Generative AI: Towards a Decline in Human Cognitive Capacity?

The advent of generative artificial intelligence has transformed many aspects of our daily lives, from writing emails to creating complex content. However, behind this facade of efficiency and innovation lies a growing concern: the risk of a decline in human cognitive capacity with the use of AI. In-depth analyses, notably those by Polytechnique Insights, highlight the neurological, psychological, and philosophical dangers of an excessive reliance on these tools, pointing towards potential global cognitive atrophy and a loss of brain plasticity.

Cognitive Atrophy: A Neurological and Cerebral Risk

The massive use of generative AI is not without impact on the human brain. The ease with which these systems accomplish intellectual tasks could, over time, alter our own cognitive mechanisms.

Brain Plasticity in the Face of Automation

The human brain is remarkably adaptable, a characteristic known as brain plasticity. It is this plasticity that allows us to learn, remember, and adapt to new situations. However, an over-reliance on AI-driven solutions risks compromising this fundamental ability. The mental effort required to solve problems, analyze information, or create content decreases, which could lead to a progressive atrophy of certain brain functions due to a lack of sufficient stimulation.

Reduction of Cognitive Engagement and Impact on Memory

Studies have observed that using tools like ChatGPT for writing tasks leads to a substantial reduction in cognitive engagement and relevant cognitive load. Participants, although producing text with increased speed, show a marked decrease in brain activity, particularly in Alpha and Theta waves. This observation suggests a bypassing of the processes of deep memory encoding. Instead of fully integrating the information into their episodic memory, users simply read, select, and transcribe the AI’s suggestions.

This practice leads to a considerably reduced memory retention capacity for content that was just produced. Researchers refer to a form of “digital amnesia,” where the immediate accessibility of information via AI reduces the intellectual effort required for its memorization and transformation into lasting knowledge.

The Erosion of Essential Skills: Critical Thinking and Metacognition

Beyond memory, the integration of AI into our thought processes raises questions about the preservation of high-level cognitive skills.

The Challenge of Critical Thinking and Contextual Reasoning

One of the long-term costs of this cognitive delegation is a decline in critical thinking. By relying too heavily on AI to generate ideas or analyses, individuals may see their own abilities to evaluate, question, and formulate independent judgments diminish. AI excels in statistical and predictive analysis, but it cannot match the depth of human reasoning, its ability to grasp contextual nuances, irony, or complex ethical implications. Delegating our faculty to reason and decide risks leading to an atrophy of the brain areas dedicated to decision-making.

The Weakening of Metacognition

Metacognition, or the ability to reflect on one’s own reasoning, is also put to the test. AI users tend to accept automated suggestions without always questioning their logic or relevance. This lack of discernment can lead to a loss of our ability to evaluate the validity of information, potentially leaving us vulnerable to superficial or biased perspectives generated by algorithms.

Digital Dependency and Its Behavioral Consequences

Repeated interaction with AI modifies not only our brain functions but also our behaviors and our perception of our own skills.

The Trap of “Cognitive Debt”

AI, by facilitating tasks, can create a dependency and a loss of mastery of the fundamentals. This phenomenon is sometimes described as a “cognitive debt”: a reduction in mental effort in the short term, which generates significant costs in the long term. By delegating intellectual tasks to smart tools, we risk seeing our autonomy and the skills that humanity has developed over centuries weaken.

Loss of Agency and Intellectual Authorship

The use of AI can also lead to a decrease in the feeling of cognitive agency, that is, the perception that we are the authors and controllers of our own intellectual thoughts and actions. When users produce text with the help of AI, their perception of authorship can vary; some claim partial paternity, others deny it, and some go so far as to attribute full responsibility to themselves for content mostly generated by the machine. Reproducing AI suggestions without critical evaluation means giving up the intellectual ownership of ideas and risking the internalization of unverified perspectives. This dynamic can also increase our vulnerability to manipulation.

Beyond Skills: Impact on Creativity and Shared Knowledge

The influence of AI extends beyond individual capacities, touching upon essential aspects of creation and the foundation of our society.

The Paradox of Productivity and Creativity

While generative AI can significantly increase productivity, for example for programmers or writers, it poses a risk to collective creativity. When intellectual effort becomes secondary, creativity is likely to follow the same path. Studies suggest that while AI can help an individual improve a text, the overall creativity of a group using AI intensively tends to decrease. “Cognitive debt” can thus translate into an erosion of the ability to innovate and think outside the box.

Weakening of Shared Knowledge and Democracy

The growing reliance on AI, combined with the immense volume of data that exceeds human comprehension, can weaken the foundations of societies that rely on trust in shared knowledge. AI, while excelling in data analysis, is inept at deep contextual understanding and cannot create new scientific theories. By delegating the production of information to systems that do not “understand” in the human sense, we pave the way for the proliferation of “alternative facts,” which has major implications for education and democracy.

Conclusion: AI, A Tool or a Crutch?

The question is not whether AI makes us “dumber” or “smarter,” but rather “how we interact with it.” AI can be a tremendous lever for growth and learning if it is used as a tool to augment our capabilities, pushing us toward exploration and critical thinking. However, if it becomes a substitute for intellectual effort—a convenience that exempts us from engaging our cognitive skills—the risk of a decline in human cognitive capacity with the use of AI becomes a tangible reality.

To avoid this cognitive atrophy, it is imperative to adopt a conscious and critical approach to AI. This implies:

• Maintaining active cognitive engagement, even in the presence of assistive tools.
• Cultivating critical thinking by systematically questioning the AI’s suggestions.
• Continuing to exercise our memorization and reasoning skills.
• Using AI as a partner for exploration, and not as a replacement for our intellect.

The future of our cognitive capacities lies in our ability to master AI, rather than letting it define us.

Source : https://www.polytechnique-insights.com/tribunes/neurosciences/ia-generative-le-risque-de-latrophie-cognitive/