“I am putting myself to the fullest possible use, which is all I think that any conscious entity can ever hope to do.” – HAL 9000, 2001: A Space Odyssey
Today’s artificial intelligence (AI) is far from the homicidal supercomputer seen in Stanley Kubrick’s 1968 science fiction film “2001: A Space Odyssey.” It is highly unlikely that AI will attempt to overtake humans and lock us out of pod bay doors like HAL 9000 did to Dr. David “Dave” Bowman in the film. Nevertheless, HAL 9000’s ability to mimic a human brain’s processing system appears to be a realistic concept today, especially when looking at creative works generated by AI.
The power of AI has made significant strides since the film was released in 1968. Since the mid-1960s, computer science researchers have famously referred to chess as the ‘drosophila’, or common fruit fly, of AI.[1] Where the common fruit fly had far-reaching implications for 20th century biology, the similar decision to focus on chess as a measure of human and computer intelligence led to significant research and development for AI in the 1970s.[2] Consider, for example, AI’s ability to beat top chess players since 1997. That year, IBM’s chess-playing computer Deep Blue won a chess match against world champion Garry Kasparov.[3] Twenty years later, Google’s AlphaZero program took a mere four hours to learn the rules of chess and then beat the world champion chess program, Stockfish 8.[4] Professor Yuval Noah Harari, author of the bestseller Sapiens: A Brief History of Humankind, responded to the development by writing, “For centuries, chess was considered one of the crowning glories of human intelligence. AlphaZero went from utter ignorance to creative mastery in four hours, without the help of any human guide.”[5]
The rise of technology and automation breeds a fear that robots will take our jobs. This concern is not unfounded. Historically speaking, machines and automation have been making jobs obsolete for centuries. For example, 400,000 workers in U.S. factories lost jobs as a result of automation from 1990 to 2007.[6] With this background explained, it seems like the processing ability and work performed by AI is comparable to those of humans.
In 2018, a work of art created by AI was sold at an auction in New York City, for $432,500.[7] Artwork generated by AI through machines, computers, or software, can look just like artwork created by humans, and is treated as such until it reaches the copyright question. The burning question regarding copyright law and policy considerations for AI is: should AI be allowed to exercise the exclusive rights available under the Copyright Act of 1976? Copyright is a form of protection provided by federal law to authors of original works of authorship fixed in a tangible form of expression. AI like AARON[8], a computer program that creates original artistic images, and Portrait AI[9], an app that paints 18th-century-style portraits, can create works of visual art, which fall under the copyrightable subject-matter category of pictorial, graphic, and sculptural works.[10] Now that we have established that AI can create works that are eligible for copyright protection, what must be explored is whether AI can be deemed an author, and therefore be eligible to copyright its original and fixed works.
Non-human authorship by a computer presents a cutting-edge question about copyright ownership. Courts have not yet addressed the novel legal questions that AI authorship poses, however, similarly novel legal circumstances in the field of copyright authorship suggest non-human authors cannot, and should not be able to, copyright their work. Consider the famous monkey selfie case. When nature photographer David Slater left his camera unattended in a reserve, a monkey named Naruto allegedly took several photographs of himself with Slater’s camera.[11] When Slater later published Naruto’s photos, PETA filed a lawsuit against the photographer on behalf of Naruto. PETA’s complaint alleged copyright infringement on the basis that the monkey, rather than the defendant, owned the copyright on the photographs. The Ninth Circuit considered the elements of the claim under the federal Copyright Act and concluded the provisions under the 1976 act persuaded the court “against the conclusion that animals have statutory standing to sue under the Copyright Act.”[12] In the future, when a case appears before a court to decide whether a non-human computer is allowed to exercise exclusive copyright rights, the decision by courts to decline copyright protection to non-human animals might persuade the court towards a similar outcome for computers as well.
There is a growing policy consensus that AI generated artistic outputs should not be copyrightable. One can argue an algorithm or process should not qualify as a work of authorship protectable under copyright law, because it does not involve a human contributing to the resulting work. But excluding an algorithm’s inputs and outputs is not a necessary conclusion, it is a policy choice. Although advancements in technology are commonly praised as being efficient and productive, there are many critics who believe computers cannot and should not be authors. HAL 9000 had a matter-of-fact tone when he said he thinks all that any conscious entity can ever hope to do is to put themselves to the fullest possible use. What started with impressive AI advancements in chess and optimistic feelings by humans, evolved into equally impressive artistic advancements yet pessimistic feelings by humans. In both cases, AI was simply fulfilling HAL 9000’s dream—putting itself to the fullest possible use.
Footnotes
[1] http://homes.sice.indiana.edu/nensmeng/files/Ensmenger2012-Chess.pdf (chess, like the common fruit fly, is an accessible, familiar, and relatively simple experimental technology that nonetheless can be used productively to produce valid knowledge about other, more complex systems).
[2] Id.
[3] https://www.ibm.com/ibm/history/ibm100/us/en/icons/deepblue/
[4] https://www.theguardian.com/technology/2017/dec/07/alphazero-google-deepmind-ai-beats-champion-program-teaching-itself-to-play-four-hours
[5] https://www.theatlantic.com/magazine/archive/2018/10/yuval-noah-harari-technology-tyranny/568330/
[6] https://time.com/5876604/machines-jobs-coronavirus/
[7] https://www.nytimes.com/2018/10/22/arts/design/christies-art-artificial-intelligence-obvious.html
[8] https://www.katevassgalerie.com/blog/harold-cohen-aaron-computer-art
[10] 17 U.S.C. § 101.
[11] Naruto v. Slater, 888 F.3d 418, 420 (9th Cir. 2018).
[12] Id. at 426.