To understand why I stopped assigning essays, we must acknowledge that AI has become a master 
of the identification tasks that once formed the backbone of Art Appreciation. It is not that AI is the 
best student in a holistic sense. It lacks the lived experience, emotional resonance, and personal 
critical thinking that a human brings to a gallery, yet it is an unparalleled expert in pattern recognition. 
Through deep learning and neural networks, AI is learning to see by recognizing the same elements 
and principles of design I teach. With vision-language models (VLMs), AI can describe what it sees 
at a microscopic level to identify the medium, tools, and techniques employed in an artwork with 
extreme precision. 
Advanced models even utilize 3D topography to analyze paint texture, identifying individual 
brushstrokes to determine authenticity. I put this on the table not to argue that we should stop 
teaching these concepts, but to acknowledge the environment in which we teach. Students still 
need to learn the elements and principles of design, but we must move beyond identifying what is 
on the canvas to exploring how that knowledge can be used to navigate the world of art. An essay 
that merely identifies a medium or elements and principles is now a task that can be easily done by 
a machine, and I have learned this the hard way.
I chose to navigate this reality by granting myself permission to experiment within what I call the 
Autonomy of Dependency. I draw the term dependency from Theotonio Dos Santos and his 1970 
work, The Structure of Dependence, which explores how technology serves as a primary mechanism 
of subordination. For Dos Santos, dependent groups expand only as a reflection of the dominant 
Figure 1. Real examples of visual outcomes generated by Copilot using student prompts.
45
TRENDS 2026 	
	
PAGE

View this content as a flipbook by clicking here.