Books are linear because paper is linear. Understanding isn't. This is the long-form case for where TruNaut is headed, held for people who've already opted into wanting the fuller story.
Direction, not feature. Everything past "What exists now" below is unbuilt. None of it should be read as a promise or a timeline — it's the reasoning behind the roadmap, not the roadmap.
You highlight a passage. The highlight opens a conversation. The conversation becomes a note in your own vault. That loop is real, and it's the whole pitch for now — everything else on this page is direction, held in reserve until it's true.
“Books are linear because paper is linear. Understanding isn't.”
You read a chapter on entropy in a physics text, a paragraph on it in a philosophy essay, a tossed-off line about it in a novel eight months later — and your brain quietly files all three under the same concept. No reading tool does. Every highlight you've ever made lives in the book it came from and nowhere else.
The tools got good at capture — clip it, tag it, export it — and never touched the harder problem underneath: tracking what you actually come to know, as it accumulates across everything you read, watch, and think about.
TruNaut is built to close that gap. Slowly, and honestly — starting with the part that's true today.
Each step earns the next. A canvas has to exist before it's worth clustering handwriting into knowledge; capture has to work across formats before generation has anything real to draw from.
A reader isn't where the thinking happens for everyone — some people think by writing, not chatting. The next step is one infinite page per document, where you handwrite against the source directly, and the system quietly clusters what you've written into atomic units of knowledge, the same way a highlight does today.
A lecture, a voice memo, a thought recorded on a walk — captured and summarized with the same rigor as a highlight, written into the same vault, tagged into the same system. Reading was never the actual unit of value; thinking was. Audio is just another channel it arrives through.
A listening companion for audiobooks and podcasts, doing for your ears what the chat does for your eyes — and a browser layer that meets you on YouTube and the open web, so the system isn't limited to what you opened inside the app. Capture stops being tied to a format and starts being tied to you.
A system that knows what you've read, what you struggled with, what you've already connected — and uses that to hand you the next right thing. Not a static reading list. A curriculum that rewrites itself around a knowledge state only you have. This is the point where the product stops being a better way to read and becomes a way to learn that no static course or book can match, because it's built from your actual history, not a syllabus written for a stranger.
Not for sharing passages, which everyone already does, but for sharing thinking. Someone else's insight on a page becomes a starting point for yours, instead of a screenshot that dies in a group chat.
One thing has to be built first and never redone: a shared way of representing knowledge that every layer above reads from and writes to.
Raw signals — highlights, notes, conversations, re-reads, what you abandoned — stored as events, never collapsed into a score too early. A concept map that lets "entropy" in a textbook and "entropy" in a conversation six weeks later resolve to the same node. And an inference layer on top that estimates what you actually know, and gets better as more signal arrives.
This runs silently from the first version. It has no screen of its own. Its only job is making sure that someone who's been using the reader for a year arrives at every future version of TruNaut with a head start instead of a blank slate.
“Time is the entire advantage. There's no shortcut to it, which is exactly why it's defensible.”Why this is worth building at all
The moat isn't other people. It isn't a network effect where the product gets better because more users join. It's depth — the longer you use it, the more accurately it reflects what's actually in your head, and that record is worthless to anyone else and impossible for a competitor to hand you on day one.
A system that knows what you know — from every book, article, lecture, and conversation you've fed it — and uses that to decide what you should encounter next. Reading is the first thing it learns from. It won't be the last.
The reading experience must never feel secondary to the AI features, at any layer we add.
Every layer connects back to a highlight, a passage, a captured thought. Nothing floats free.
The value compounds through stored discussions — this is the entire premise the knowledge layer is built on.
Generation hands you the next right thing — it doesn't decide what you think about it.
This page is direction. The app's pitch is only ever what's actually built — see Features for that.
I'm building TruNaut alone, nights and weekends, as my own way into EdTech after years in a very different kind of service. Everything shipped so far got built because I needed it for my own reading, not because it looked good on a roadmap — and everything on this page stays unbuilt until the same is true of it.
— Sid, Indian Army officer turned solo builderOne email when TruNaut lands, nothing else.