Roger-Luc Chayer (Images: Gay Globe via Copilot & Deepai.org)
In recent years, artificial intelligence (AI) has made significant strides in numerous fields, including the arts. AI’s ability to create, interpret, and even compete artistically has opened new avenues for creativity and innovation. One notable example of this trend is the artistic competition between two advanced AI programs: Copilot and OpenAI. Since we regularly use these two software programs to illustrate the magazine and social media of the Gay Globe Group, I thought it would be interesting to push the experience to the maximum by giving both programs the same instructions to see what would emerge.
The Rise of AI in Art
AI’s foray into the arts has been marked by a variety of projects that showcase its potential to mimic and even surpass human creativity. From generating music and visual art to writing poetry and prose, AI has demonstrated an impressive range of abilities. These developments have not only challenged traditional notions of creativity but also expanded the boundaries of what is considered art.
Copilot: The Creative Assistant
Copilot, developed by OpenAI in collaboration with GitHub, is primarily known for its ability to assist programmers by suggesting code snippets and completing programming tasks. However, its underlying technology, based on the GPT-3 model, is also capable of generating creative content in natural language. Copilot’s strength lies in its vast training data, which includes diverse textual sources, allowing it to produce coherent and contextually relevant outputs.
In the context of artistic creation, Copilot can generate narratives, write poetry, and even compose lyrics. Its ability to understand and manipulate language makes it a powerful tool for writers and artists seeking inspiration or assistance in their creative processes. However, Copilot’s primary function as a programming assistant means that its artistic capabilities, while impressive, are not its main focus.
OpenAI: The Artistic Visionary
OpenAI, the organization behind Copilot, has also developed other AI models specifically designed for artistic purposes. DALL-E, for instance, is a model capable of generating images from textual descriptions. This ability to create visual art based on written prompts showcases a different aspect of AI’s creative potential. DALL-E can combine elements in unique and unexpected ways, producing artworks that range from realistic to fantastical.
The Competition: Copilot vs. OpenAI
While both Copilot and OpenAI’s artistic models showcase impressive capabilities, their competition highlights different approaches to AI-driven creativity. Copilot, with its focus on natural language processing, excels in generating written content and assisting with creative writing tasks. Its ability to understand context and produce relevant suggestions makes it a valuable tool for writers looking to enhance their work.
On the other hand, OpenAI’s models like DALL-E demonstrate a more direct approach to artistic creation. By generating images from textual descriptions, DALL-E bridges the gap between language and visual art, offering a novel way to create and experience art. This capability allows for a more intuitive and interactive form of artistic expression, where users can see their ideas come to life in visual form.
The competition between Copilot and OpenAI’s artistic models has significant implications for the art world. Firstly, it challenges traditional notions of authorship and creativity. If an AI can generate art that is indistinguishable from human-created works, what does this mean for the role of the artist? This question has sparked debates about the nature of creativity and the value of human input in the artistic process.
Secondly, the use of AI in art democratizes creativity by making advanced tools accessible to a broader audience. Artists and writers who may not have had the resources or skills to realize their visions can now leverage AI to enhance their work. This accessibility can lead to a more diverse and inclusive art world, where new voices and perspectives can emerge.
As AI continues to evolve, its role in the art world is likely to expand. Future developments could see more sophisticated AI models capable of creating complex and nuanced artworks across different mediums. Collaborative efforts between human artists and AI could lead to new forms of hybrid art that combine the strengths of both.
Moreover, AI’s ability to analyze and interpret art could open up new possibilities for understanding and appreciating artistic works. By analyzing patterns and trends in art history, AI could provide insights into the evolution of artistic styles and movements, offering a deeper understanding of the cultural and historical context of art.
In this experiment, I asked both systems to create the most beautiful artistic painting on an LGBT theme, and both programs set to work. Personally, I am surprised by the difference in the results. Despite my numerous attempts to break Copilot out of its usual mold, the results are often similar and always resemble candy-colored paintings from the 1960s, without much artistic interest, as you would agree (see the image at the beginning).
With DeepAI.org, the results are spectacularly different. Firstly, rather than giving us a single bland result, the software offers us a hundred different style choices ranging from modern and cubism to romanticism, and more. The image in the middle of the article was generated with DeepAI.org with minimal instructions, and it is truly magnificent; it almost looks like a Gauguin. As far as I’m concerned, DeepAI.org wins this artistic competition on an LGBT theme hands down.
To use DeepAI.org, simply visit the website at deepai.org; the services are free. For Copilot, you need to download the Microsoft Edge browser with the Copilot feature; the software is also free.