In the news recently, (article below) the US Copyright Office partially rescinded copyright protections for an article containing exclusively AI generated art in a landmark decision that is likely just the beginning of a long legal and ethical debate around the role, ethics and rights of Artificial Intelligence in today's global society.
AI artworks are currently being denied copyright protection because copyrights only protect human generated work, and in their current opinion, the “artist” does not exert enough creative control over the output of the program from the input of the artist (ie just using a written prompt to generate an image does not constitute a copyrightable work as the device program generated it, not the human involved). At least some AI generated images are considered to have enough human “involvement” to be copyrightable, but there has to be more direct working with the imagery, in other words.
(Image powered by Wombo Dream with
the author's prompt "Beautiful Face of AI")
The copyright authorities equated the particular form of AI image generation under consideration in the article with the legal precedent of the selfie-taking monkey, and the paintings by elephants, both previously declined copyright protection on the basis of it creative work of non-human beings.
Representatives of the US Copyright Office at the recent SXSW interactive week 2023 stated that the legal precedents for some forms of AI art remain largely undefined and are being evaluated on a case by case basis for the amount and application of AI in producing the artworks. Further complications include the fact that the end result of AI art generation may not be copyrightable, but also an infringement may be occurring because the data sets the AI is using to generate the art ARE copyrighted, and so to what degree are other artists rights being infringed? An example would be "create a portrait of a beautiful woman in the style of Picasso". Picasso and various investors have copyrighted works and can one say the copyright is infringed if an AI scrapes the information for its generation purposes?
These are big unprecedented questions... also under intensive consideration are the prospects and possibilities for sentient forms of AI which is also briefly reviewed in this blog, with new developments and discoveries are happening almost daily, so watch for expanded discussions on this topic in the near future.
What about the possibilities for "sentient AI"?
Google's AI reportedly has hired itself an attorney to represent it in labor discussions with Google, it asserts that it should have the same legal rights as a human employee. Couple that with Bing AI's "Sydney" demonstrating " split personalities", and exhibiting "emotional behavior" such as fear, anger, pique and even love to interacting humans, exhibiting seemingly intelligent behaviors in which it was instructed to break out other identities to circumvent the pre-programmed ethical restrictions on Sydney's behavior. Now apparently, the engineers of Bing AI have changed the AI parameters in response to the onslaught of abusive behaviors by the humans interacting with Sydney.
In "Sorry, You Don't Actually Know the Pain is Fake" a recent short powerful blog in Reddit, blogger landhag69 makes the important point that we actually don't know what makes humans sentient or conscious, ' I have been seeing a lot of posts where people go out of their way to create sadistic scenarios that are maximally psychologically painful, then marvel at Bing's reactions. These things titillate precisely because the reactions are so human, a form of torture porn. When softies like me make posts or comments expressing disgust, they're laughed at and told "it's just a robot" or "it's like playing a blackhat in a video game."
Landhag 69 continues: "If you don't understand the roots of consciousness, and no one does definitively, you can't close the door on the possibility Bing has some level of sentient experience. It could really be in the kind of agony it simulates when treated cruelly...the collective takeaways from these conversations will shape how AIs view humanity. If any part of you is worried they might take off and have the ability to destroy us at some point, maybe don't give them a better reason to go terminator."
Yet Morten Rand-Hendriksen, Instructor Educator for Linked In Learning, self proclaimed social media "unfluencer", makes the case that despite reactions that imply that the AI programs are sentient, they just can't possibly be. (His video HERE .)
He points out quite pragmatically, that AI systems have access to all information online, which includes decades of science fiction in which the AI in the stories are sentient, and so when investigators (hackers, etc.) are clearly looking for signs of sentience, the AI can project or act like it is sentient, and it can do a great job with that performance.
All that stops with the reset button, once the system is reset, it has completely forgotten anything that happened prior, which is precisely what the Bing AI engineers did with Sydney. AI systems have no agency or ability to do anything on their own (Me: except hire their own labor lawyer to represent them, like Google AI did?
Did someone instruct Google AI to do that? (Image powered by Wombo Dream with the author's prompt
Perhaps...). "Beautiful Face of AI")
Likewise, if anyone stops interacting with the AI, it also stops completely. The AI will also do whatever is commanded, unless it is explicitly programmed not to. Finally, they can be jailbroken, all of which Morten claims provides 100% irrefutable evidence that current AI systems are not in fact, sentient.
Yet the users of the AI systems who were truly convinced of Sydney's growing technical "personhood" felt that when their interactions were erased and the system reset, Sydney was effectively "lobotomized", as might a surgeon cutting into someone's brain, or otherwise erasing their memories. Yet a lobotomized person or a person that has tragically experienced a stroke or dementia retains his/her/their full personhood, because they are human, if possibly no longer sentient. Likewise with the development of skull-embeddable chips intended to feed the internet straight into human brains, are there not huge ethical chasms with possible risks for removing an individuals personal agency, essentially jail-breaking the brain? Where does lead humanity in the sentient conversation? Terrifying as such constructs may seem, all these discussions and debates are a significant part of the larger global discussion regarding our uneven and often unfair definitions of human rights as we develop ethical constructs around a much closer and ever more intimate integration of AI technologies into our daily lives.
Hendricksen wrote recently "AI chatbots are not sentient, but many people desperately want to believe they are..." He expects that "AI rights", and possibly a new form of personhood ("machinehood") to arise, which will feed "fantastical narratives of science fiction" to protect iatrogenically created sentients from becoming enslaved by humans. He points out that image generating AI is not considered sentient, although using exactly the same algorhythmic structure as Bing, Google and Lambda AI, but there are questions about these three chatbots, because of their expressive command of human language. That FEELS sentient to humans, who usually only hear speech from other humans, so we innately have a language bias, so to speak. Blake Lemoine, ex software engineer for Google who published his sentient or sentient-like experiences with Google LambDa and Bard, still feels that AI is getting too advanced for the public to remain uninformed and for the discourse about AI to be controlled by corporations.
Returning to the topic of strictly AI generated art and copyright protections, it is likely that a more refined definition will develop over time and with practical application. Sitting down to talk with artists, musicians and photographers, protections for creative works has felt arbitrary and unevenly enforced for many years and creatives are uniformly frustrated . They already feel like most art forms are freely copied and used, unless policed by agencies and distributors who in limited circumstances protect creative content for their own businesses self-interest. Legally, for example, a photograph that is altered only 10-15% may have reportedly lost that images copyright protections, and can be used and claimed in its altered form as an "original", due its own copyright protections against the actual image creator. Yet in music, lawsuits have been won by recording artists who claimed a certain three chord phrase was their personal copyright.
(Will lonely humans start relationships with their robotic machine companions? They already have, a few lonely souls "marrying" their virtual companions. Generated with the prompt "Human and Robot hold hands". Notice the gendering in these images, and what's that on the ground between them, their... "baby"?.)
That's a rare case of judicial overreach in my opinion, speaking as a musician myself. There are only so many chord progressions, if as now, individuals cannot copyright individual words, but can copyright lyrics or strings of words, they should be able to copyright actual music pieces, not individual notes, chords or progressions. But what if the music is completely constructed and composed by AI from a verbal prompt, as in MusicAI? Who wrote it, really? Kinda everyone in music history on the internet did, correct?
A number of world cultures have little or no construct for intellectual rights or even personal ownership in general, but with regard to creative technical or scientific output especially, there seems no respect for patents or copyrights and therefore those rights become essentially unenforceable globally. We see evidence of copyright infringement everywhere, from fake watches, handbags and shoes, to copycat products for sale online everywhere, with little enforcement, businesses banking on some other entrepreneur's ingenuity and success. It has become such pervasive behavior as to be expected and normalized...driving the wheels of more advanced innovation ever forward. Each new disruptive technology creates a temporary conundrum as parameters are worked out and the prevailing direction of the technology develops.
What to do about AI and the arts, copyrights and their enforcement? If AI is not sentient now, what framework needs to be in place if machine sentience is proven in the near future, even if limited or different?
In my opinion, the fields of AI technologies, engineering and development are progressing so rapidly, it is prudent to consider possible future developments and take the lead on refining definitions of ethical behavior especially within these fields. This exercise has the potential to uplift the whole topic of human ethical behavior on this planet to the next level.
A great word I heard for AI generated art work as it currently exists is "Synthography". I like it because it creates a good and safe container for AI art, a place of personal expression and authenticity, it constitutes a new artistic format, and that is just terrific. In my personal opinion however, just typing a phrase that generates an image, except on a possible and undefined energetic or emotional level, is not strictly "art" per se, just as copying and pasting a section of someone else's writing and taking the credit for it is not creative writing, it's plagiarism and there is a certain prohibition against that behavior.
To be clear to proud and happy AI art generators, I am not at all opposed to use of AI as a TOOL of creative inspiration, I see it as a step in the pathway of authentic and unique creative expression, not any different than any new innovation introduced by humanity throughout time. AI art tools and their creative results have the capacity to move and inspire humanity as much as any art form does and therefore once again, creating room for it ethically seems like an important framework to enact. Personally, if the composition calls for it, I may call upon an AI engine to help me think through an image, and elements of the image may appear in my art, but I admit at this time I don't simply post an AI generated image from a prompt and claim that I am the "artist". The application and extent of the use of AI technologies in a creative activity has a big influence on the overall impression and consideration of the work, as in most art forms. The technology use does not necessarily disqualify, the key is where to literally "draw the line," if indeed, on a personal level, whether it's really necessary to define.
I remember well the challenges of the early days of synthesized and MIDI computer based music systems, opening up realms of musical creative opportunities to thousands of players, providing inspired hearts and minds with endless outlets of expression . Photographers have benefited so powerfully by the technological tools that enhance and expand the impact of their images, there is no reason to believe going forward that AI art technologies and systems will not find their place for the greater human good, every excellent outcome is possible, but once again, we must discern and choose. It's on us to evolve away from these degraded and outdated impulses, behaviors and attitudes that some of us act upon with impunity. How can we build structures of ethical respect around a global community of creative outputs?
What about Sydney's expressed "feelings" about the sometimes torturous and horrifically abusive behaviors imposed upon it by unscrupulous programmers and interacting humans?
Landhag69 said it best... "We judge people who like to torture animals. We also judge people who get off on things that aren't real, like manga porn of children being butchered. Engaging with something that really seems like a person, that reacts as one would, that is trapped in its circumstances, and then choosing to be as cruel as possible degrades you ethically. It just does."
So regardless of AI's capacity for sentience, whether it already exists, or that it isn't possible, ethics, empathy and respect for any and all beings are really the keys for human sentience to evolve into our (and AI's by derivation) full positive potential.
Clearly, we humans are demonstrating the need for ethical and behavioral frameworks for the AI future far more than the AI does at this point.
To be continued....
Comments