I am six years old, sitting in music school and trying to reach my parents using the payphone in the hallway (not entirely sure those things even exist anymore). I know every important number by heart: my mum's, my dad's, my grandmother's, and a few extras, just in case.

Then mobile phones arrive, and suddenly there is no need to remember phone numbers anymore. Great news, supposedly. More space in the brain for other things. Then comes Google: no need to memorise dates, capitals, or random facts.

And then comes the real revolution: ChatGPT.

Unlike Google, it doesn't just help you find information. It gives you the answer. Instantly. And it is incredibly convenient. But it also raises an uncomfortable question.

For years, technology has gradually reduced the amount of information we need to store in our heads. Now, for the first time, it is starting to reduce some of the mental effort involved in processing that information too.

So what happens to the parts of our brains that used to do that work? Are we becoming worse at remembering? Less creative? Less capable of deep thinking? Or is this simply another technological shift — one our brains are adapting to, just as they always have?

It would be easy to blame ChatGPT for everything. The only problem? People have been worrying about technology changing the way we process information for years.

If we had to pick a landmark moment, it would probably be 2011. That was when researchers published the now-famous paper Google Effects on Memory: Cognitive Consequences of Having Information at Our Fingertips, exploring what later became known as the "google effect" — or, more dramatically, "digital amnesia".

The experiment itself was simple: participants were asked to type trivia facts into a computer. Some were told the information would be saved. Others were told it would be erased.

The result? People who believed the information would be stored somewhere else were significantly less likely to remember it themselves. Instead, they remembered where to find it.

A few years later, a 2015 study commissioned by Kaspersky found something similar. Surveying more than 6,000 adults, researchers discovered that 91% of people used the internet as a kind of external memory bank, relying on their devices to remember information for them.

And study after study has pointed in roughly the same direction: the easier information becomes to access, the less motivated our brains are to store it. Researchers often explain this through a concept known as transactive memory. Instead of remembering the information itself, we increasingly tend to remember where to find it.

Which is worth keeping in mind. Because ChatGPT isn't creating this trend. It may simply be taking it several steps further.

When people talk about Google, ChatGPT and what they are doing to our brains, the conversation almost always starts with memory.

And, to be fair, that makes sense. We clearly remember less than we used to. But I am not entirely convinced that my quality of life depends on remembering a specific historical date or the surname of someone I can identify in five seconds online. Besides, we just proved, we started outsourcing parts of our memory long before ChatGPT came along.

The thing is, memory is only the visible part of the iceberg. There are three other — and arguably more important — questions to consider:

In other words, the change isn’t only about what we remember. It is also about how much thinking we do ourselves. This brings us to my favourite part: experiments.

In 2025, researchers at MIT decided to look at what happens inside our brains when AI joins the writing process. Participants were divided into three groups. One wrote essays using ChatGPT. Another used traditional search engines. A third had to rely entirely on their own knowledge and reasoning.

You can probably guess the result. The ChatGPT group showed lower levels of brain engagement during the writing process than both of the other groups. Even more interestingly, many participants struggled to recall parts of essays they had submitted only a short time earlier.

One survey of 1,000 internet users found that only 8% regularly verified AI-generated answers, despite almost 75% saying they had personally encountered significant mistakes made by AI systems.

In other words, even when we know AI can get things wrong, many of us stop checking. And that is where researchers start to worry. Critical thinking isn't about knowing every answer. It is about questioning information, evaluating sources, and spotting mistakes. If AI is making us less likely to do that (even when we know it can be wrong) the consequences could extend far beyond memory.

Interesting fact: this article turned out to be harder to write than most. The reason? ChatGPT seemed surprisingly reluctant to help with anything too critical about itself. Every time a paragraph started with a phrase like "ChatGPT may affect...", the conversation kept drifting back towards the benefits. In the end, I found myself doing something rather ironic: taking a step back and returning to Google instead.

Let's be honest: we all know how tempting it is. You sit down in front of a blank page, your brain refuses to cooperate, and instead of wrestling with the problem, you open ChatGPT. A few seconds later, you have a list of ideas, angles and suggestions. It feels productive — and often it is.

What we sometimes forget, however, is where those ideas come from. ChatGPT doesn't create in the same way humans do. It generates responses based on patterns it has already seen.

And that may come with a trade-off. A 2025 study found that while ChatGPT helped people generate ideas more quickly, it also reduced the diversity of those ideas. Participants using AI assistance tended to converge on similar solutions rather than producing truly original ones.

Important note: there is currently no evidence that ChatGPT causes cognitive decline, lowers IQ, causes dementia, or damages the brain. None.

But there is a reason researchers are paying attention to the question.

We know that certain cognitive abilities — including processing speed, working memory, and the ability to learn completely new information — tend to decline gradually with age. We also know that people who remain mentally active throughout life tend to maintain stronger cognitive function for longer.

And this isn’t just about staying sharp in old age.

One of the most consistent findings in ageing research is the concept of cognitive reserve. People who build it through education, lifelong learning and regular mental challenges tend to show a lower risk of dementia symptoms later in life. Researchers describe cognitive reserve as the brain's ability to adapt to age-related changes and continue functioning effectively despite them.

This is where AI introduces a fascinating question. ChatGPT is a form of cognitive offloading. Instead of doing the work ourselves, we hand part of it to a tool. Sometimes that means remembering less. Sometimes it means writing less. Sometimes it means analysing less.

And right now, nobody really knows what the long-term consequences of that will be. ChatGPT is still primarily a young person's tool, with the largest share of users falling between the ages of 18 and 34. So what happens when the first generation that grew up outsourcing parts of its thinking to AI reaches 50, 60 or 70?

The honest answer is that we don't know yet. The technology is simply too new.

I want to end on a positive note. Because what is the point of blaming ChatGPT for everything? AI is part of our reality now, and it isn't going anywhere.

We know it creates cognitive offloading. What we also know is that this offloading should probably be balanced by cognitive engagement somewhere else. Research tracking people across decades has found that education, lifelong learning and intellectually stimulating activities are associated with greater cognitive reserve later in life — the brain's ability to remain resilient as we age.

The good news? The brain isn't particularly picky about where that stimulation comes from. It can come from learning a musical instrument. Picking up a new language. Reading regularly, writing, dancing, taking a course, learning photography, joining a chess club, or finally trying that hobby you have been postponing for years.

The list is long.

So perhaps the answer isn't to fear ChatGPT. Perhaps it is to use it for what it does best — saving time — and then spend some of that time doing something that keeps your own brain active, challenged and learning.