Intellectualism Has Hampered Generative Art (2018) (www.tylerxhobbs.com)

The essay argues that generative art has been dominated by intellectualism — a preoccupation with clever algorithms, optimizations and technical novelty driven largely by mathematicians and engineers — and that this focus has weakened the work’s capacity to communicate to viewers. The author insists that what matters in the moment of active viewing is perception and emotional engagement: technical backstories or esoteric toolchains don’t make an image resonate. Quoting Agnes Martin’s distinction between the “intellectual way” and the “inspired way,” the piece says generative artists must prioritize making images they truly love that can draw others in, rather than relying on cleverness that mainly impresses peers and academics. For the AI/ML community this is a practical challenge and design imperative: models, loss functions and pipelines are enablers, not ends. The implication is to shift evaluation and development toward human-centered metrics — perceptual testing, iterative aesthetic refinement, curator/artist-informed priors, and cross-disciplinary study of traditional visual practices — instead of optimizing solely for novelty scores or algorithmic sophistication. Doing so means treating generative systems as tools for expressive goals, aligning technical choices with viewer experience, and measuring success by emotional engagement rather than technical impressiveness alone.