AI as a Catalyst for Deep Thinking and Real Learning

Although our core mission at Mindplace centers on mental health, psychotherapy, and psychiatry, our day-to-day practice is deeply intertwined with education and learning.

Many of us on the team split our time between clinical practice and teaching. On any given week, we see patients struggling with academic challenges, and we conduct a significant number of neuropsychological evaluations, particularly for ADHD and dyslexia.

In our consulting rooms, we see a steady rise in attention difficulties, impaired memory retention, lack of patience, and a growing struggle to pursue long-term goals. At the same time, we are witnessing a decline in genuinely original, personal thinking, the kind of work where you can see a person's authentic voice and how they have truly internalized knowledge.

Of course, no neuropsychological phenomenon has a single cause. Yet we cannot ignore the massive footprint artificial intelligence is leaving on every area of our lives. Formal education is, without question, one of the most impacted spaces.

So, what do we actually know? How does AI usage reshape our cognitive processes, and how can we leverage it as a catalyst for critical thinking rather than a substitute for the human mind?

1. Cognitive Prosthetics and Mental Muscle Atrophy

From a neuropsychological perspective, the brain is remarkably efficient: it constantly seeks to conserve energy and take the path of least metabolic resistance. When we outsource complex mental tasks to technology, we reduce cognitive load, but we also risk understimulating the neural networks that support working memory, sustained attention, and executive function.

Consider a striking neurobiological finding: the brain’s metabolic shortcutting. When we process information superficially or offload analytical thinking, glucose and oxygen consumption in the prefrontal cortex drops precipitously.

True learning doesn’t happen by passively consuming the right answer. It happens when we navigate what cognitive psychologists call "desirable difficulties"—that subtle, active effort of recalling, connecting, questioning, and reconstructing information. AI provides instant answers, effectively stripping away the incubation phase of thought. Without cognitive friction, there is no deep neuroplasticity. When processing lacks effort, knowledge fails to consolidate into long-term memory and never becomes part of a student's core identity.

2. The Neurobiological Power of Paper: Deep Reading and Handwriting

In an era dominated by digital speed and language models, neuroscience has uncovered surprising evidence supporting two analog tools we often take for granted:

High-density electroencephalography (EEG) studies led by Dr. Audrey van der Meer at the Norwegian University of Science and Technology (NTNU) reveal that writing by hand triggers widespread neural connectivity across alpha and theta frequencies that simply does not happen when typing. The motor complexity of shaping letters requires fine sensorimotor integration between parietal and frontal regions, which primes the brain for conceptual encoding. Typing, by contrast, relies on a uniform button-press motion for every letter, depriving the brain of that crucial proprioceptive anchor.

Meanwhile, a massive meta-analysis by the European research network E-READ, encompassing data from over 170,000 participants, revealed a striking statistic: reading print material offers a deep comprehension advantage equivalent to roughly three years of academic growth compared to reading on screens. This occurs because human cognition relies on spatial navigation. The brain uses physical cues—the tactile feel, thickness, and layout of pages—to map out a mental architecture of the text. Screen reading fragments this spatial mapping, pushing the brain into a rapid-scanning mode that hinders complex inference and narrative empathy.

3. From Metacognition to "Externalizing the Self"

One of the most concerning trends we observe in clinical practice is how technological outsourcing impacts metacognition: our ability to monitor and evaluate our own thinking processes.

When students delegate synthesis, writing, or problem-solving to an algorithm, two major shifts occur:

• Loss of personal voice: Knowledge becomes homogenized and impersonal. That distinct human signature, the ability to infuse ideas with personal experience, style, and critical judgment, gets swallowed by statistical language averages.

• The illusion of competence: It triggers an attribution bias where the brain confuses the AI’s clear output with its own understanding. Students experience a false sense of mastery, but when asked to apply that knowledge in a novel context, their mental mental model falls apart due to a lack of synaptic consolidation.

  
  

4. Toward an Integrative Model: AI as a "Socratic Companion"

At Mindplace, we don't advocate for banning or demonizing AI—that is both impractical and ineffective. Instead, we advocate for rethinking our relationship with digital tools. AI should transition from acting as an oracle (a dispenser of final answers) to a Socratic mirror (a tool that prompts deeper questioning and expands perspective).

To foster learning grounded in genuine neurobiology and psychology, we recommend shifting how we use these tools:

• Encourage critical friction: Rather than asking AI to write an essay, students can ask it to stress-test their own handwritten drafts for logical flaws or weak arguments. This keeps higher-order executive functions—judgment, inhibition, and evaluation—centered in the human brain.

• Evaluate the process, not just the product: Education must shift toward valuing marginal notes, handwritten outlines, step-by-step reasoning, and oral defenses. Authenticity isn't found in a polished final piece, but in the cognitive journey taken to build it.

• Protect analog spaces: Intentionally safeguard time for print reading and handwriting to activate the Default Mode Network (DMN)—the neural network that engages when the mind wanders, links distant concepts, and consolidates creative and emotional learning.

  
  

Final Thoughts

The real risk of Artificial Intelligence isn't simply that machines will start thinking like humans, but that humans will begin thinking like algorithms: reactively, rapidly, superficially, and predictably.

Protecting genuine learning is, ultimately, a matter of mental health. Training the mind to sit with uncertainty, embrace the effort of discovery, hold onto deep reading habits, and write by hand to think are the strongest neuropsychological tools we can offer future generations.