For those who claim that AI is the doomsday of art, feel free to go back to the ancient discussion about whether photography kills painting. The adaptability of a new medium in mimicking the visual quality of an old medium is usually considered a merit of the former and a weakness of the latter, but this argument is established upon a premise of a dogmatic and immortal aesthetic standard that lionizes certain artistic movements or styles. Look how this A.I. painting resembles Van Gogh! Even the most brilliant painter today can not recreate such vivid brushstroke! My interrogation towards such statement would be, who is real author of this final product born out of countless Van Gogh painting input? Would Van Gogh keep creating paintings of his signature style, were he alive today? 

​

The first question probes into the idea of authorship, which is now feverishly debated on Twitter. A.I. becomes the theft of human artists’ work fits perfectly into the Marxism idea of alienation: the more people use A.I. to create images based on the existing dataset, the less means of production “original artists” own. From my point of view, the debate of authorship can be analyzed with “waypoint or endpoint.” Ian Cheng, interviewed in this reading, celebrates A.I. assisted tools as a prototype tool that helps individual artists increase efficiency in the experimental phase. The waypoint argument resolves the authorship dilemma by integrating artists’ personal intention. I think whoever uses A.I. to generate end product, is not creating original works and is not entitled with authorship. 

​

The second question considers the idea of originality and the definition of art in the contemporary context. Tim Schneider considered A.I. as both an available tool for artists to use and a handcuff restricting how artists can and should use them. What Schneider want to say, although not explicitly expressed, might be that A.I. image generator is just another tool/medium, like printmaking or photography, except its technological aspect is not demystified, unlike other mediums. I agree with Ian Cheng that A.I. can be both tool and tyrant, depending on how people use it. Art today can not be discussed solely in its visual quality. A quotidian photograph of a starving child, taken by artist Sebastião Salgado, denotes not only the child’s disconcerting health, but also the complicated political relation between the North and the South. A photo realism style painting, looks exactly the same, does not has the equivalent power to provoke the meaning. That’s how documentary photography’s commitment of objectiveness comes into play. Similarly, A.I. assisted works have to commit to something internal to its medium instead of creating a visually compelling piece that mimics other mediums. The former method uses A.I. as a tool, the latter is constricted by it. Those artists are good examples of how to use A.I. as a tool. Feminist Data Set, a work by Caroline Sinders, is interesting in terms of how she interacts with every step in machine learning - data collection, data labeling, data training, selecting an algorithm to use…She deals with the internal texture of artificial intelligence. Also see Ross Goodwin in how he represents the interplay, mediated by algorithm, between image and text in his work Word.Camera. Also check out how Keaton Pattin trains a bot with thousands of hours of Batman film and makes the bot writes a screenplay(he didn’t claim himself an artist but I think he artistically heightened the iconic tropes pervasive in American contemporary culture.)