When I was younger, I had big dreams about becoming a writer. I wanted a human interest column in a local paper, I wanted to be able to travel the world and talk to interesting people, and I wanted to write books. This ultimately drove me to study journalism.
While the format changed over the years (ie: a newspaper column became a blog, the interviews became startup founders and customer discovery calls), the core desire stayed the same: To share ideas and stories with the world.
Since I've been using my blog to test out ideas in real time, least year, I tried remixing a lot of content about one topic (fractional work) convert into a book form. This took me one week.
I shared that resulting ebook with a literary agent. But they made it clear: Any use of AI was a hard no for traditional publishing. Then I spoke with a New York Times best-selling author who told me that people publish nonfiction books for one of three reasons:
To self-publish and own a niche
To get NYT best-seller status and do high-paid speaking gigs
To write a very short, topical and niche idea and make sure the very best 50 people read a book
When I told him I didn't relate to any of those ideas and just wanted to share a few ideas with a broad audience, he looked at me like I was crazy.
"You know it's considered a good things if even 10,000 people read your book right?"
That's about when I stopped being interested at all in traditional publishing.
A like a lot of things about regular publishing on the Internet. It forces me to crystallize ideas quickly, and it invites others to participate in the real-time discourse.
But there are still a lot of gaps in how I'd prefer to create and consume content online. Things that I think new technology can really help transform in incredible and new ways.
So, since I've abandoned my dream of traditional book publishing, here's my pipe dream for AI-enabled online publishing:
Consider: Customizable formats, styles, and tones for every reader
Think of a book you've always wanted to read but never managed to commit to. I'll take The Power Broker as my classic example here; it's sat on my shelf for at least 10 years, but I've never been able to get past the first 100 pages without distraction. In this case, I want to know the information, but I lack the focus to to sit down and read the book from cover-to-cover as it's designed.
Now imagine a version of "reading" that book, on your terms. I could instruct the mechanism to rework the storylines of The Power Broker and relate to a topic or theme more interesting to me, for instance the startup business world. I could ask it to rework the text to speak to me in analogies or with a more charismatic tone, pulling in modern-day references of New Yorkers in power whose names I may recognize, to bring it closer to home.
This is only one way to remix a book. Consider other remixes too–like turning any book into a graphic novel, a series of listicles, or even TikTok memes that help you remember key ideas. With today’s AI tools, this future isn’t far off.
For those who think this would be a serious offense to book authors who spent years on original content pieces, I'll add that this would be a totally optional way of consuming content (and would ideally involve author consent).
I cant speak for the industry at large, but I'll speak for myself. As a prolific blogger who knows humans can't keep up, I'd be more than happy for people to consume my blog content in different forms that resonate more with them.
Some options for this blog might be:
Highlights reel only (skip the daily digest, only read the summaries each week - I made this version already, powered by AI)
Anecdote-free version (skip all the stuff about my personal life, just read for critical ideas and themes)
Comic book mode (skip the text entirely; just consume a graphic novel version of this blog)
Tweet length posts (turn every 800-word posts into 280 characters, then decide if you want to dig in)
Critical review version (read every blog post from the frame of reference of a character or persona that you invent)
Split big ideas into daily, weekly, or thematic learning journeys
Any long-form content piece that I discover on the Internet should have a "read it later" button that invites me to receive the information either at a later date, split apart over several days, or woven in with additional content and information.
Think about something complicated, like The Bitcoin Whitepaper, or something long, like the 200-page Developer Report written by Electric Capital each year. Rather than commit to sitting down in one sitting and understanding each concept individually, you could ask an AI to "split it apart" for you and help you learn key themes, at a drip cadence on your own terms.
If I were to break apart the Bitcoin whitepaper, I might invite the AI to teach me one conceptual concept each day, in a way that both teaches the core concepts and also invites in some more common explanations that aren't included in the original source material.
Imagine books that update, grow, and adapt over time
We've lived most of our lives with the understanding that books are relatively static reading experiences. You read alone, the text is generally fixed. By contrast, blogs are more dynamic and invite reader comments. Consider an AI-powered publishing platform that's even more fluid. At the author's discretion, the source material itself evolves and adapts in real time. A book on "vibe coding with AI, no-code tools" stays fresh as reader comments introduce new tools and the author incorporates the latest technological changes. My book on fractional works has anecdotes that auto-fresh as people who read along share their own stories. The source materials and links can also evolve fresh as new information emerges.
At the risk of rewriting history, it will obviously be important to index all changes onchain, as proof of timestamp for when certain edits and changes were made. But for people like me who want to contribute to an ongoing conversation around key content clusters, I'd love the idea of "starting a book I never finish" online, one that lives and breathes and grows with me and my audience over many years.
One thing that this instant remixability of content introduces is that it forces all of us as readers to consider more deliberate intent. Here are three questions to ask before consuming content:
Why are you reading?
There are many reasons why we read; before starting a book, an article, or a research paper, we might ask ourselves why we are reading, as this will inform the style and format of the ideal output. If it's to appreciate the art of writing or dissect an author's original wording, it's probably best to stick with the original source material. But if it's simply to consume information, to learn some interesting stories, or to teach yourself how to do something, that's where these remixes get interesting.
How much interaction are you looking for?
There's a big difference between sitting on a beach reading a romance novel, and reading through the latest Jonathan Haidt book, The Anxious Generation. The first is likely a private, self-contained experience for personal enjoyment. The latter invites conversations; you might want to engage with other parents considering screen-free experiences for children or read further research.
How do you learn best?
For moments when you are trying to learn or consume new information, AI-enabled technology invites us to design our own learning environment for anything. Whether you learn best through facts and figures, through historic facts, or personal anecdotes, you can "read" a version of a piece of material that meets you where you're at. This is incredibly powerful.
One of the reasons I am continuing to publish so much content today is so that I have a lot of play doh to remix it tomorrow. If you've seen any cool content remixing tools out there that I should try for this blog, or other content that I consume online, please let me know.
I'm excited to see how digital publishing evolves in the months and years ahead.
Bethany Crystal
Over 500 subscribers
i found the data presented very convincing.
When I was younger, I had big dreams about becoming a writer. I've largely swapped out newspaper & book dreams for blogging, etc.. But there are still a lot of gaps in how I'd prefer to create and consume content online. A few ideas for AI-enabled publishing: https://hardmodefirst.xyz/a-pipe-dream-for-publishing
Really like these ideas!
Excited to chat more with you this week!
The Power Broker is a timeless classic, but for most, it's a snooze fest. I love your idea about remixing this content into something that resonates with the reader so they get the memo.
I spent nearly a decade in book publishing. I’ve seen tech advancements get branded an industry “killer.” First it was blogs, then ebooks, then social media. Yet the industry always survived. Your AI-powered ideas feel like the next evolution: personalizable, living texts that finally close those gaps in creation and consumption. Having lived through these panic cycles, I’m curious how the industry will embrace/hate on AI-tools and what will ultimately become accepted, especially re: reader-driven publishing.
In an industry with as much history and precedent as publishing, it can sometimes feel like fighting a brick wall to imagine any change. What brick do you think will move first?
always comes from the self-publishers and independent authors/publishers first. they're more likely to experiment. for example; 50 Shades of Gray was originally published as Edward and Bella Twilight fan fiction. A PDF was being passed around (if i rememeber correctly) and ereaders were quite popular with the indy crowd. not sure what will pop off first, but whatever it will be it will be a book, story, or author plucked from these experimental forms and brought into the trad publishing world.
Exploring the future of digital publishing, @bethanymarz shares visions for AI remixability in writing. Content should adapt to individual learning styles and can evolve over time based on reader interactions. Embrace customization and fluidity in how we consume knowledge!