Mind Matters Natural and Artificial Intelligence News and Analysis
Film studio with cameras and movie equipment
Film studio with cameras and movie equipment exposing edge of facade

Could AI Authentically Create Anything?

Brendan Dixon: The first question posed to me as an artist was,“What are you trying to say?”

Oxford’s current Professor of the Public Understanding of Science, Marcus du Sautoy, wrote in New Scientist recently that he believes that AlphaGo, the Go-playing AI, demonstrates creativity. In earlier posts, I’ve considered and rejected that and questioned du Sautoy’s criteria for creativity. In this final post, I’ll explore if an AI — as we presently understand such an entity — can ever be authentically creative.

First, what is “authentically creative”? Artists initially rejected photography as a medium: Real artists, the theory went, painted; pretenders took photographs. But as the medium developed, artists grew to accept photography. As I learned in my course with the Rocky Mountain School of Photography, anyone can take a picture; creating a photograph is work. The photograph’s composition, balance, subject, and more must convey the purpose or intention of the artist. The first question posed to me in those classes, when my own attempts were on display, was always: “What are you trying to say?”

AI creates kitsch, not art

In his new book The Creativity Code: Art and Innovation in the Age of AI, Du Sautoy, a Professor of Mathematics at Oxford offers four traits that define a creative work — novelty, surprise, value, and original intention. I noted earlier that creative works can lack the first three, but that the last — capturing the artist’s intention — matters.

If creative work captures an intention, then can an AI—as we presently understand it — independently form intentions which it can then convey?

Du Sautoy suggests, and I agree, that forming an intention requires consciousness:

…we must first ask what drives our own urge for artistic creation. And for me, that is bound up with the hard problem of consciousness: the difficulty of explaining the true nature of felt experience in ourselves and other sentient beings. Because it is impossible to get inside each other’s heads to experience what another person’s pain or ecstasy feels like, we create works of art as a kind of functional MRI scan to reveal our conscious world and share it with others. A novel or a musical composition or a painting is our best way to help gain access to another person’s mind.

Marcus du Santoy, “True AI creativity is coming and will reveal the minds of machines” at New Scientist

AI is bound to its programming; it is not conscious and is not creative. So-called self-modifying programs — such as neural networks that tune their parameters through training — are not an exception. They do so under the hand of the program’s author, embedded in the original code. The difficulty we sometimes face in tracing a neat line from the original code to the final product does not imply consciousness in the program; it only demonstrates the complexity of the events.

Du Sautoy believes, nonetheless, that AI will “in the distant future” achieve consciousness. For that, we have no evidence. It is a statement of religious faith akin to that of Anthony Levandowski’s AI Church.

We can use AI to create things, including art. If the product captures our intentions, it is creative. But the creativity comes from us. Jackson Pollock dripped paint across canvases to create paintings that sell at record prices. No one suggested his sticks and tools were the true artists though some question if Pollock was (see, for example, Jackson Pollock is Trash and Abstract Expressionism Needs to Die). But that is an in-group dispute among artists. Tools remain tools and they do not create. We are the creators and we create those tools, whether they are sticks, brushes, or a complex AI computer system.

See also: Part I: Why AI appears to create things: When AlphaGo made a winning move, it exhibited no more creative insight than when it played pedestrian moves (Brendan Dixon)


Part II: Why AI fails to actually create things Only one of the traits du Sautoy suggests is an essential part of creativity (Brendan Dixon)

Also by Brendan Dixon: Does the Butterfly Effect sharply limit AI’s power?

Brendan Dixon

Fellow, Walter Bradley Center for Natural & Artificial Intelligence
Brendan Dixon is a Software Architect with experience designing, creating, and managing projects of all sizes. His first foray into Artificial Intelligence was in the 1980s when he built an Expert System to assist in the diagnosis of software problems at IBM. Since then, he’s worked both as a Principal Engineer and Development Manager for industry leaders, such as Microsoft and Amazon, and numerous start-ups. While he spent most of that time other types of software, he’s remained engaged and interested in Artificial Intelligence.

Could AI Authentically Create Anything?