Artificial intelligence continues to transform societies at an unprecedented pace. It is becoming increasingly clear that AI technology differs fundamentally from previous technological breakthroughs. As virtually every field – from education and health care to industry, the economy, biotechnology, and defense – undergoes transformation in response to new dynamics, an entirely new technological ecosystem centered on artificial intelligence is emerging. On the one hand, this rapid transformation is generating growing excitement and expectations across all segments of society; on the other, it is also intensifying concerns and anxieties. Perhaps no previous technological development has ever evoked these two contrasting emotions simultaneously and with such intensity.

The history of the relationship between humans and machines is, in fact, also the history of humanity's efforts to transcend its own limitations. Throughout this journey, from the earliest simple tools to complex machines, humans initially sought to enhance their physical capabilities and overcome the constraints imposed by nature. Agricultural tools, mechanical devices and eventually the machines introduced by the Industrial Revolution complemented human muscle power, took over a substantial portion of physical labor and dramatically increased productive capacity. With the advent of electricity and automation technologies, this relationship entered a new phase. Machines were no longer merely tools that enhanced human physical strength; they became capable of carrying out certain production processes with progressively less human intervention. The development of computers extended human-machine relations beyond the physical domain, bringing computation, data processing, information storage, and certain decision-making processes within the scope of machine operations. Thus, throughout history, humanity has used machines first to expand its physical capabilities, then to automate repetitive tasks, and finally to begin transferring certain aspects of its cognitive activities to machines.

Artificial intelligence, however, is now taking human-machine relations into an entirely different dimension. While earlier technologies enhanced human physical capabilities, computers could perform certain cognitive operations within the boundaries of predefined rules and algorithms. Artificial intelligence, particularly generative AI technologies, is entering a much broader domain that encompasses not only what humans do but also how they think, interpret information, solve problems and generate new content. Moreover, this involvement is not limited to supporting human cognitive capacities; in some cases, AI can directly substitute for human thinking and creative processes. This is precisely where the historic turning point in human-machine relations emerges: Throughout history, humans have developed machines to enhance their own capabilities, but today, for the first time, they face the possibility of routinely delegating their own cognitive activities to machines on a widespread scale.

Recent experimental studies suggest that certain cognitive processes associated with critical thinking, independent problem-solving, and memory may be adversely affected, particularly when AI tools are used as substitutes for individuals' own thinking and problem-solving activities. Of course, these effects vary depending on how AI tools are used. Nevertheless, we are now confronted with an ecosystem whose dynamics we have never experienced before and that requires us to exercise considerable caution in every aspect of human-machine relations. Within this emerging ecosystem, AI technology is increasingly reshaping social stratification, introducing new divisions based on skill levels alongside traditional distinctions of class and social status. Moreover, the risk we face is not merely that existing skill disparities may widen. A new divide may also emerge between those who can use AI technologies effectively and those who become dependent on the ready-made outputs these technologies provide. In such a scenario, significant differences in skills and capabilities may develop even among individuals who use the very same technology. In short, what ultimately matters is not simply access to these technologies, but the nature of the relationship individuals establish with them.

In this context, it is worth repeatedly emphasizing two critical issues concerning AI tools. First, the transformation in human-machine relations discussed above may increasingly develop into a cycle of cognitive dependence. As individuals delegate their thinking, analysis, and problem-solving processes to AI tools, they reduce their own mental effort in these activities. This is because, by their very nature, these tools are designed to take over cognitive processes that are fundamental to what makes us human. A sustained reduction in mental effort may, over time, weaken individuals' ability to perform these tasks independently. As a result, the need for AI tools will increase, while greater reliance on these tools will further accelerate the transfer of cognitive processes to machines. Humans are thus being drawn into the turbulence of a potentially dangerous cycle. Moreover, as this cycle of dependence intensifies, it may not only weaken individuals' capacity for independent problem-solving but also cause them to feel inadequate when they cannot access AI tools.

The second issue concerns the accuracy of the information generated and the question of agency in the resulting products, both of which become even more critical as cognitive processes are delegated to machines. While these tools have made it possible to produce virtually any type of content rapidly and at negligible cost, the question of information accuracy has become more critical than ever before. This is because AI tools can not only reproduce biases embedded in their training data but also present information that has no basis in reality in an extremely convincing manner. Therefore, it is essential to evaluate the accuracy of AI-generated content before using it as is. However, when cognitive processes are increasingly delegated to machines and dependence grows, as described in the first issue, the very capacity to verify such information may be undermined, thereby magnifying the scale of potential harm.

Another issue closely related to the accuracy of information is the question of agency in the production of content. While the issue of accuracy concerns the relationship between the information produced and reality, as well as the human responsibility to evaluate that information, the issue of agency concerns the extent to which humans remain active agents in the production process. We are therefore confronted with two distinct but equally important consequences of the same transformation. On the one hand, individuals risk distancing themselves from their responsibility to critically evaluate the information produced; on the other, they risk losing their central role in the production process itself. As noted above, when virtually any type of content can now be produced rapidly and at negligible cost using these tools, who is the actual agent behind the resulting product? Here again, the position humans occupy within human-machine relations becomes crucial. Furthermore, research suggests that when agency shifts from humans to machines, individuals may become alienated from the products they generate and may struggle to recall even basic details of content produced only a short time earlier, indicating a weakening of memory. In addition, particularly in written texts, a substantial decline in the diversity and richness of arguments has been observed.

In short, the central concern underlying the issues discussed above is how to protect human beings within this relationship. In other words, how can we preserve the qualities that make us human in our interactions with artificial intelligence? Both issues discussed above are directly related to the potential threat that machines pose to fundamental aspects of human existence. This is precisely where the manner in which artificial intelligence and related tools are used becomes decisive. It is therefore clear that we need a pedagogical framework for human-machine relations that does not diminish human participation in cognitive processes, continuously maintains awareness of the need to evaluate the accuracy of generated content, and ensures that agency over the final product remains with humans.

At this point, it is worth revisiting the Turkish Statistical Institute's (TurkStat) 2026 statistics on artificial intelligence. The data indicate that the use of AI for professional purposes remains considerably more limited than its use for personal purposes. For example, in 2026, approximately 86.4% of individuals using these technologies (compared with 80% in 2025) reported using them for personal purposes, while only 37.3% (compared with 33.8% in 2025) reported using them for professional purposes. These figures suggest that although AI adoption is rapidly expanding in Türkiye, this expansion has not been reflected to the same extent in professional applications and, consequently, in professional productivity. However, this difference in the purposes of AI use does not, by itself, reveal the level of awareness or skill with which individuals use these tools. Precisely for this reason, we need to measure not only the prevalence of AI use but also the quality of that use. The extent to which individuals remain actively engaged in cognitive processes when using these tools, how thoroughly they verify AI-generated content, whether they preserve their independent problem-solving abilities, and whether they can translate AI use into professional productivity have now become fundamental questions requiring systematic investigation.

Ultimately, the central issue in the age of artificial intelligence is not how widely these technologies are adopted, but where human beings stand within this new relationship. Are we able to preserve our abilities to think, question, verify, and create while using artificial intelligence? Or are we gradually delegating these abilities to machines and becoming increasingly dependent on the capabilities they provide? Perhaps the defining distinction between societies in the future will not be who has greater access to AI technologies, but who is better able to preserve human cognitive autonomy while using them. For this reason, the most important educational challenge of the AI age is not merely to raise individuals who know how to use these technologies, but to cultivate individuals who can strengthen their capacity to think and create alongside AI without surrendering the responsibility for thinking to machines.