Throughout the course of “DTC 338: AI (Artificial Intelligence) in the Arts”, learning the capabilities of artificial intelligence and how it can be used to bring creativity into life has been a fun learning process to be a part of. AI can augment the capabilities of artists by offering tools that can simulate complex physical processes which can help assist artists to visualize ideas quickly that can begin the jumpstart of an artist's creative process. Utilizing AI tools in this way has been less time consuming and has allowed creativity and inspiration of an artist to be expanded. Users having access to tools like AI can also bring limitations in augmenting the creative process. One of the limitations I was constantly challenged throughout the semester in this course was generating images that were consistent all together with an overall theme, style, and appearance. The way AI generated art is generated is through the algorithms and data it has access to produce art from a given prompt. The technique of language a user prompts the AI with is most often not the best. The language AI is prompting images from can be too broad if a user does not correctly specify details and needs special technique language to be approached from, for an AI to produce art in the direction where a user desires. This was a learning experience throughout the semester as I was new to utilizing AI to generate art and now as we near the end of the course, I have developed strategies in the way I approach AI to generate any generative art.
The question I want to explore for my final project is "How is AI transforming the traditional art of storytelling? (particularly in character development and plot visualization?) And just how much AI can help bring a fictional world to life through its tools?
My creative project inspiration comes from the assignments from the course that involved creating a fictional world. With this assignment I was able to explore creating a fictional world with assistance of AI, specifically Chat GPT where I was able to train a GPT so it could assist me in a more accurate way as it had the correct context behind the fictional world I was building. With the assistance of ChatGPT, I was able to establish a fictional world with a well-defined era, setting, environment, food, culture, economy, religion, education, technology, and uniqueness of this fictional world. Having these key elements of this fictional world be well defined and given to the trained GPT, it was able to be trained to assist me to generate visuals and creatively as I configured my fictional world. ChatGPT was trained with context surrounding my fictional world where I could just ask it for suggestions creatively, and it would be able to offer insights and support as I navigated through building my fictional world in the direction I envisioned while receiving the assistance from AI.
The DTC 338 course helped guide me in being able to create a fictional world with guidance of what to prompt into the AI to help guide what my ideal fictional world would be. But in the real world, this more than likely will not be the case. As someone who often has a tough time expanding my creativity without having visuals to guide my thinking, the AI tools I have learned about in the brief time of the course, have helped expand my creativity in the arts. An example of this is with creating visuals of characters for my fictional world, I was able to be inspired in creating story ideas just by seeing the visuals that were generated from AI. I think with AI tools becoming accessible, artists who may be stuck in creativity or inspiration due to not having the resources to be able visualize their ideas, these artists will be able to now use AI tools to take the initiative of visualizing their ideas and gain the first step to push their creativity further. By taking this course, I have already gained much inspiration by being aware of what these AI tools can help artists push their creativity. An example is becoming interested in how AI can assist with storytelling and how it can provide visuals which can help push boundaries with the way you begin to develop stories, similarly to how I created my fictional world. Before this course, personally I really struggled with storytelling and visualizing characters and their fictional worlds, but with the AI tools I have learned about I now can easily input a small prompt of my ideas and generate a visual which can help guide where I take my story or characters to. These are some examples of just how AI tools can be able to push creativity and inspiration to build the process of art and storytelling.
For my creative project I am building on from my previous fictional world but with a twist of generating an art gallery display of ice castles and sculptures from my fictional world that are realistic as if they were in a real-life museum. My vision with this creative project is that this would mirror how a fictional world setting can be visualized and be generated to be displayed in a setting where people could go learn about the era, setting, environment, food, culture, economy, religion, education, technology, and uniqueness of this fictional world in a setting that can seem realistic. My reasoning behind this is that artists could utilize these same AI tools to help visualize a fictional world of their own and help create visuals of how their fictional world or characters would look. This project generated images from my previous fictional world work and with the correct technical language and consistency generated images that were created into realistic looking sculpture artifacts from this fictional world, fit to be placed into a museum. In my mind this is like how art museums operate in the placement of historical artifacts and visual displays. With this creative project I am creating that same historical visual for my fictional world, which can be proof of just how a fictional world can be completely turned into a visual which can serve as a guide to be built on furthermore such as a story. These images were all generated with the help of ChatGPT and with the help of Eleven Labs, a sample dialogue was generated that might be coming from speakers or delivered by tour guides as they lead visitors through the ice castle sculptures exhibited in the futuristic museum dedicated to the Kingdom of Glacium.
During the 338 course, one of the crucial issues I struggled with AI was generating images that were consistent in the aesthetic, style, and overall theme to create a harmonious story. So, for my final project my objective was to be able to generate images of my fictional world where I could be consistent in style to tie my story together while creating a scenario that could be realistic in my fictional world.
During the course, I was able to create an entire GPT that relied solely on my fictional world’s environment. Using this GPT, I was able to build and revise its instructions that influenced the way it generated images and prompts to my fictional world. This GPT fed environmental factors about my world, culture, technology, and was told how to generate the images and what to capture in each image. This GPT throughout the course got trained to perfect the vision I had of my fictional world, and I was able to revise the final instructions for this final project. I told the GPT what my final project description was and how it would be tied to my fictional world and its GPT.
Generating the images for this museum for my fictional world still had its limitations as there were still images that were generated that did not fit my vision. An example was how ChatGPT generated images that were too extravagant for my vision like this.
I wanted a smaller and more interactive museum where people could use touchpads to learn information about the kingdom of Glacium as the displays laid in front of them mimicking how the castles and sculptures looked at the kingdom. So, I had to reword my chat to produce images that fit what my vision was as well as to correctly generate the images to make it seem like they belonged at the same location/scenario in the museum setting. I had to remind my ChatGPT of the description for my museum as it sometimes could go off track, but once I let it know that no that is not what I wanted but rather this, and I fed it a good description, it would go back to generating images that would fit my vision.
An example of the final outcome of my vision with the assistance of AI:
With the help of AI tools like Chat GPT and Eleven Labs I was able to create a museum dedicated to my fictional work, The kingdom of Glacium, that provides visuals of the kingdom and historical background. I was able to maintain a consistent style, color, and overall visual of each image that helped tie the whole scenario together to help establish the scenery for this art museum gallery of the kingdom of Glacium. The generation of AI art pieces is far from perfect, but this project helped display that by knowing how to use and manipulate these tools you can be able to bring to life a fictional world and visually attempt to generate it into a realistic matter. This project helped emphasis how AI tools such as Chat GPT and Eleven Labs can be used to help artists and authors to enhance their creativity and inspiration by assisting in creating visuals for fictional worlds that are in our minds and can be brought to life with tools that don’t have to necessary be as time consuming and easy once you know how to utilize these tools. AI can help transform the traditional art of storytelling, become much easier for artists, and be able to inspire creativity to build exciting stories with tools involving AI.