Artificial intelligence could destroy humanity not through a sci-fi "machine uprising," but via a gradual process of intellectual decline where humans simply stop thinking for themselves. Researchers are warning about the massive threat of so-called "enfeeblement," which quietly erodes society's cognitive abilities. Amid rapid digitalization and the integration of neural networks into all spheres of life, this process is becoming systemic, threatening the fundamental skills required for human survival and civilizational development.
The Threat of Cognitive Exhaustion and the Enfeeblement Phenomenon
The term "enfeeblement" is used by the scientific community to describe the gradual atrophy of human intellectual, planning, and emotional skills due to the constant and unchecked delegation of routine and creative tasks to algorithms. As Sheffield Hallam University senior robotics lecturer Dr. Samuel Vineanza explains, this process unfolds unnoticed against the backdrop of a psychological habit of "cognitive offloading"—the natural human urge to conserve mental effort using external digital tools.
The more heavily people rely on algorithms to write texts, write code, conduct deep research, and make strategic decisions, the harder it becomes for them to critically question AI outputs and timely identify its errors. Studies show that this dangerous degradation of critical analysis has already affected high-level specialists across diverse fields—from medicine to military operations.
Real-World Examples from Medicine and Military Analytics
Healthcare experiments serve as a clear illustration of this trend. During tests involving 27 qualified radiologists, doctors made accurate diagnoses 80% of the time when the AI gave correct hints. However, the situation changed dramatically when artificial intelligence was intentionally fed false prompts: the diagnostic accuracy of experienced physicians dropped below a catastrophic 20%, demonstrating professionals' blind faith in machine authority.
Similar risks are documented within national security sectors. Incidents are known where the US military nearly launched a dangerous boarding operation on a civilian Chinese vessel because a chatbot generated a false report alleging the transport of components for a secret nuclear program. Catastrophic escalation was averted solely due to the vigilance of a single individual who questioned the machine-generated data.
Contradictory Data
While pessimistic researchers like ControlAI founder Andrea Miotti argue that the rapid handover of control over hospitals, power grids, and armies to autonomous systems could strip humanity of its leverage in the coming years, skeptics urge not to exaggerate the threat. Some technology experts and AI developers insist that automation tools serve a purely auxiliary function, and responsibility for final decisions always rests with humans. Nevertheless, even moderate skeptics acknowledge the necessity of strict legal and ethical regulations to prevent the emergence of a "fragile generation" of specialists.
Educational Crisis and Global Consequences for Civilization
Of particular concern is the impact of algorithms on education and the training of future professionals. Professor Seun Kolade conducted an indicative experiment in which ChatGPT successfully wrote complex academic essays. The papers received top marks from independent examiners and easily bypassed standard plagiarism-checking software like Turnitin. This cultivates an entire generation of technically proficient yet critically vulnerable young people who can produce flawless-looking results without being able to assess their deep qualitative value.
Surgeons, engineers, lawyers, and teachers are currently at high risk of losing foundational expertise. As Andrea Miotti summarizes, superintelligent AI will not even need malicious intentions: in a blind pursuit of resources, the system might treat humanity the exact same way humans treat random obstacles during major highway construction.
