The Robots Are Coming, the Robots Are Coming

Source: Image produced by Midjourney

Artificial Intelligence and Visual Art

Artificial intelligence (AI) will insinuate itself in healthcare, education, transportation, commerce, politics, finance, security, and warfare (Ford, 2021) and replace many human jobs. AI is well suited to take over routine tasks, but will human creative endeavors be spared? In exploring this question, I recently concluded (Chatterjee, 2022) that AI will deeply affect how we think about art and what it might mean to be an artist.

Art, when most powerful, can evoke deep emotions and promote a new understanding of the world. The vast repository of online images on platforms like Instagram, Facebook, and Flicker is labeled with people’s preferences. These rich stores of data, growing every day, means that AI programs can be trained to identify patterns in images that people like and recapitulate these patterns in their productions.

AI need not be confined to producing banal images that reduce to a low common denominator, like liking. Historically, in empirical aesthetics, scientists have asked participants whether they like a work of art, find it interesting, or consider it beautiful. Imagine a taxonomy of words that describe the potential impacts of art that include general assessments such as “liking” and “interest” (Chatterjee & Cardillo, 2021) but also nuanced terms like “moving,” “challenging,” or “elevating,” or “disgusting” that capture the emotional range of art engagements (Christensen et al., 2022).

A large database of art images labeled by such impacts could serve as a training set for an art-appreciating AI. The objection that AI does not know what it is looking at or experience emotions is irrelevant. All it needs to be able to do is label a novel image with descriptions and impacts- a more complex version of labeling an image as a black cat, even if it has never seen that particular cat.

Sequence transduction or transformer models are making strides in processing natural language. The ability to use text to display images is an important step in producing artistic images. Dall-E 2, Imagen, and Midjourney are gaining popularity as art generators that produce images when fed prompts (Kim, 2022). As an example, I used Midjourney to produce this image below with the prompt “a still life with fruit, flowers, a vase, dead game, a candle, and a skull in a Renaissance style.”

Source: Image produced by Midjourney

Ambiguity often fuels an artwork’s power, forcing viewers to ponder what it might mean. This ambiguity is a strength for AI-generated art, which need not worry about accuracy in the way that is imperative for an autonomous car to recognize children playing in the street. There is no right answer in art.

Most theories of creativity include divergent and convergent thinking (Cortes et al., 2019). Divergent thinking includes coming up with many possibilities. When confronting a problem that needs a creative solution, generating many possibilities doesn’t mean that they are the right or the best.

An evaluative phase then narrows the possibilities, converging on a solution to identify a useful path forward. Artists presumably shift back and forth between divergent and convergent processes as they work towards their final product.

Artists could easily use AI as a tool (Kim, 2022), as artists have used the technology of their times (Hockney, 2001). They could type their intent for a picture and then evaluate the images generated. They might choose which of many images lie closer to their intent, expand their own imaginings, and select or further modify them. The generative phase of creativity would be informed by AI; the role of the artist would be to choose the input text and then evaluate the output for the final image.

ChatGPT

If AI-generated images can be used as a tool to spur creativity in visual artists, can thinkers and writers use AI to generate and consider new ideas expressed in language? ChatGPT suggests that using AI to generate ideas is possible. ChatGPT is a state-of-the-art language processing AI model that was just released by OpenAI for public use. It generates human-like text in response to queries. The model was trained using text databases with 570GB of data obtained from books, web texts, Wikipedia, articles, and other pieces of writing on the internet.

Artificial Intelligence Essential Reads

Tech Interface and AI Enable Stroke Survivor to Communicate

Fulfilling These 3 Human Needs Can Ease the AI Revolution

For example, when posing the following question:

In 500 words, speculate about the future role of AI in generating art.

I received a response in less than a minute that is eerily similar to my recent peer-reviewed paper (Chatterjee, 2022), which took over three months from submission to acceptance. As I did, the ChatGPT response also speculated about a fully autonomous AI art maker. ChatGPT answered:

The future role of AI in generating art is a topic that is ripe for speculation. It is clear that AI has the potential to revolutionize the way that art is created and that it could open up new avenues for artistic expression.

One possible future scenario is that AI will become a tool that artists use to create new works of art. For example, an artist might use an AI program to generate a rough draft of a painting, which they could then refine and perfect using their own skills and creativity. This could allow artists to experiment with new styles and techniques and to create art that would be impossible for a human to create on their own.

Another possibility is that AI will become a fully autonomous art creator capable of generating original artworks without human input. This would require the development of AI systems that have a deep understanding of the principles of art and that can use this knowledge to create aesthetically pleasing works of art. It is likely that such AI systems would be able to generate a wide variety of art, from traditional paintings and sculptures to more experimental forms of art, such as installations and digital art.

In either case, the integration of AI into the world of art will likely have a profound impact on how art is created and experienced. It could lead to the creation of new forms of art that blend the best of human creativity with the power of AI and could even change the way that we think about what it means to be an artist. Ultimately, the future of AI in art is an exciting and unpredictable one, and it will be fascinating to see how it develops in the coming years.

Conclusion

As AI continues to insinuate itself in most parts of our lives, it will do so with art and possibly literature. The hints of art production and writing samples from AI now might be like the video game Pong that was popular when I was in high school. Pong is a far cry from the rich, immersive quality of games like Minecraft in the same way that Dall-E and Midjourney images and ChatGPT verbiage might be a far cry from a future “cultured” machine.

The idea that creative pursuits are an unassailable bastion of humanity is untenable. For many people, AI will be a powerful tool and a partner in creativity, especially when there is no “right” answer. Aesthetically sensitive machines will challenge our views of beauty and creativity and perhaps our understanding of the nature of art.