| Théâtre D'opéra Spatial | |
| Artist: | Jason M. Allen |
| Year: | 2022 |
| Medium: | AI art, Midjourney |
Théâtre D'opéra Spatial (French for 'Space Opera Theater', pronounced as /fr/) is an image created by Jason Michael Allen with the generative artificial intelligence platform Midjourney. The image won the 2022 Colorado State Fair's annual fine art competition in the photomanipulation category on September 5, becoming one of the first AI-generated images to win such a prize.[1] [2] [3]
Allen said he used at least 624 text prompts and input revisions to get Midjourney to create the image, which he then manipulated with Adobe Photoshop, and enlarged with the Gigapixel AI tool.[4]
He disclosed his use of Midjourney when he entered the image in the contest's category of "digital arts/digitally-manipulated photography". The two judges for the category later said they did not know that Midjourney used artificial intelligence (AI) to generate images, but that they would have awarded Allen the top prize anyway.
Some artists accused Allen of cheating.[5] He responded: "I'm not going to apologize for it. I won, and I didn't break any rules."
On September 21, 2022, Allen, through his attorney Tamara Pester of TMBTQ Law,[6] submitted an application for registration to the U.S. Copyright Office. In subsequent communications with the Copyright Examiner prior to the first formal refusal, Allen's attorney provided further detail about the 624 prompts that he input to create the version of the work satisfactory to him, and the Copyright Office Examiner indicated that registration of the prompts alone may be permissible. Allen declined to limit his claim to the portion of the work that excludes the entirety of the Midjourney-assisted image or the Gigapixel AI enhancements.
In December 2022, the Copyright Office issued a first formal refusal, noting that "all of the pictorial and graphic content within the deposit is attributable to the AI -- your client did not paint, sketch, color, or otherwise fix" the submitted work. In January 2023, Allen, through attorney Pester, filed a first Request for Reconsideration. In June 2023, the Copyright Office continued the refusal, again inviting registration of the prompts and Photoshop enhancements alone, but indicating that the portion of the work created using the Midjourney and Gigapixel AI tools was not. Allen filed a Second Request for Reconsideration on July 12, 2023, noting that case law as well as public policy reasons supported registration of the work, encouraging the Copyright Office to more carefully consider the matter rather than rejecting anything AI-assisted. On September 5, 2023, the United States Copyright Office Review Board made a final determination and found[7] that Théâtre D'Opéra Spatial was not eligible for copyright protection as the rules "exclude works produced by non-humans".[8] This decision[9] continued the Copyright Office's rejection of AI as a creative tool.[10] The decision also creates a quandary because it indicates that Allen's work did not contain human authorship (suggesting that the AI technology is the author), in contrast to a recent court case, Thaler v Comptroller-General of Patents, Designs and Trademarks[11] which found against Thaler on the basis that he listed an AI program as the author, which is also impermissible. In the Thaler case, the decision focused on how AI cannot be an author; a work eligible for Copyright protection must have human authorship -- a working principle used by the Office. The case leaves a black hole in terms of authorship of AI assisted works: if Allen is not the author and AI can't be an author, to whom is credit for the creation owed? Allen insists he will continue to try to gain copyright registration:
Allen was dogged in his attempt to register his work. He sent a written explanation to the Copyright Office detailing how much he'd done to manipulate what Midjourney conjured, as well as how much he fiddled with the raw image, using Adobe Photoshop to fix flaws and Gigapixel AI to increase the size and resolution. He specified that creating the painting had required at least 624 text prompts and input revisions.[12]