Pieces of art like Théâtre D’opéra Spatial, created by Jason Allen using Generative AI, which won the 2022 Colorado State Fair’s annual fine art competition, raise the questions about how we perceive art in the current digital age, and whether AI could be used by humans to create artwork showing human’s own creativity.
While the human part of an artwork represents the author’s ideas, culture, and heritage, the technology part represents the implementation phase of the work, where prompting is used to derive an outcome matching the artist’s imagination. The artist must be aware of the capabilities of the AI system he uses based on knowledge about its data and algorithm. This kind of knowledge represents the domain in which the artist can create his work, just like musical notes for a musician and colors for a painter.
The more the details, in the form of specific instructions, provided by the artist to the Generative AI model, the more clear is the artist’s contribution to the artwork. Accordingly, the copyright for such a piece of work should belong to the artist who created it just like the copyright of a work of music belongs to the composer, not the instrument. It seems we’re in a new era of art innovation where creative artwork requires knowledge in both art and technology fields.
A modern artist has to be acquainted with AI principles at some basic level to be able to apply the technology efficiently in his work, though it always depends on what he wants to create. On a basic level, the ability to work with AI tools, generative models, and their capabilities enables artists to try new forms of expression. However, when it is about really knowing how to fine-tune models, manipulate algorithms, and integrate AI with classic techniques, then more unique and peculiar things can be created. Although technical skills are not necessary, a conceptual understanding of the potential, limitations, and ethical implications of AI allows the artist to push creative boundaries while maintaining artistic intent.
Mohamed el Nawawy
The understanding needs to be enough to realise the point of the work. AI is being integrated into most computer applications. It is often there when using a search engine, spelling and language correction when editing a document, noise reduction in digital images shot in low light. None of these require any understanding of AI. Generating an image or 3D model from a text prompt may have a different meaning or impact depending on the context of the given prompts or what is behind the decision to use generative AI. I think the data used by generative AI needs some consideration (is it another form of collage?). For commercial art the requirements are, and should be different. Much of the data used to train generative AI models has not been paid for or licensed for commercial use.
Brian Buxton
Thank you for your valuable contribution. My explanation about why we are not required to understand AI when using it for editing a document while we need to understand AI when using it to generate artwork would be by giving an example from cloud computing. Using AI for document editing is like using Software as a Service in cloud computing, which represents a ready made tool that is suitable for a wide range of users. On the other hand, using AI in generating artwork is similar to using Infrastructure as a Service in cloud computing which requires more technical knowledge by the user, yet it provides him with an increased customizability, a feature which is necessary for the artist to realize his view.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders