This Easter weekend, Andrej Karpathy’s post on LLM-powered knowledge bases went viral — and for good reason. He described a self-maintaining, self-correcting knowledge system built entirely by AI, and then shared the concept as a natural-language “idea file” instead of code, letting anyone’s agent build their own version. For readers who collect highlights and annotations, this is bigger than it sounds: you’re already closer to a living knowledge base than anyone else — you just need the right architecture to unlock it. That architecture has two dimensions: vertical structures that capture each book’s argumentation and logic, and a horizontal layer that connects insights across books, guided by what you’re actually trying to learn or solve.

Books have always done two things for society: they preserve what we know, and they help us create what we don’t yet know. Now AI can produce, search, and recombine text at a scale no human can match. But a closer look at what AI has actually achieved — and what it hasn’t — reveals why books remain irreplaceable.

Any book can be summarized by AI in seconds. So why would anyone still read the whole thing? The answer requires rethinking what books actually are. A book is a mediator — one possible bridge between an insight and your mind. Sometimes it’s the best bridge. Sometimes AI is. And sometimes the shortcut you think saves time quietly destroys the thing you were trying to acquire.

Choosing a book is not like picking something to watch tonight. It is a serious commitment of hours, sometimes days — and the opportunity cost of every other book you won’t be reading during that time. That’s why readers have always needed to trust their sources. But here’s the counterintuitive part: that trust has never really been about books. It’s been about finding people who have already found the books you were looking for.

When anyone can write a book, more people will. More topics, more perspectives, more niche knowledge that was previously locked inside practitioners’ heads. But what about the top — the works that change how we think? AI has had access to virtually all of human knowledge for three years now. It has not produced a single original explanation for any unsolved problem. What does that tell us about the future of books?

For five thousand years, books have done something no other technology has managed: let you enter another person’s mind and reconstruct their thinking in your own. In 2026, the most powerful technology we’ve ever built communicates the same way — as a chain of words on a screen. But in a world racing toward interfaces that bypass text entirely, will the book survive?

There are two ways to know a book. One changes how you think. The other gives you answers to concrete problems. Most people drift toward one or the other. Combining the two is where you stop consuming knowledge — and start creating it.

There are two ways to know a book. One changes how you think. The other gives you answers to concrete problems. Most people drift toward one or the other. Combining the two is where you stop consuming knowledge — and start creating it.

A new AI tool exploded this week – 100,000 GitHub stars, Mac Minis selling out, three name changes in seven days. But unlike most tech hype, OpenClaw might actually deliver on a promise we’ve been waiting for: AI that gives you time back instead of stealing it.

My books should work for me. Not sit passively in highlights waiting to be searched, but actively remind me of what I’ve read in the right moment, suggest connections I haven’t seen, and help me apply what I’ve learned to real problems. Here’s how I built that system with Claude Code.

What if an AI could search every book you’ve ever read and find the exact insight you need — condense years of reading on a topic, or compare how different authors approach the same question? I’m experimenting with Claude Code to make this real: teaching an LLM to navigate my entire Kindle highlight collection and become an intellectual sparring partner.

Your reading history contains years of insights, but they’re scattered across dozens of books. When you need to pull together everything you know about a topic, those highlights feel unreachable.
We’re building new cross-library AI workflows for DeepRead that transform this scattered knowledge into structured outputs. Through guided conversations, the AI helps you generate topic memos or mind maps that synthesize insights from across your entire library. The key difference from traditional AI chat: the system has persistent access to your complete reading history, not just what you upload. Every interaction creates something tangible and makes your knowledge system more valuable.
We’re developing these workflows now and looking for early users to test features and share ideas.