Is artificial intelligence turning off our minds?

Despite its undeniable advantages, using AI programs introduces latent threats such as biases, mental laziness, and cognitive debt

Pilar Jericó

JUL 17, 2025 – 12:35 CEST

Artificial intelligence is like living with a teenager: sometimes brilliant; other times, a little clumsy, but it never ceases to amaze us… When we interact with artificial intelligence (AI) programs like ChatGPT, Gemini, Claude, Grok, and so many others, we are amazed by their tremendous potential to transform the future, both at work and in our everyday lives. These are applications that inspire curiosity and in which we also place our trust. In fact, a recent study published in Harvard Business Review lists among the most prominent uses of AI in 2025 its role as emotional support. Beyond generating texts or automating tasks, these tools are being used as therapists, to organize our lives, and even to help clarify our life purpose. And this is just the beginning. As Dr. Silvia Leal, technology expert and advisor to the OECD, predicted: “We will talk more with chatbots or AI applications than with our partners.” She published this in 2017; today, many people spend more time with AI-powered tools or devices than with real humans. To make proper use of this technology, we need to understand the risks our minds face when they come into contact with it. Let’s look at three of the most relevant ones.

The first risk stems from the world of bias. Dr. Leal explains: “AI advances without being able to avoid data biases or internet lies. Furthermore, it develops in our own image and likeness, and humans are far from perfect.” In other words, this technology will be fast and surprising, but it can fail. Furthermore, the belief that AI is infallible accentuates our automation bias, whereby we don’t question the results it offers us or we blindly follow a browser, even if it takes us to a site we think is wrong.

The second risk arises from our brain’s energy conservation. Our gray matter is programmed to save effort; it weighs 2% of our body, but consumes more than 20% of our daily energy. For this reason, we tend to avoid certain unnecessary efforts and fall into what is known as cognitive laziness. Due to laziness, we stop remembering phone numbers or practicing mental calculations if we have address books or calculators on our phones. Also because of laziness, when faced with a doubt, we have an unconscious tendency to look for a quick solution to reduce the discomfort, according to researchers at the University of Texas, after analyzing how frequently students reviewed their exams before handing them in. With this natural starting point, it’s not surprising that AI accentuates cognitive laziness and begins to weaken our memory when it comes to recalling information.

Finally, and related to the previous point, the third unconscious risk may be the most worrying of all: the brain activities we “sacrifice” beyond memory. In a recent experiment conducted by MIT, the activity of 32 brain regions of volunteers writing essays on philanthropy was measured over the course of four sessions and using different resources. The first group had no help; the second could only rely on search engines like Google; and the third could use ChatGPT. After the sessions, they discovered that the third group recorded lower brain activity, poorer memory of what they had written, and a weaker sense of authorship compared to the other groups, especially the first. Or, as the researchers themselves noted, the groups that used the AI “consistently underperformed at neural, linguistic, and behavioral levels.” But even worse, the last essay the volunteers had to write had to be a compilation of the previous three, and in this case, they were asked to switch resources. Those who had used AI switched to relying solely on their own reflections, and so on. What emerged was that the group that had initially used ChatGPT maintained the same limited mental activity, even when they no longer had access to it. This phenomenon is known as “cognitive debt.” In other words, intensive use of AI can provide short-term benefits, but at the cost of impaired learning, the ability to form complex ideas, creativity, motivation, or a deep connection with what they’ve learned, among others.

The three difficulties mentioned above warn us of something important: with AI, we run the risk of becoming increasingly dependent and of exerting less mental effort in our daily activities. Furthermore, we know that the brain, in its efficiency, tends to weaken the functions it doesn’t exercise. Therefore, we need to train critical thinking, make conscious and limited use of artificial intelligence, turning it into a coach and not a crutch that could replace many of our mental processes. We must question the results this technology offers us based on our experience or that of other sources, use it as a mental sparring partner, but not turn it into a shortcut that compromises our learning or into an all-knowing oracle.

Pilar Jericó is the author of the Happiness Lab blog.

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition

Society El País in English on Facebook

Society El País in English on Twitter