The Misarchitects of AI
By Briar Harvey
Wanna Try Something New With Me?
I have dreamed of being a published author since before I could write.
Truly. All I have ever wanted in the world is a stack of books with my name on the cover.
But I’ve never written a book.
That’s because traditional publishing is broken.
As a reader, I buy multiple copies of the same book—print, digital, audio. These days, I use digital to decide. But I read in print, with a pen or a highlighter. And if it’s dense or something I want to remember, I’ll listen at the same time.
Somehow, despite the complete and utter chokehold that exists in publishing, no one has figured out how to bundle all the formats (PDF, epub, audio, print), in a way that you ACTUALLY OWN, while fairly compensating the author.
DRM tried to solve piracy, but it actually broke everything else.
Kindle locks you into their ecosystem.
And you don’t own that Audible book.
The Math Ain't Mathing
In many ways, the enshittification of publishing was predictable.
Despite all the technological differences, the industry is still basically the same. Enormous advances for an elite few, and no effort for anyone else, despite lipservice to the contrary.
Elizabeth Gilbert got 200k to go figure herself out and write Eat, Pray, Love. The advance for All the Way to the River was not disclosed, but based on its marketing and PR package, it was probably high six, maybe even seven figures.
Meanwhile, most authors get no advances. And I’ve known more than a few authors who actually lose money publishing, either from using an overpriced vanity press, to weird KDP requirements, to the kinds of bullshit that made me say…nah. It’s not worth it.
The math is definitely not mathing.
In fact, relative to most other commodities, (and a book IS a commodity), the price of a book hasn’t changed much in the last 25 years.
Reading Itself Is Changing...
The other side of the equation is that nobody reads anymore.
According to research conducted by the National Endowment of the Arts in 2022, fewer than half of Americans read a book for pleasure that year. Our attention spans are on the decline. And we make hundreds, if not thousands of micro decisions about content a day.
Plus, with the advance of Generative AI, the way we consume information is fundamentally changing.
Why read a book when you can ask Chat for a summary? It doesn’t even matter if it’s wrong (and I can’t stress this enough—it WILL BE, up to 40% of the time, regardless of model), that’s what you’re expected to do now, right?
At this point, the ONLY acceptable reason to read a book is in pursuit of depth.
In other words, I’m not just pushing back on what publishing has become, I’m pushing back on what it’s about to become, too.
I’m trying something different—because a book that needs updating in real time requires innovation.
The Misarchitects of AI is a living document, and I’m writing it in real time.
The speed at which AI is changing (and the resulting collapse that sits on top of us) can't wait for a traditional publishing cycle. But if anyone can figure out a way to make content work at scale, it's me.
I’ll walk you through what I see coming, and how to get out from under it now.
PRESALE BUYERS RECEIVE:
Everything I'm talking about in the book is publicly available information, and I don't think I'm the only person to put the pieces together in this way.
But I seem to be one of the few people willing to say the quiet part out loud.
If our stance on AI doesn't change soon, we will be responsible for the consequences.
PURCHASE BY FEBRUARY 15th FOR LOWEST PRICE!!!
Traditional Publishing
How it works now:
$
70+
You pay for the same content three times anyway
PRINCIPLED
PUBLISHING
How it COULD work:
$
150 +S/H
BONUS (for teams):
Buy 10 OR MORE copies for you and your team, and I'll personally walk you through your AI Dependency Audit, PLUS give you my personalized recommendations.
FAQ
Then you get a book documenting why I was wrong, how the pattern evolved differently, and what that teaches us about intelligence systems. The methodology is still valuable.
Plus: If I'm wrong, updates continue longer. More content, not less.
Then you have a case study in pattern recognition, a dependency audit that applies to ANY vendor, and preparation frameworks that work for the next paradigm shift.
You shouldn't trust anyone's prediction blindly.
That's why I'm showing you the research (Apple's GSM-Symbolic study), the financials (OpenAI's unsustainable burn rate), the personnel patterns (researcher exodus), and the intelligence methodology (how I connected these dots).
Read the preview chapters in January. Decide if the analysis holds up. The rolling updates let you validate in real-time.

