The Cognitive Economics of Reading in the Age of AI

In the ever-evolving landscape of AI, we are still reading like it's 1995. I am as guilty as the next person of growing up a millennial, when information was scarce and the libraries and encyclopedias were the best sources of information. As we have advanced more and more with AI at the forefront, the way we consume information hasn't kept pace. I still catch myself feeling guilty when I don't finish a book cover to cover, or when I skim an article instead of reading every word. That guilt? It's a holdover from an era when information was precious and hard to access.

But here's the thing: I came across the idea that perhaps I am not reading wrong. We are just reading everything the same way—and that's the problem.

We've been taught that reading means starting at page one and grinding through to the end. But here's what nobody tells you: that approach made sense when information was scarce. It doesn't anymore.

Though I cannot take credit for the original framework, I will be expanding on it and adding tips & tricks that I have found work best.

So let's dig into the framework

There's three kinds of reading, we just arent naming them

Every time you pick up a book, article, or paper, you're actually doing one of three fundamentally different activities. Recognizing which one you need changes everything.

Awareness Reading:

This is you scrolling through X, it is radar monitoring. You're scrolling headlines, skimming book jackets, scanning abstracts. You're not trying to remember everything—you're mapping the landscape. What arguments exist? What's the field talking about? This is reconnaissance, not study. And it is a way for us to keep up with the AI ecosystem (or any subject for that matter)

Retrieval Reading

All of those AI or SWE books you have on your shelf? That's retrieval reading where you need a specific fact, technique, or answer. When you Google how token architecture works or search for a Python library, you're doing retrieval. It's instrumental. You get what you need and move on. No one is expected to read the Automata Theory book cover to cover (or at least we shouldn't)

Neural Reading (or at least that's what I call it)

Now, this is the hard one. This is reading that literally rewires your brain—forming new neural pathways, building new patterns of thought. Research suggests it burns significant glucose because your brain is working overtime to integrate new structures. You can't do this with everything. Your brain doesn't have the energy. This is the reading we were taught to do for everything we came across.

Why AI Changes the Game (But Differently)

A fear we keep hearing is that AI destroys reading. The reality is more interesting: AI lets you be strategic about which kind of reading you're doing. Take a fact-heavy technical book. That's perfect for retrieval reading with AI assistance. Describe what you already know, what you're trying to learn, and let the AI navigate you to exactly the sections you need. A 10-hour cover-to-cover read becomes 2 hours of targeted learning. But a book like Ethan Mollick's Co-Intelligence—where each chapter builds on the last, where the value isn't in isolated facts but in how ideas interconnect? No shortcuts. You need to read it end to end and let it sink in. The skill isn't using AI to read less. It's using AI to identify which 20% deserves your deep attention.

The Real Problem Isn't "Knowledge Rot"

Some argue we're experiencing knowledge decay because AI just recycles and regurgitates information. That misses what's actually happening. We're not producing less knowledge. We're drowning in it—good information, valuable expertise, important insights. The problem is our filters have collapsed. This is why I have an endless list of bookmarked articles, saved posts and a pile of books sitting on my nightstand. We are in desperate need of better ways to sort through the avalanche. And AI, properly used, can be part of that solution rather than the problem.

Making It Practical

Here's the decision rule: If a book might change how you think about a domain, it gets neural reading. If it's teaching you specific facts or skills, it gets retrieval reading. Everything else gets awareness reading—or just skipped entirely. This isn't about reading less. It's about reading strategically. Awareness reading keeps you current without drowning. Retrieval reading gets you specific knowledge efficiently. And you save your deep reading energy for books that might actually change how you think. As for me, I finally got some color dot stickers to classify my books based on this framework, now the next thing I need, is to develop a framework on 'energy-state matching' meaning, when should we apply each reading type.

Until next time!