It’s quite remarkable how far artificial intelligence has come in a relatively short amount of time. Computers, and the software that powers them, can do more today than we could have ever imagined just a few years ago, and they aren’t far from being as smart as—or even smarter than—their creators.
While that’s great in terms of technological advancement, it raises myriad ethical questions. Chief among those questions is intellectual property. If a computer becomes smart enough to create its own computer programs, creative products, and even art, who owns the copyrights to what it creates?
The question of who owns the output of non-human creators was presented recently in a rather strange way. When a photographer travelled to Indonesia and figured out a way to coax monkeys to take a photo of themselves, the resulting images took the internet by storm. But then started the legal problems. As The Guardian reports: “the images became the subject of a complicated legal dispute in 2014, when Slater asked the blog Techdirt and Wikipedia to stop using them without permission. The websites refused, with Wikipedia claiming that the photograph was uncopyrightable because the monkey was the actual creator of the image. The US Copyright Office subsequently ruled that animals cannot own copyrights.”
The monkey selfies—at least in the eye of the photographer—came as a result of “human guidance”, therefore he believed he owned the copyright. As Forbes notes, there is a similarity in that argument to the current state of AI: “Today AI systems are still largely human guided, meaning that even creative algorithms like Google’s Deep Dream are still dependent on the input of a human artist to select both the training images to build the neural network and the image to manipulate. What happens, however, as deep learning algorithms become increasingly capable, eventually operating more and more without human oversight?”
The fundamental question that underpins the AI and IP debate is this: does artificial intelligence have rights akin to those afforded to humans? For now, most AI is what is considered to be “self aware” rather than fully autonomous. Self aware AI can “still be shut down at whim by its corporate owners simply pulling the plug”, whereas once we reach the fully autonomous stage that won’t be the case. In the not-too-distant future, it is conceivable that an AI system could “fork their code in novel ways to create the AI equivalent of a child.” This is where the real tension arises.
For some, the answer to this question is simple: AI is a robot, not a human, therefore there is no moral or ethical ambiguity to speak of. Since “intellectual property law is entirely based on understanding and valuing human intellect and ‘creative capacity,’ the answer is clear cut.” This was similar to the judge’s ruling in the monkey case, where he stated: “There is no way to acquire or hold money. There is no loss as to reputation. There is not even any allegation that the copyright could have somehow benefited [the monkey],” said Judge N Randy Smith. “What financial benefits apply to him? There’s nothing.”
That line of thinking—that a human is a human, plain and simple—may suffice for now, but the truth is that there is bound to be disagreement in the future about the very essence of what constitutes humanity and existence. While a computer engineered by a human to think for itself doesn’t have rights, does a computer engineered by another computer have any? The answer to that question could very well depend on how many science fiction novels you’ve read.