Preeti
Data artist Robert Hodgin recently created a feedback loop between Midjourney and ChatGPT-4, generating images of an old man in a messy room wearing a VR headset. This process involved prompting MJ to create an image, asking ChatGPT to describe it, and feeding that description back into MJ to generate another image, repeated 10 times. The project resulted in a series of images showing the progression of the AI-generated scene.
This recursive art project reflects a long history of similar experiments, such as Alvin Lucier’s “I Am Sitting in a Room” from 1969. However, it also highlights the challenges faced by AI companies in keeping their training data free from material created by AI. Ted Chiang has compared ChatGPT to a blurry JPEG of all the text on the web, emphasizing the approximation and lack of exact data in AI-generated content.
The project raises questions about the quality and accuracy of AI-generated content, especially when it comes to creating images based on textual descriptions. It also serves as a reminder of the limitations and potential pitfalls of relying on AI for creative tasks. The recursive nature of the project, similar to repeatedly saving JPEG images, hints at the potential degradation and distortion that can occur over multiple iterations. Overall, the AI image feedback loop project by Robert Hodgin showcases the intersection of art, technology, and AI, sparking discussions about the capabilities and limitations of AI in creative processes.
Source link
Source link: https://kottke.org/24/04/ai-image-feedback-loop
GIPHY App Key not set. Please check settings