Summary:

- This article discusses the concept of "generative aesthetics" in the context of AI-generated art and poetry. It explores how AI systems can sometimes get "stuck" in formal patterns, leading to repetitive or predictable outputs.

- The article examines how AI models trained on large datasets may struggle to break free from the formal structures and patterns present in the training data, resulting in a sense of "formal stuckness" in the generated content.

- The author suggests that understanding and addressing this challenge could lead to more diverse and expressive AI-generated art and poetry, allowing the technology to move beyond its current limitations and unlock new creative possibilities.