Your Brain in the AI Age: How to Keep Your Memory Sharp

Nov 21, 2025 • 10 Min Read

Source:
Freepik

Recently, my sister asked me for good fiction book recommendations.

Despite reading every week and having many books that I loved, I struggled to think of one to recommend. Yet, when I cast my mind back to my pre-Kindle days, I can easily remember many, many books to recommend.

It made me wonder: Is there something about digital reading that makes it harder to remember what we consume? Or was I imagining it?

Given my bias towards evidence over anecdote, I went looking for what the research actually says.

Along the way, I fell down a rabbit hole that stretches from how we read on screens to how AI changes what we remember to what neuroscientists are learning about how memory really works.

The short version is that my Kindle experience is not unusual, but the story is more nuanced than “screens are bad, paper is good”.

It is a story about attention, meaning, and how we choose to utilise the digital tools around us.

Does format matter? What the research says about print vs digital

Researchers have been comparing paper and digital reading for years, and the findings are mixed – which is precisely what you would expect in a messy, real-world phenomenon.

A 2018 meta-analysis of studies comparing paper-based and digital reading found that, on average, people understood and remembered paper texts slightly better than digital ones, particularly when reading under time pressure and when the texts were informational rather than narrative in nature.

The authors suggested that screens often nudge us towards skimming, multitasking and ‘good enough’ reading, which weakens comprehension.

More recent work complicates the picture.

A 2024 meta-analysis reported no significant overall difference in reading comprehension between digital and paper formats once factors such as text length, task type and how familiar people were with the medium were controlled. In some contexts, digital did just as well; in others, paper still held a small edge.

Therefore, weaker recall for digital texts is not simply a matter of pixels. It reflects how we tend to use digital media.

Screens are often:

Surrounded by distractions. For example, notifications, tabs, messages, etc.
Used in short bursts rather than extended stretches.
Associated with scanning and jumping rather than slow, linear reading.

Paper books, in contrast, come with built-in cues that support memory. For example, the physical feel of the book, the repetition of seeing the book’s cover, a sense of where you are in the chapter, and a more precise mental map of the text. There is also less temptation to flick away.

In other words, it is not that a Kindle or e-reader cannot support deep memory. It is that you and I are more likely to read in ways that undermine deep encoding when we are on a digital device.

To understand why that matters, we need to take a look under the hood at how our memory works.

Your memory is not a filing cabinet

Popular debates about memory often treat forgetting as failure. If you cannot recall an author’s name or the exact statistics from a report (or, in my case, a good fiction book to recommend), something must be wrong.

Neuroscientist Charan Ranganath argued that this is the wrong way to think about memory. In his book Why We Remember and in a widely discussed 2024 New York Times op-ed on President Biden’s memory, he argues that forgetting is normal and often adaptive.

Memory has evolved not to store everything, but to help us make good predictions and decisions in a changing world.

So, what is going on in your brain when you read and try to remember? At a simplified level, there are three key pieces to the story.

1. Short-term and working memory

Short-term memory is the ability to hold a small amount of information in mind for a brief period, seconds to minutes. Working memory is the more active version of this; it is the mental workbench where you juggle and manipulate information, such as holding a sentence in mind while you determine its meaning. This depends heavily on networks in the prefrontal cortex.

Short-term and working memory are limited.

You cannot deeply encode what you never fully pay attention to. Suppose you are reading while glancing at an email or half-listening to a podcast. In that case, your working memory is constantly disrupted, and the likelihood of transferring that material into long-term storage drops dramatically.

2. Long-term memory and the hippocampus

Long-term memory stores information over days, months and years.

A structure deep in the brain, the hippocampus, plays a central role in forming new episodic memories (events and experiences) and in binding together the different elements of an experience – sights, sounds, ideas, and feelings – into a coherent narrative.

Over time, through a process known as consolidation, those memories become more widely represented in the neocortex, the outer layer of the brain. Sleep, repetition and revisiting ideas all support this process.

3. Synaptic plasticity and changing your wiring

At the cellular level, new learning is associated with changes in the strength of synaptic connections, a phenomenon known as synaptic plasticity.

When you learn something new, your brain cells actually change how they communicate with each other; that is, they strengthen or weaken their connections (i.e the synapses that connect neural pathways).

When you practice or repeat something, these connections can become permanently stronger. This lasting strengthening is one of the main ways your brain builds memories and learns new skills.

The key point for the rest of us is that how we engage with information matters.

Deep, focused processing – linking ideas, generating examples, questioning, teaching others – sends a strong ‘this matters’ signal to the hippocampus and strengthens those neural pathways. Shallow, distracted exposure does not.

Which brings us back to screens, AI and why some things do not stick.

Outsourcing your brain: the Google effect and AI

Source: Macrovector from Freepik

Humans have always offloaded memory onto tools: clay tablets, notebooks, calendars and search engines.

The question now is what happens when we outsource more and more of our thinking to digital systems and AI.

One of the earliest warnings emerged long before the advent of generative AI.

In a 2011 paper in Science, Professor of Psychology Betsy Sparrow and colleagues showed that when people believed information would be stored on a computer, they were less likely to recall the information itself and more likely to remember where to find it.

This so-called “Google effect” or digital amnesia does not mean our brains have stopped working. It means we are reorganising what we remember: from content to location.

As AI has become more capable, researchers have begun to notice a related pattern in skill development. In an article originally published in The Conversation, Researcher Tapani Rinta-Kahila described how an accounting firm discovered that, after years of relying on smart software to handle fixed-asset accounting, their staff struggled to perform the work manually once the software was removed. Critical expertise had quietly eroded.

The same article introduced the notion of “automation complacency”: the understandable but risky assumption that if a system is usually accurate, you no longer need to monitor it or maintain your own competence.

A 2024 review in an educational technology journal went further, arguing that overreliance on generative AI for tasks such as writing and problem solving may damage memory retention and critical thinking over time, precisely because it removes the need for active cognitive effort.

Active learning – effortful engagement with ideas – is what drives consolidation and durable memory.

Leaders are also starting to worry about this in the workplace. While most executives see significant productivity gains from AI tools, some are concerned about skill atrophy and loss of foundational knowledge among employees.

There is a genuine tension here.

Offloading some tasks to technology frees up mental bandwidth. That can be good for higher-order thinking. At the same time, if we offload too much, too early, we risk weakening exactly the neural circuits and habits of mind we value.

The question is not “should we use AI?” but “what do we still need to know how to do ourselves, and how will we stay in practice?”

Attention as the gateway: why context matters

Psychologists describe attention as the gatekeeper for memory.

If your attention is scattered across many inputs, very little makes it through the gate with enough force to trigger strong memory encoding.

The combination of always-on devices, constant notifications and AI tools that produce instant answers can keep that gate in a perpetual half-open, half-closed state.

This does not mean we need to abandon our devices. It does mean we should be more deliberate about when we seek deep understanding and how we shape the conditions for it.

Building better memories in a digital, AI-rich world

Source: Freepik

What might this look like in practice?

Here are evidence-based strategies to help you remember more of what matters, whether you are reading on paper, on a screen or working alongside AI.

1. Read with a purpose
Before you open the book or article, ask yourself: What do I want to take away from this?

Setting a clear intention helps the brain decide that this material is worth encoding. Memory research indicates that meaning and relevance significantly influence what we retain.

2. Make digital reading more “paper-like”
The problem is not the Kindle or e-reader; it is passive scrolling.

Get deliberate as you read by:

Creating conditions for focus: one device, notifications off, dedicated reading time.
Annotating as you go. Highlight sparingly and write short notes that explain why an idea matters.
At the end of each chapter, close the device and summarise the key ideas in your own words. This active recall practice is one of the most potent techniques for strengthening memory traces.

3. Use spacing and retrieval, not just exposure
Memory consolidation benefits from spacing – revisiting material over time – and retrieval practice, where you test yourself without looking at the text.

Instead of rereading, try:

Jotting down what you remember the next day, then checking against the text.
Creating a summary or model that you could explain to a colleague.
Using simple digital flashcards for key concepts or definitions.

4. Make it meaningful and emotional
We remember what is connected to existing knowledge and what carries emotional weight. Context, meaning and emotion all strengthen memory networks.

As you read:

Link ideas to real problems you are tackling at work or what you already know.
Notice where you agree, disagree or feel surprised.
Capture one ‘So what?’ after each reading session: what might you do differently because of this?

5. Treat AI as a thinking partner, not a substitute
AI can be a powerful ally for learning if you stay in the driver’s seat.

Consider:

Drafting your own summary of a chapter, then asking AI to critique, extend or challenge it.
Asking AI to quiz you on key concepts, rather than asking it to summarise the book for you.
Periodically doing tasks “AI-free” to keep your skills sharp and stay competent without autopilot or smart software.

The discipline is to let AI accelerate the grunt work, without handing over the parts of thinking you actually want to own.

6. Design for attention, including time in nature
Given the link between attention and memory, even small design choices make a difference. For example:

Create device-free blocks for deep reading, even if they are only 25 minutes
Where possible, pair reading or reflection with time outdoors – a short walk in a park can restore attention and boost working memory performance. Studies have shown that spending time in nature can improve your attention and memory.
Reduce multitasking. Every context switch is a small tax on your working memory.

7. Protect the brain systems that support memory
Finally, the unglamorous basics matter: regular sleep, physical activity and stress management all support the hippocampus and prefrontal networks that underlie memory.

You cannot out-read or out-AI a chronically exhausted brain.

Remembering what matters

When I interrogated my own ‘Kindle amnesia’, the answer turned out not to be a simple indictment of screens.

It was a reminder to be more deliberate. When using a paper-based book, I was more likely to slow down, underline, pause and reflect. On a device, I was more likely to skim between other tasks, trusting that I could always search for the idea again later. My brain responded accordingly. It remembered what I had signalled, through my behaviour, was important.

In a digital, AI-enabled world, setting that signal requires conscious control. We can choose to let technology hollow out our skills and memories, or we can design a different relationship in which digital tools expand our capacity without eroding our competence.

The difference lies less in the devices we use and more in the habits and environments we build around them.

In the end, the real question is not “Can you remember everything you read?” It is “Are you remembering and practising the things that matter for the work and life you want to lead?”

Republished with courtesy from michellegibbings.com.