Diego Rodriguez, co-founder of Krea.ai, discusses the challenges of evaluating AI-generated content, particularly concerning human perception and aesthetics. He highlights the current limitations of AI models in understanding subjective human reactions, contrasting them with how easily humans assess images, even those with obvious flaws. Rodriguez traces this issue back to how AI is trained on human data, including compressed images from the internet that already contain artifacts. He criticizes standard AI metrics, such as FID scores, for failing to align with human perception, leading to models being misjudged. Ultimately, Rodriguez advocates for the development of new, perceptually aware evaluation metrics that account for human opinion and the inherent biases in training data, especially as AI advances in areas like real-time generative media and cross-cultural communication.