Author: Feraye
Professor Alex Jackson
MFA Visual Arts 3rd Residency
September 1, 2023
As strange as it sounds, the conscious experiences in our brain cannot be found or reduced to some neural activity. Although there are several theories, experiments, and claims by neuroscientists, there is no conclusive evidence to prove that consciousness emerges from our physical brains. Consciousness is a subjective, first-person phenomenon; therefore, it is difficult for science, which is an objective third-person endeavor, to explain it (Schurger and Graziano 1). Consequently, scientists focus on characterizing the kind(s) of neural activity that are reliably correlated with consciousness, instead of trying to explain what consciousness really is or how and from where it emerges. We still don’t know why and how humans and other organisms have qualia or subjective experiences. This is called the hard problem of consciousness in science.
Cognitive scientist and philosopher, David Chalmers, first defined the hard problem in his paper "Facing Up to the Problem of Consciousness" (Chalmers 3) and expanded upon it in his book The Conscious Mind. He explains how the ability to discriminate, categorize, and react to environmental stimuli; the integration of information by a cognitive system; the reportability of mental states; the ability of a system to access its own internal states; the focus of attention; the deliberate control of behavior or the difference between wakefulness and sleep can be explained by neural activity, hence scientifically eventually. He calls these easy problems of consciousness. The really hard problem of consciousness is the problem of experience. When we think and perceive, there is a whir of information-processing, but there is also a subjective aspect. This subjective aspect is experience (Chalmers 3). Our taste buds can make us taste a piece of chocolate but cannot provide the experience we feel while eating that piece of chocolate, which is a private, individual, inner experience. We can never know what another person is experiencing internally until they tell us, which is still limited to symbols such as language, letters, and sounds. We cannot know what it is to be like that person.
Not surprisingly, the study of consciousness became a forefront scientific topic with the advent of Artificial Intelligence. After the Google engineer Blake Lemoine made headlines in 2021—and got himself fired—when he claimed that LaMDA, the chatbot he’d been testing, was sentient, a group of 19 computer scientists, neuroscientists, and philosophers has come up with an approach: not a single definitive test, but a lengthy checklist of attributes that, together, could suggest but not prove an AI is conscious (Finkel). In a 120-page discussion paper, the researchers proposed 14 criteria, and then applied them to existing AI architectures, including the type of model that powers ChatGPT (Butlin et al. 19). Results: none of the AIs ticked more than a handful of boxes; none is a strong candidate for consciousness. In other words, our inner experiences cannot be copied.
My interest in Artificial Intelligence as an artist is about its capacity for making art. If AI does not have consciousness but only intelligence, this means the art it creates can only be replications, not originals. It can be argued that human art is also partly replication, and we are influenced by other artists, but we still cannot know what is going on in an artist’s mind and heart while making that piece of art. That experience cannot be replicated or repeated, because it is a private inner experience, regardless of how the art piece ends up looking.
As artists, we are limited to depicting our inner experiences with tools that can be perceived only by five senses. Music can make us hear joy, a painting can make us see agony, and so on. However, no one can have the exact same experience even if they can understand the concept of the art piece. Joy alone has no definite limit or amplitude. Mark Holder, Ph.D., who leads a research team at the University of British Columbia dedicated to understanding and enhancing happiness, explains that the most common way researchers assess happiness is through self-reports (Holder). This does not fit the definition of empirical evidence. "Empirical evidence includes measurements or data collected through direct observation or experimentation," said Jaime Tanner, a professor of biology at Marlboro College in Vermont (Bradford and Gordon). No one can directly observe what another person is feeling or even thinking, unless that person expresses it with their words or actions. Even then, we are limited with language, sounds, and movement. We experience much more than we are able to report to ourselves or to others.
Physicist, engineer, inventor, and entrepreneur Federico Faggin, who is best known for designing the first commercial microprocessor, the Intel 4004, claims that after dedicating decades of his life to machine learning and AI technologies, he is convinced that AI can never have human consciousness. Simply because, consciousness is a purely quantum phenomenon, unique to each of us in that, according to the quantum no-cloning theorem, it is not reproducible, so no machine can ever recreate it. This means it is not reducible to mechanisms (Faggin). He also explains how consciousness cannot be measured or implemented through classical physics. This is exactly why machines like AI cannot become sentient. And without feelings, emotions, and inner experience, whatever is created cannot be called art, or at least it must be categorized as machine-made art (MMA) or AI-produced art (APA) to make sure it is clearly differentiated from human-made art. Meaning, the art someone is buying from these machines could be easily and quickly replicated, unlike original human-made art.
In conclusion, it is established that consciousness is a private experience, and scientists are not even sure if it is an emergent property of the brain. Furthermore, AI badly failed the most comprehensive consciousness test that is currently available. Inner experience is the main component of artmaking. And, up to now, science cannot prove that AI has inner experience or will ever be able to have it. These facts alone are a good starting point to understand why AI art must be classified as machine-manufactured objects. It is imperative to come up with these definitions if we would like to support the idea of redefining AI art and clearly distinguishing it from human made art. There are many more questions and problems to address on this topic. I will be covering those in the upcoming research papers for the semester. Until then, let’s remember Paul Cezanne’s affirmation, “A work of art which did not begin in emotion is not art.”
Works Cited
Bradford, Alina and Jonathan Gordon. “Empirical evidence: A definition.” Live Science Website. Published on 08 Feb.2022.
https://www.livescience.com/21456-empirical-evidence-a-definition.html
Butlin, Patrick et al. 19. “Consciousness in Artificial Intelligence: Insights from the Science of Consciousness.” Cornell University, Arxiv.org Website. Published on 17 Aug.2023
https://arxiv.org/abs/2308.08708
Chalmers, David J. "Facing up to the problem of consciousness" (1995). Consc.net Website. Accessed 15 Aug. 2023
https://consc.net/papers/facing.pdf
Faggin, Federico “La coscienza, la vita, i computer e la nostra natura.” 21 Oct. 2022
https://www.raicultura.it/filosofia/articoli/2022/11/Federico-Faggin-Irriducibile--10f11b52-4e16-4b30-a8fb-200b353b933d.html
Finkel, Elizabeth. “If AI becomes conscious, how will we know?” Science.org Website. Published on 22 Aug. 2023
https://www.science.org/content/article/if-ai-becomes-conscious-how-will-we-know
Holder, Mark. “Measuring Happiness: How Can We Measure It?” Psychology Today Website. Published on 22 May 2017.
https://www.psychologytoday.com/us/blog/the-happiness-doctor/201705/measuring-happiness-how-can-we-measure-it
Schurger, Aaron and Michael Graziano. “Consciousness Explained or Described?” National Library of Medicine Website. Published on 21 Jan. 2022.
https://pubmed.ncbi.nlm.nih.gov/35145759/
Further readings:
https://royalsocietypublishing.org/doi/10.1098/rstb.2017.0357
Comments