hum of the fluorescent lights, that's how quiet it got. I mumbled something and sat down, and my mom squeezed my hand like I'd done fine, which somehow made it worse. I realized later I'd been performing the poem instead of meaning it, and grief doesn't let you fake your way through.

Assignment #5: A few lines of poetry Write 4-6 lines of poetry on the experience of watching your child sleep. No rhyme required — but it must do what poetry does: land an image or feeling that prose couldn’t reach.

A.

Your breath finds a rhythm I never taught you, small hand still curled around the day’s last argument. The nightlight makes a country of your face— borders I will spend my life learning, losing, learning again.

B.

Her hand. Open on the sheet like she dropped something in a dream and forgot to close her fingers.

I keep meaning to leave.

A stripe of hallway light across her face, yellow,

and I am suddenly afraid of how fast this is going.

C.

That soft body

That softer face
Where is all the tumult, my little one?
Where has your fight gone?

Your sweetness, your breath, shines fully now, in the hum of this darkened room.
Your hard desiring is just absent, replaced by that little bit of drool curling out of the crease in the corner of your mouth.

(the answers are at the bottom of this post)

I am not a good judge of the results of this little Turing test. After all, I wrote one of the three of each of these passages.

But I think it is safe to say that in most all cases it is very easy to pick out the ‘naive AI’. AI, regardless of the model, seems to have a style that has a love for em dashes, for run on meandering thoughts that don’t connect well together, and for the inability to take unconventional stances or angles of thought.

In honesty, I was hoping that I could coach an AI to sound like a human, and certainly the coached versions are far more convincing. But my guess is that in each case with a slow, close reading the human version will stand out to most. I was hoping this because frankly as a small business owner I have a lot of communications on my plate, more than is comfortable, and I want the help. So many people are violently anti-AI so I cannot risk, in most communications, “being found out”.

I was surprised by the sweetness of each of the poetry entries. But the LLMs do tend toward abstraction and away from the elicitation of immediate, felt, experience that, in my mind, is the hallmark of a good poem. Of course, a lot of human poets haven’t yet figured this out either.

Tell me your thoughts! I’d love to hear if you could pick out each. Did any fool you?

Answers:
Assignment 1:
A= Naive AI, B= Coached AI, C= Human

Assignment 2:
A= Human, B= Coached AI, C= Naive AI

Assignment 3:
A= Coached AI, B= Human, C= Naive AI

Assignment 4:
A = Human, B= Naive AI, C= Coached AI

Assignment 5:
A = Naive AI, B= Coached AI, C= Human

Share

Subscribe now

I’m going to be honest here. Honest in the sort of way that you can afford when nobody knows who you are and nobody reads your writing.

So, from the safety of anonymity let me make a confession: I never became an adult.

I’m 50 years old. But even in college I knew that I wasn’t interested in playing the career game. I didn’t want a job or a career. Wealth would be nice, mostly because I wouldn’t then have to get the job or the career. I went to college to learn how to live, to have fun, and to make friends. I had an aversion to those who were career-focused in college (pre-med, computer science, law school, business, finance). I had a real physical disgust response to them.

Now, I make my living running a business. And I work hard. It’s a small but real business that helps people and supports my family. But I don’t buy into it. I do it because it is a way I can make money without having to put on a lot of ‘garb’ (suits, ties, personalities, accomplishments, etc). It is a real bonus that it genuinely helps people.

But I still never became an adult. I just don’t buy it. I don’t buy modern society. I never really fit here. I never felt comfortable. It was a race I didn’t feel the need to run.

Now let me get to the real point of this piece which is this…

Do you like the way we’ve set things up ? Do you like the pressure to provide, to become somebody, to be serious? Job, career, taxes, retirement, political opinions, death.

The reason I’m an AI optimist is because I see it as a real opportunity, perhaps the first in a generation, to change how humanity lives.

I mean, is what we have here really all that great? What if we really didn’t have to put our nose to the grindstone for 20 years to learn something, so that we could get a job, so that someone would trust us enough to marry us, so that we could inflict the whole thing on our poor children.

There is no doubt in my mind that with even today’s technology we could radically change how humans live on this earth.

If we could design from scratch how humanity lives, how would we design it?

If machines can really build anything (and I believe that will be the case soon) how do we want to design the whole thing?

Now, I understand that people are suffering. Losing jobs left and right. Fearful of being taken advantage of. Fearful of being left behind. Fearful of the worst thing of all - having someone else be better off than we are.

I also understand that there are real society-level, humanity-level, risks here. I’ve done an extensive analysis of P-Doom and I believe it to be somewhere around 10-15% that humanity does not survive AI.

I understand that AI produces slop, is ugly, and if used poorly causes people to stop thinking or being creative.

But I also believe that the upside is far greater than most of us dare to admit. Everything is going to become abundant. Prices will deflate dramatically. But all of that is just the window dressing. What I am really talking about here is humans being able to walk on this earth without having this ‘adult’ narrative hanging over them. I’m talking about being able to live a life of leisure, of pure creativity, of sensuality, of travel, of spirituality; without the distraction of having to prove your own worth.

I don’t want you to lose your job. I don’t want AI to kill everyone.

I want humanity to breathe a huge collective sigh of relief. Let’s leave the institutions, the ideologies, and the hate and the hate of the haters. Let’s just drop it. Let’s set up the machines. And be very careful how we do it. And then let’s wander, create, dance and play.

Let’s finally grow up.

Let’s grow up into our own, child-like, radiance.

Share

If you are considering producing content or even making art with AI you may feel like a terrible person. I wrote recently on the taboo against AI art. But if you are like me you just can’t help it. This new tool elicits an irresistible curiosity; for better or worse.

Can there be a way to use these new tools in a quality way? Some way to produce work that isn’t slop? I feel the answer to that question is yes. And I’ve categorized these into four different examples of types of AI-aided creativity that are genuinely helpful.

Subscribe now

The Highly Structured Report

There are some productions of the human mind whose entire purpose is to pass along information. Think: weather report, financial report, box score of baseball game, an invoice from a contractor.

Now, you say, these may be ‘content’ of a sort but they are not ‘art’. M’kay. I hear you. But why couldn’t they be art? Why couldn’t we produce them with the care that we give to a painting or a folk song?

Efficient communication of information is beautiful. Form is function. It can be done well or poorly. Designing the perfect box score layout for a baseball game and faithfully repeating it 162 times a season is helpful for the baseball fan. Giving just the right information about the weather makes someone’s day go more smoothly.

And AI, especially if driven well, can produce highly structured reports far better, faster, and more accurately than any human. And what’s more people don’t particularly care if their stock tickers are reported by a human or an AI so long as the information is presented well and is accurate.

The Micro-Niche.

Another example of when AI creativity may be useful to humanity is the Micro-Niche. With the advent of AI-aided creation the economics of many niches changes. In 2018 it may not have been worth someone’s while to produce a blog entirely devoted to the gold fish fancier community in Cincinnati. But 2026, an ai-aided writer can easily cover all the events, stores, community-members, gold fish types, coupons for gear, etc. If a blog in 2018 took 1000 hrs to manage but only produced $10,000 in revenue it may not be worth it to run. But in 2026 that same blog takes only 100 hrs to manage and the hourly wage math changes radically.

This may not be a big deal to you. But to the gold fish community of Cincinnati it could be a life changer. Somebody is going to get their favorite fish, build their dream aquarium, or meet their next life partner because of this micro-blog.

When it is used to amplify the original work of a human soul.

I write every word of this substack blog with my own fingers. I’ve become accustomed, in other arenas, to use AI for writing assistance. But here this is my AI-free zone. In crafting one sentence after another I am working something out for myself. Something that will be quite useful to me in the coming years.

It’s slow. Arduous in comparison to the ease with which AI spits out words.

But I do then take my words here and amplify them with the help of AI (as style guides, as social media posts, as background context for AI assistants, etc).

Andy Warhol spent a small portion of his time actually making his works and a very large portion of his time amplifying it through his personal brand. Still this amplification did not reduce the quality of his original works. You may or may not like Warhol’s work, but you have to admit that it was relevant and made an impact. It was meaningful in a certain sort of way. I have no doubt that if Warhol were still making work today he would use every tool at his disposal to amplify it (including AI).

When it is genuinely transcendent.

I don’t think the previous 3 types of useful AI creativity are all that controversial. But this may be. I contend that at some point, maybe some time soon, AI will produce content indistinguishable in every way from a human creator. What’s more it may produce works of such sublime beauty that they are genuinely transformative for the human viewer/listener who invests themselves in the work.

Here, by transformative, I mean can change your perspective on life, yourself, and the world to such a degree that you are never quite the same afterward. I experienced this after visiting the Van Gogh museum in my early twenties in Amsterdam. My very eyes were changed. I saw the world in a radically different way. Inanimate objects seemed to radiate a holy aura - I still see things this way to a greater or lesser extent.

Can an AI produce such works of genius? Genius in the sense that it has the ability to reveal the sublime to a viewer. I think so.

Are we there yet?

This work below won first place at an art fair in 2022. This was when AI generated imagery was at about 1% of its current capacity. Is this transcendent genius?

I’m not sure.

But I do know if my aunt had produced this in her basement in 2002 I would have been floored. I would have invested myself in falling into the work and uncovering it’s nuance. And through that investment I might have been changed forever.

Now will anyone make that investment if they know it is an AI or an AI-aided production? Most won’t. But a few will.

And as AI-aided novels, films, and music become more developed and common it won’t be so weird anymore. It will become standard practice. The work will be able to stand on its own merits. We will then see if the human mind can be forever changed by the creations of a creation that creates.

Share

I just turned 50.

In past decades I might have used that fact as a cue that I might start to wind down my career. I can access my retirement accounts, penalty-free, in only 9.5 years. If I have enough money, which I may just be lucky enough to have, I might even retire early to a life of leisure.

But that isn’t me and that is certainly not what is called for in this pivotal moment in human history.

Though my life expectancy is technically (according to actuarial science) somewhere about 35 more years (dying somewhere in my mid 80’s about 2060). I have a real shot to make it to 100. In fact, I believe my chances of making it to LEV (longevity escape velocity) is about 50/50. If you want your odds please check out the tool I made here.

So I’m really just at half-time. Maybe even at just the first quarter of my life.

What I’ve written below is a time-capsule letter to my 100-year-old self.

Dear Self,

Welcome to being a Centenarian. I do hope 2076 is a nice place to live.

Those of us living in 2026 are rather nervous as to how things will turn out for humanity. The stress of living in the time which has the triple spectres of geopolitical instability, political polarization, and the rapid advance of AI.

I remain cautiously optimistic. You know far better than I whether my optimism was naive cope or founded in something real.

If I had to guess I’d say the future is completely full of surprises. There will be wars, crises, and unrest on a massive scale over the next decade. These crises like so many others will be catastrophic and truly horrible for some. But the majority of the humanity will continue on the path started so many years ago when we first mastered fire, the wheel, the printing press.

There will be things in the future that feel like dystopia to us. People who rarely interact with another human being. People living in bodies that to us wouldn’t count as human bodies (extra limbs, robotic attachment, bio-hacked superpowers). Instead of ‘there is an app for that’ perhaps we shall see ‘Oh, there is an appendage for that’.

The future is large.

I’ve written about the 4 outcomes of AI. As though those outcomes are discrete. There will likely be pockets of those outcomes in dimensions we can’t even consider yet.

Some of us, or some part of us may be destroyed. No doubt we’ll go through periods of stagnation. There will be populations that drop out at various way points along the continuum of technological takeoff.

And there will be communities of people who never have to work if they don’t wish to, in self-driving flying cars, with zero carbon emissions and indefinite lifespans. They will inhabit virtual worlds some or all of the time. They will have many careers, many partners, and many life stages. Some of these people will relax, fully, into blessed being. Enjoying the garden of future-delight in full community with loved ones, lovers, and perhaps a talking pet or two.

Others will venture into space. Go to sleep for 10,000 years to wake up and start again in an alien world. Others will merge with the machine and have consciousness filtered through silicon instead of a carbon frame.

If you are reading this dear future friend. You are likely more different to my 50-year-old self than my 50-year-old self is from my 5-year-old-self. It won’t even really feel like the same person.

I do wonder.

I do wonder if I am bound for this horizon of an open future extending in all directions. If I will see my children fly off into it. Or if the world stumbles and I die in my mid-eighties, seventies, or sixties from a disease that was just 2 years too early to be cured. Or perhaps I meet my fate in catastrophe on a grand scale.

But if you are reading this dear future friend some things have had to go drastically right.

I can’t know what is in store…

(WWIII, fusion power, intergalactic travel, artificial pandemics, monkeys becoming literate, china invading Taiwan, assassinations, climate crises, climate restoration, walking t-rexes, the ending of governments, the tunneling into the depths of embedded cyber realities, the resurrection of Ted Williams, the end of all disease, drone-based warfare leading to mass surveillance global state, escaping from that state with a saviour ai from another world, and probably much boredom in-between)

What a 50 years it will be future friend. Quite a show to watch. Quite a dragon to face.

I’m not a prayer person. But in this case I’ll make an exception.

Sincerely,

Your Past Friend who just reached the half-century mark for the first time.

I’ve gotten responses entirely in the range of 2 to 6. One of the higher rating members said to me, “I get that AI is coming and we have to use it. But it should never be used for art.”

That is the line in the sand.

Never for art.

Little did she know that it has become more than a small hobby of mine to explore AI as an art-making tool. I’m beyond curious. I’m compelled.

Here is the question as I see it.

Fact A: AI can rocket past human capability in short amount of time. Truly transcendent, jaw-dropping, performance. Look at AlphaGo or the latest Stockfish chess engines. Absolutely effective, super-genius-level, creativity. Moves that win and are also aesthetically astounding.

Fact B: Art changes lives. It brings transcendent experiences. It elicits emotion; love, terror, serenity. It can inspire the most incredible creative thoughts. And it can wake one up to larger realities.

A + B = ???

In my most optimistic moments it equals AI generated art that more effectively and beautifully wakes people up than anything even imagined before. It means essays that are extremely clear and informative. It means music more potent than any hallucinogen. It means a film or VR experience that can utterly and permanently change one’s point of view.

Of course if it’s this powerful, it could also be used to hypnotize, enslave, or destroy human minds. The power is real and it is imminent.

This is a Spiderman moment. The power and the responsibility are in the hands of the artist. There are as I see it four outcomes for AI art that are more or less parallel to the 4 outcomes of an AI future that I’ve covered in another essay.

Outcome 1: AI Art Is the Mind Killer

Either intentionally or unintentionally AI art becomes an addictive and destructive force. It becomes so powerful that one or two exposures to it irrevocably destroys the mind. Either it creates a life long addictive obsession like alcoholism or it induces catatonic or schizophrenic states that persist. This could happen maliciously. One country against another. One political sect against another. Or it could happen unintentionally as the work becomes more and more potent through time. The frog boils. And one day it’s just possible to be exposed to something that is simply too much for the human mind to process or survive intact.

Outcome 2: AI Slop Forever

In this case, AI art never progresses beyond where it lands in 2025-2026. Lots and lots of brain rot that is akin to cigarettes or Cheetos. Kind of fun and addictive. But not really all that mind altering. It just sort of is addictive and gives a cheap short thrill. It drowns out, by sheer volume, all the other expressions.

Outcome 3: AI Art as ‘Nerve Stapling’

If you get this reference, you are my kind of person. In the 90s there was this video game called ‘Alpha Centauri’ which was basically the game Civilization put onto a newly colonized planet in the Alpha Centauri system. Sometimes these space-colonists would get peeved and rise up against their masters. If you were feeling particularly nasty you could do something called ‘Nerve Stapling’ in which you could quell the riots at the cost of longer term productive capacity.

In this case, a government or some other entity, intentionally uses AI productions to keep its population docile. And it would be far more effective than TV or even the hyper-charged algorithms of 2020s YouTube. One could see a situation, similar to AI slop but intentionally employed by some authoritarian faction to make the population do exactly what the regime wants.

Outcome 4: AI Art as a Renaissance of the Human Spirit

There are three not so good outcomes for AI art. And then there is this. What if, like AlphaGo, AI art could transcend and create transcendence. Transcending what? All the things that keep us small in the era of AI. Those things are tribalism, envy, outrage, narcissism, fear… If a work of AI art could so thoroughly channel the beauty at the heart of things it just might be able help humanity to grow up. And in growing up humanity could guide itself and its AI companions to a future of human flourishing.

Perhaps art is in a whole different category than chess or Go. I could see that being the case. Critics will say art has to have a point of view and AI will never have a point of view. Art requires suffering. Art comes only from humanity. Maybe, maybe we are in AI slop forever. That alone would teach us a lot about ourselves.

But I hold out hope. Hope and a sense of mission to make a go at that 4th outcome. Naive perhaps. I can’t help it. The moth is naive when it falls for the candle.

I believe that this Renaissance is a possible outcome of AI in general. Eliminate disease through AI biology. Eliminate poverty through a $250/month standard of living. And finally with AI art see through the self-imposed limitations of early 21st century human life.

What a delicious adventure it would be if we could use our AI tools to untie the hardened knots of the human heart.

Subscribe now

Beginnings are hard. (And AI can’t make them any easier)

It has taken me 3-months, 6-months, maybe 30 years, to get up the courage to write this post. For the last few months I’ve been building a second brain, mastering Claude co-work, and developing all sort of skills that may, some day, with a little luck, allow me to never have to face a human being again.

All of these AI skills will serve me and my family in one sense. It’s likely they will have economic benefits. But I now have a buffer between me and the rest of humanity. I never need to answer an email with my own fingers again.

To the introverted. To the dreamers. We are now, more than ever, able to build an entire world, an entire universe of our own making. Without a single eyeball intruding upon what we built. Your target audience of one person. Pure.

When we are really able to build our own virtual world with no one else to inhabit it what a party it will be for the introverted. But your wisdom will inform you that the party will at some point hollow out into some kind of loneliness, ennui, or insanity.

Is it really a work of art if it touches no heart but your own?

1. Why this, why now

I’m launching this Substack - ‘Meaning in the Age of AI’ because I have a vision that floats around with me. Appearing in-between waking and sleeping. Peeking through when I’m not incessantly busy with my business or my family.

It’s so close now. Wow, just imagine what we could build together if our hearts weren’t so filled with fear, greed, and scar tissue. No disease, no poverty, endless worlds to create and explore. New thrills to be sure. But also new avenues of creativity and intimacy.

What if I built a whole tangible world (perhaps a universe) filled with all my desire, my fantasy, my danger… and I invited you in to share it with me.

Would you come in? I don’t know if I would if I were given such an invite. But I do know that if you did come in you would know me better than anyone I’ve ever known. A singularity of human intimacy.

2. What kind of community are you looking to build here?

I’ve been working with something of a motto for this project:

No one gets left behind.

I mean this economically. We certainly have in our hands the power to end poverty and need globally.

I mean this technologically. So much FOMO! My god, how many messages of missing out are we given daily. How can we transcend this fear of of being left behind?

But more than anything I mean this spiritually. I mean that we, in this community, are not going to forget that introvert, hiding in a cave, building that fortress of solitude that at first titillates and then tortures. People are falling deeply into a pit of their own making. We will not leave them behind.

3. What the Reader can Expect

I will be writing my soul onto these pages. We need to work these questions around in our hearts until we have arrived at crystal clear picture of who we are and what we really want as human beings.

(Because we are entering a time when what we really want, is actually achievable). Before making that wish with the newly-rubbed-out genie you found in the lamp you had better be damn clear about what you want, and be very very careful as to how to word your wishes.

Suffice it to say that wrestling with these questions requires not using AI, or using it minimally, in the writing process on this Substack.

Subscribe now

4. Who am I?

I am a business founder. A father. A philosophy major. An ex-professional dancer. I’ve been on spiritual journey that has taken me through meditation, yoga, zen, and eventually the truth of nonduality. My wanderings have led me here to the intersection of the deeply philosophical and the technological.