A digital reading specialist’s take on the most important shift in how we consume knowledge
Most people think they are reading more than ever. They are finishing more books, running through more summaries, and ticking off longer reading lists. But if you asked them what they actually retained six months later, the silence would be telling.
We have confused volume with depth. And AI just made that confusion a whole lot easier to maintain.
I have spent years inside the world of digital reading technology. E-readers, reading apps, annotation systems, knowledge workflows. I have watched how people build habits around books, how they highlight without ever reviewing, how they chase the 100-books-a-year milestone like it is a personality trait. And now I am watching something new happen. AI has entered the reading experience, and it is changing not just how we read, but what reading even means.
This article is not here to tell you AI is bad for reading. It is here to tell you that if you do not understand what is shifting, you will optimise yourself straight out of the thing that makes reading valuable in the first place.
The Way We Actually Read Now 📱
Before we talk about AI, we need to be honest about how most people were already reading digitally, and the picture was not pretty.
The average digital reader reads across multiple apps, multiple devices, with notifications sitting one swipe away. They highlight compulsively but never build a system for those highlights. They buy books they never open, open books they never finish, and finish books they never think about again. This is not a character flaw. It is a design problem. The digital reading environment was never built for retention. It was built for acquisition.
The people who read well digitally have built intentional systems. A dedicated e-reader, not a phone. A curated app stack that separates reading from distraction. A workflow that connects what they read to how they think. That kind of setup does not happen by accident. It requires understanding what the tools are actually doing to your attention.
There is a reason the Kobo versus Kindle conversation is never really about specs. It is about philosophy. One ecosystem nudges you toward purchasing. The other nudges you toward reading. The difference feels small until you are three years in and notice what your habit actually looks like.
Where AI Entered the Room 🤖
AI did not knock. It just showed up inside the tools you were already using.
It started with summarisers. You could paste any non-fiction book into an AI tool and get the key ideas in four minutes. This felt like a superpower. And honestly, for certain use cases, it still is. If you are curating a reading list and want to decide whether a 400-page book deserves your full attention, a smart summary is a legitimate tool.
But something else started happening. People stopped asking whether they should read the full book. They just did not. Why would you spend eight hours on something you could process in four minutes? The answer, which most AI evangelists skip past, is that the eight hours is not just about the information. It is about the thinking that happens with the information.
Then came tools like NotebookLM. This is where the conversation gets genuinely interesting, because NotebookLM is not just a summariser. It is a thinking partner. You upload your sources, your highlights, your notes, and you build a personal knowledge base that you can have a conversation with. For a serious non-fiction reader, this changes everything. It turns passive consumption into active dialogue. That is a real upgrade.
The question is whether people are using it that way, or whether they are just uploading other people’s books and asking AI to do their thinking for them.
What We Are Quietly Losing 🧠
Here is the take that makes people uncomfortable.
Non-fiction books have been getting worse for a decade. Not all of them, but enough that the pattern is undeniable. Publishers know that most buyers will not finish the book. So authors pad ideas that could fit into a long essay into 300 pages of repeated examples, repackaged frameworks and motivational filler. The reader feels busy. They do not feel changed.
AI did not create this problem. But it is accelerating the dynamic in a dangerous direction.
When you summarise a book that was already mostly padding, you are not losing much. But the reading culture that normalises summarisation does not stop there. It starts applying the same logic to books that are genuinely dense, genuinely worth the struggle. Books where the difficulty is the point. Books where the slow, effortful processing is where the real learning happens.
There is also the highlight problem, and it goes deeper than most people realise. Research consistently shows that highlighting while reading creates an illusion of learning. You feel like you are capturing the idea. But the act of highlighting actually reduces the effort your brain makes to encode the information. You outsourced the memory work to the highlight, and then you never went back to the highlight anyway. AI-assisted reading is doing something structurally similar at a much larger scale. It is creating the feeling of knowing without the friction that produces actual knowing.
This matters enormously for anyone who reads to think, not just to be informed.
A Smarter Hybrid Workflow ⚙️
The answer is not to throw AI out of your reading life. That is performative, and it misses the point. The answer is to understand precisely where AI helps and where it quietly replaces something it cannot actually replicate.
Here is the distinction that changed how I work. AI is exceptional at the surface layer of a text. It can identify main arguments, extract key examples, map the structure of a book, surface contradictions between sources, and organise your existing notes into useful formats. These are real jobs that used to take significant time, and AI does them well.
What AI cannot do is sit with an idea long enough to let it reshape how you see something. That requires time, discomfort, your own context, and your own life experience pushing back against the text. That is the process. That is what you are there for.
A practical workflow that actually works looks like this. Use AI to qualify your reading list. Not every book deserves full engagement, and a smart summary helps you decide. For the books that do make the cut, read them slowly and without shortcuts. Highlight sparingly and with intention. After reading, use a tool like NotebookLM not to extract what the book said, but to interrogate it. Ask it questions. Push it against other things you have read. Let the AI help you think, not think for you. Then write something, even briefly. A daily reading log, a short reflection, a note to yourself about what shifted. The writing is where the understanding becomes durable.
This workflow is not about doing more. It is about being honest that reading without processing is just exposure, and exposure without retention is a very expensive way to feel productive.
The Readers Who Will Thrive 🎯
AI is not making readers obsolete. It is making a certain kind of reader obsolete. The reader who measures success in books per year, who treats knowledge as a collection rather than a foundation, and who confuses familiarity with a book’s arguments for understanding them. That version of reading was always a bit hollow. AI is just making the hollowness visible.
The readers who will thrive in an AI-assisted world are the ones who know exactly what the human part of reading is for. They use AI to handle the surface so they can go deeper with what remains. They read fewer books and get more from each one. They build systems, not streaks.
The question worth asking yourself is not how many books you read last year. It is how many of them changed how you think.
If the answer makes you uncomfortable, that discomfort is exactly where your reading practice needs to begin.
What has AI changed about the way you read? Drop your experience in the comments. I am genuinely curious whether others are feeling this shift or whether it is still flying under the radar. 💬
Leave a Reply