Definition–SinatraMan17

A.I.’s methods of “Creation”

Artificial Intelligence, again and again, is proving to be a major threshold event in our society’s history, the likes of which could be compared to the invention of electricity, the internet, and the following periods of change that ensued. Within the past 5-10 years, technological advancements in machine learning have paved the way for an exponential increase in the capabilities of artificial intelligence. While many applications of this technology have been recognized as having positive effects on the efficiency of our lives, one particular family of AI has arisen sparking worldwide controversy regarding its future: Artificially Intelligent Artists.

To fully understand the philosophical implications of AI Artists, we must first examine the technological breakthroughs which sparked this investigation, and their strengths and weaknesses. 

In November of 2022, the Artificial Intelligence research company OpenAI released a demo of their latest project, “Chat GPT-3”, free for use by the public. The site, which as of March 2023 is still free with the addition of a “Premium” tier, represents the largest advancement in written-language generators of its kind. Up until this invention, the written works of A.I. were laughable in their structure and rhetoric and never came close to emulating the abilities of an educated human writer. In the words of Kevin Roose, a technology columnist at the NY Times, “For most of the past decade, A.I. chatbots have been terrible – impressive only if you cherry-pick the bot’s best responses and throw out the rest.” Chat GPT-3 differs from its predecessors in that it produces coherent and intelligent responses and, if prompted, “artful” ones as well. The concept which powers this technology is called “machine learning”, which is defined by IBM as the process in which “[AI] focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy.” Similar to the way humans learn from teachings and mistakes, A.I. learns from each inquiry inputted into its system. 

Another form of this AI Art technology is that of the visual arts: A.I. painting and photograph generators. Rising in popularity within the last few years are apps such as “Lensa” and “Wonder: AI”. Text can be inputted into one of these programs and an image will be created based on the commands. For example, the prompt “Mona Lisa in the style of Van Gogh” could be requested, and the A.I. would generate this using a technology known as “Convolutional Neural Network”. In the book Machine Reading Comprehension, Chenguang Zhu, a research manager at Microsoft, defines a C.N.N. as an “advanced and high-potential classical artificial neural network model which can tackle and handle higher complexity data, difficult compilation, and preprocessing of data.” An important takeaway within this technology overview here is that CNNs employ “compilation”, an idea that is prevalent when discussing creativity. Moreover, Zhu sums up the uses of this technology as being “used for image processing tasks that involve image analyzing, image recognition, video analysis, [and] segmentation of image.” Basically, the software blends recognized visual patterns associated with each part of your written prompt and compiles each pattern visually based on its vast database of already existing imagery.

The fundamental argument against both written and visual art generators is the fact that they are not “creating” anything from scratch. Nothing is an original work of inspired creativity, but rather a vastly complex act of digital plagiarism. But to understand why this is so, the answer to “what is creativity?” must first be explored.

The broad idea of creativity is impossible to completely define, as it is an infinitely complex and individual concept. However, as defined by myself, works of creativity are born from imagination and inspiration and are heavily correlated with their originality. Sometimes creativity occurs through deeply personal emotions and our reactions to them, or the spark of an idea that comes out of an unsuspecting event or experience. Jeanette Milne of the 2020 British Journal of Nursing, points out that “without curiosity, there is little creation. Clearly, both are needed to generate ideas, patterns, and combinations that can lead to new and innovative products and services.” While her insight is referring to products and services, the principal also applies to art. Curiosity and creativity are needed to produce anything that is completely new and original.

The creative weakness of AI, which is the reason why it will never truly be able to replace the human artist, is that it can never generate a new idea or work of art without stealing from previous works. However advanced the machine learning or CNNs get in their ability to extract stylistic trends and compile them coherently, they will always be doing a form of just that: compiling. Just as an essay whose author copied work from another source, yet changed the language to make it “their own” is considered plagiarism, so too is a work of AI-generated art that does the same. Simply changing the way you express each idea does not change the origin of the idea; the idea is not yours (or AI’s) if it is simply rephrased.

Therefore, if artificial intelligence could somehow possess the ability to create new ideas, it would by definition no longer be an “artificial” intelligence. It would possess a skill known only to living beings: new thought. However close we may seem to this God-like invention, regarding Written-word and Art Generators of the 2020s, it is fundamentally and definitively: fiction. Art of any form is something that is viscerally human and is created by its creator through new thought and inspiration, therefore artificial intelligence, by definition, can never fully replace human artists or create new art.

References:

Roose, K. (2023). The Brilliance and Weirdness of ChatGPT. The New York Times. https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html 

What is Machine Learning? | IBM. (2016). Ibm.com. https://www.ibm.com/topics/machine-learning#:~:text=the%20next%20step-,What%20is%20machine%20learning%3F,learn%2C%20gradually%20improving%20its%20accuracy

Zhu C, Zeng M, Huang X. SDNet: Contextualized Attention-based Deep Network for Conversational Question Answering. arXiv.org. Published online 2019.

Milne. (2020). What is creativity? British Journal of Nursing (Mark Allen Publishing), 29(12), S4–S4. https://doi.org/10.12968/bjon.2020.29.12.S4 

This entry was posted in Definition, SinatraMan. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s