The rise of generative artificial intelligence has reignited an essential question: is AI weakening our cognitive abilities?
In both corporate settings and training environments, concerns are mounting. Some point to a loss of focus, decreased memorization, or an increasing dependence on tools. Yet, beneath these perceptions, the reality is more nuanced. AI does not reduce intelligence; it transforms the way we mobilize our cognitive abilities.
AI and cognition: a transformation of intellectual effort
Relying on tools to think is nothing new. Cognitive psychology refers to distributed cognition, a concept according to which we outsource certain mental tasks—such as writing, calculating, or searching for information—in order to go faster and further.
AI fits into this continuity, but it introduces an important shift: it is capable of directly producing structured answers. This evolution can reduce immediate effort, accelerate production, and create the impression of rapid understanding. However, quick access to an answer in no way guarantees genuine learning.
The trap: understanding without knowing how to do
A well-documented phenomenon in cognitive science is that of the illusion of competence. When an answer is clear and well-formulated, it creates the feeling of understanding. Yet, this impression does not guarantee the ability to reproduce the solution autonomously.
In other words, it is possible to recognize a correct answer without being able to produce it oneself.
How the brain actually learns
To evaluate the real impact of AI, it is necessary to return to the fundamental mechanisms of learning.
Active engagement
The brain learns when it is truly engaged. This involves reflecting, formulating ideas, and making choices. Simply reading or observing is not enough to anchor knowledge long-term.
The effort of retrieval
The work of Roediger and Karpicke (2006) shows that practice improves memorization more than simple re-reading. Answering a question, attempting a solution, or confronting one's own mistakes are key stages in the learning process.
The role of feedback
Effective learning requires rapid feedback, corrections, and adjustments. Without feedback, learning remains superficial and lacks durability.
AI: a risk… or an accelerator of learning?
The impact of AI depends entirely on how it is used.
Passive use, which consists of copying an answer, reading without interaction, or entirely delegating thought, leads to low cognitive engagement and therefore limited learning.
Conversely, active use—based on questioning, reformulation, putting one's understanding to the test, or requesting exercises—significantly reinforces learning.
AI as a tutor: a major opportunity
Educational research shows that individualized tutoring is one of the most powerful levers for learning, as demonstrated by Bloom in 1984. Today, AI allows us to get closer to this by offering real-time interaction, adaptation to the learner's level, and personalized explanations.
It thus becomes a true learning partner, rather than just a simple tool.
The real challenge in the corporate world
In organizations, a major difficulty arises: knowing how to use a tool does not mean mastering a skill. This confusion can result in difficulties acting without assistance, a partial understanding of subjects, or errors in real-world situations.
The challenge is therefore clear: it is about developing solid, transferable, and sustainable skills.
Rethinking training in the age of AI
Traditional formats are now showing their limits. Content is often passive, engagement is low, and practical application is insufficient. AI paves the way for a new, more dynamic approach.
The most effective training programs now integrate concrete simulations, decisions to be made, errors to analyze, and personalized feedback. The learner thus becomes an active participant in their own progress.
How Complement uses AI to strengthen cognition
At Complement, a strong conviction guides our approach: learning happens through action, not through simple exposure.
The experience we offer is based on situations close to reality, with professional scenarios, concrete decisions, and directly applicable business context. Interaction is continuous thanks to a dialogue with a dedicated tutor, allowing users to ask targeted questions and guide their progress.
Training is personalized, with adaptation to each learner's level, identification of difficulties, and a tailored path. Finally, immediate feedback is provided, with real-time corrections and adapted explanations, to durably consolidate what has been learned.
Conclusion
Artificial intelligence does not automatically cause intellectual decline. Depending on how it is used, it can either reduce effort or, conversely, reinforce learning.
The difference lies in one key factor: cognitive engagement.
The organizations that succeed will be those that know how to leverage AI to stimulate reflection, encourage practice, and personalize learning.
📚 Sources
Bjork, R. – Memory and Metamemory Considerations in the Training of Human Beings (1994)
Roediger & Karpicke – Test-Enhanced Learning (2006)
Bloom, B. – The 2 Sigma Problem (1984)
Hutchins, E. – Cognition in the Wild (1995)
OECD – AI and the Future of Skills (2023)