New Research Identifies AI Deskilling as a Structural Problem Requiring Systemic Solutions
Key Takeaways
- ▸AI creates systemic conditions that undermine human skill development through what researchers call 'capacity-hostile environments'
- ▸Meaningful skill cultivation depends on habituation and social learning—conditions that AI mediation risks degrading
- ▸Current approaches to AI adoption often fail to consider impacts on human capacity development and long-term workforce capabilities
Summary
A new paper published in AI & SOCIETY journal argues that artificial intelligence poses a structural risk to human skill development, introducing the concept of "capacity-hostile environments" where AI mediation impedes human capacity cultivation. The research demonstrates that the problem extends beyond individual responsibility, showing how AI's influence creates systemic conditions that inhibit the development and exercise of human capacities by undermining habituation—the gradual process through which people acquire skills and learn from others.
The paper draws on philosophy and social epistemology to argue that meaningful skill acquisition requires not just learning the "how" of a skill, but also gaining agential control over capacities through prolonged practice and exposure to shared understandings of value. The authors identify a critical risk: AI systems that mediate human activity can degrade the very conditions necessary for this habituation process, leading to what they call "capacity impoverishment" across society.
The research calls for a fundamental shift in how AI applications are evaluated, proposing that societies must assess whether AI systems are conducive to or hostile toward human capacity cultivation. The authors emphasize that this requires critical reflection on the values embedded in AI socio-technical systems and argue for a societal obligation to create capacity-conducive environments rather than capacity-hostile ones.
- Society must shift from evaluating AI purely on efficiency to assessing its impact on human capacity cultivation
Editorial Opinion
This research highlights a blind spot in how we evaluate and deploy AI systems: we optimized for immediate productivity gains without adequately considering the structural costs to human capability. The paper makes a compelling case that allowing AI to erode the conditions for human skill development is not just a personal problem—it's a societal one. As organizations race to automate decision-making, recommendations, and content creation, this work suggests we need explicit governance frameworks that evaluate AI systems not just on output quality, but on whether they preserve or undermine our capacity to cultivate meaningful human skills.
