Sivaji Sahoo

maybe I wasn't ready for books alone

I had the words but not the thinking behind them.

I kept noticing the same problem. I would finish an article feeling like I understood it, confident I could use the idea. But when I tried to apply it to a different context, I couldn’t do it.

This would happen with ideas that seemed clear while reading. Why index funds beat active funds, why network effects create monopolies. I would follow the argument sentence by sentence. But later, when I tried to transfer the logic to a new situation, the structure would fall apart.

At first I assumed this was a memory problem. Maybe I was reading too passively. So I tried highlighting key points, wrote summaries, even explained the ideas to myself out loud. It helped a little, but not much.

That got me thinking: what if I just had not developed the skill yet?

Now, for a moment, think about the reading advice we hear constantly. Engage critically. Question assumptions. Test the logic. All of it sounds reasonable, but it assumes you already know how to do those things. It is like telling someone to just be creative without teaching them how.

I did not know what it meant to engage critically. I thought highlighting and summarizing were enough. But those are just recording activities. They capture what the author said.

The real skill I was missing is adversarial reading. Playing devil's advocate with every claim. When would this fail? What assumptions are hidden here? What would need to be true for the opposite to hold? That is not something I learned to do automatically. And without it, books felt one-sided to me.

Later, I started using language models to interrogate what I was reading. I would copy a sentence from a book and ask Claude: "What assumptions does this rely on?" or "When would this fail?"

The first time I tried this, I went back to a claim I had read: "Articulation is enough to reveal most errors." It sounded right. But I pasted it into Claude and asked, "When would articulation not reveal the error?" The answer pointed to cases I had not even considered. Then I understood it. Not just the words, but the conditions where the author’s idea held and where it did not. That was the first time I felt like I had wrestled with an idea I had only read before.

What the LLM did was make explicit a cognitive process I had not figured out how to do on my own. It externalized the adversarial thinking that experts do internally. It showed me what it looks like to actually challenge an idea, not just passively agree with it.

This is not a permanent solution. The goal is to practice enough that the questioning becomes automatic. But right now, I am not there yet. And that is fine. All the reading advice assumes you have already crossed this threshold. It tells you to engage critically without teaching you what critical engagement actually requires. It is advice for people who have already developed the skill.

Maybe that is why so many people read widely but still feel like they are not learning deeply. They follow the steps like highlighting, summarizing, reviewing, but miss what matters the most. The core principle is adversarial : treating every claim as something that needs to survive interrogation before you accept it.

I still highlight things. I still take notes. But now, whenever something sounds plausible, I stop and ask it questions. I try to make the idea defend itself. And the ones that survive that process stick far more deeply than anything I used to passively agree with.

Books have always required something from the reader. For a long time, I thought I was bringing that. I was not. I was just listening carefully.

Maybe books were always enough. But I was not ready for them yet.