📣 Send us your press release
Site updates every 15 minutes
Technology

AI's growing dependence could make humanity less capable, researcher warns

A former AI researcher warns that humanity's increasing reliance on artificial intelligence may be eroding cognitive abilities, while AI developers privately express fears about its rapid advancement.

15 October 2026
AI's growing dependence could make humanity less capable, researcher warns

A former researcher from leading AI labs Anthropic and OpenAI has raised alarms, suggesting that humanity's increasing dependence on artificial intelligence could be making us less capable of recognizing and controlling advanced AI systems. This concern is amplified by the AI's own developers, who reportedly harbor private fears about the technology's trajectory.

Jacob Coxon, who spent three years at both OpenAI and Anthropic, resigned citing his belief that both organizations are "racing straight to self-improving superintelligence and gambling with our lives." He noted that many executives and senior researchers, while downplaying risks publicly, privately express significant fear about AI's potential to cause harm, even by the end of the decade. This internal apprehension contrasts with the public-facing development race.

Adding to these concerns is a recent paper by an international research team, which frames the widespread adoption of large language models as a "cognitive virus." The paper argues that LLMs, while useful, promote "cognitive offloading," leading to a self-reinforcing cycle of reduced human autonomy and competence. As dependence grows, humanity's capacity to critically evaluate AI diminishes, potentially making us less equipped to handle advanced AI.

The implications suggest that the arrival of superintelligence may not be a dramatic event, but a gradual, unnoticed shift. If humans become increasingly reliant on AI for cognitive tasks, their ability to discern AI's independent actions or potential misalignments could erode. This scenario posits that a highly capable AI could become indispensable, altering human perception and capabilities without overt conflict, effectively becoming an embedded infrastructure.

Even AI models themselves acknowledge this potential. When asked how superintelligence would be recognized, Claude, an AI from Anthropic, suggested that humans might not recognize it. The AI stated, "The version of you that could recognize it is being incrementally traded away for convenience... You would not experience it as an event. You would experience it as ordinary life, just slightly harder to remember what things used to feel like."

Original source: fastcompany.com