1 / 7
The Museum of Prompt Mishaps' takes a critical eye to artificially intelligent 3D object generation technologies by generating objects based on 'confusing' or 'impossible' prompts. The goal is to consider how 3D object generation technologies understand concepts that may be abstract or non-understandable to them. Examples of this would be prompts that include phrases such as: 'make a 3D object of a bird's song', 'make a 3D object of melancholy', or a more poetic prompt that gives no particular direction, such as: 'time is of the essence and the blackbirds are singing today'. From these prompts, I have collected together a series of roughly 70 3D models that seem in some way physically broken or non-representative of any particular form, poetically expressing how the machine has been 'confused' by these prompts. 'The Museum of Prompt Mishaps' video presents as a large formless 3D space in which 50 of these animated yet seemingly broken 3D sculptures can be moved around and between. The camera moves, pans, and passes around and between these sculptures, moving you through the museum. The 3D object generation tools experimented with include platforms such as Meshy, Luma, and others.
One channel video. Under 10 minutes. mp4 file, 16X9 size. Has audio soundtrack.
Alinta Krauth
Alinta Krauth is an Australian new media artist who creates digitally innovative connections between machine learning, animal and intelligences, and interfaces. Her work often looks at topics related to animals and environments, including the use of interactive devices as a response to more-than-human agency.
