Artificial intelligence challenges what it means to be creative
When British artist Harold Cohen encountered his first computer in 1968, he wondered if the machine could help solve a mystery that had long puzzled him: how do you look at a drawing, a few small squiggles, and see a face ? Five years later, he designed a robotic artist called ARON to explore this idea. He gave him ground rules for painting and how body parts are depicted in portraiture – and then left him free to make art.
Not far behind was the composer David Cope, who coined the phrase “musical intelligence” to describe his experiments with AI-powered composition. To cope told me once that as early as the 1960s it seemed “perfectly logical to do creative things with algorithms” rather than painstakingly hand-draw every word in a story, note in a musical composition, or brushstroke in a painting . He first tinkered with algorithms on paper, then in 1981 switched to computers to help solve a case of composer blocking.
Cohen and Cope were among a handful of eccentrics pushing computers to go against their nature as cold, calculating things. The still nascent field of AI focused on solid concepts like reasoning and planning, or tasks like playing chess and checkers or solving math problems. Most AI researchers balked at the notion of creative machines.
Slowly, however, as Cohen and Cope published a flood of academic papers and books about their work, a field emerged around them: computational creativity. It included the study and development of autonomous creative systems, interactive tools that support human creativity, and mathematical approaches to model human creativity. In the late 1990s, computational creativity became a formalized field of study with a growing cohort of researchers and eventually its own field. newspaper and annual event.
Soon – thanks to new techniques rooted in machine learning and artificial neural networks, in which connected computing nodes attempt to mirror how the brain works – creative AIs could absorb and internalize real-world data and identify patterns. and the rules they could apply to their creations.
Computer scientist Simon Colton, then at Imperial College London and now Queen Mary University of London and Monash University in Melbourne, Australia, spent much of the 2000s building the mad about painting. The computer program analyzed the text of news articles and other written work to determine sentiment and extract keywords. He then combined this analysis with an automated search on the Flickr photography website to help him generate pictorial collages in the mood of the original article. Later, the painting madman learned to paint real-time portraits of people he encountered through an attached camera, again applying his “mood” to the style of the portrait (or in some cases refusing to paint whatever either because he was in a bad mood).
Similarly, in the early 2010s, computational creativity turned to gaming. Artificial intelligence researcher and game designer Michael Cook dedicated his Ph.D. thesis and first research associate work at Goldsmiths, University of London to create ANGELINA — who made simple games based on news articles from The Guardiancombining news text analysis with hard-coded design and programming techniques.
Around this time, says Colton, AIs began to look like creative artists in their own right – incorporating elements of creativity such as intentionality, skill, appreciation and imagination. But what followed was an emphasis on mimicry, as well as controversy over what it means to be creative.
New techniques that excelled at classifying data with a high degree of accuracy through repeated analysis helped AI master existing creative styles. AI could now create works like those of classical composers, famous painters, novelists and more.
An AI painting inspired by thousands of portraits painted between the 14th and 20th centuries sold for $432,500 at auction. In another case, study participants had difficulty differentiate the musical phrases of Johann Sebastian Bach from those created by a computer program called Kulitta who had been trained on Bach’s compositions. Even IBM had fun charging its AI system Watson with analyzing 9,000 recipes design their own kitchen ideas.
But many on the ground, as well as onlookers, wondered if these AIs were really getting creative. Although sophisticated in their mimicry, these creative AIs seemed incapable of true innovation because they lacked the ability to integrate new influences from their environment. Colton and a colleague described them as requiring “a lot of human intervention, supervision and highly technical knowledge” in producing creative results. Overall, as composer and computational music researcher Palle Dahlstedt puts it, these AIs have converged on the average, creating something typical of what already exists, while creativity is supposed to drift away from the typical.
To take the step towards true creativity, Dahlstedt suggestedAI “should model the causes of music, the conditions of its birth – not the results”.
True creativity is a quest for originality. It is a recombination of disparate ideas in new ways. These are unexpected solutions. It can be music, painting, or dancing, but also the flash of inspiration that helps advance the order of light bulbs, airplanes, and the periodic table. In the opinion of many in the field of computational creativity, this is not yet achievable by machines.
In the past few years alone, creative AIs have developed into style invention – into individualized rather than imitative authorship that projects meaning and intentionality, even if none exists. For Colton, this element of intentionality – a focus on the process, more than the end result – is the key to creativity. Corn he’s asking himself though meaning and authenticity are also key, as the same poem could lead to vastly different interpretations if the reader knows it was written by a man against a woman against a machine.
If an AI lacks the self-awareness to reflect on its actions and experiences, and to communicate its creative intent, then is it truly creative? Or does creativity still belong to the author who fed it data and directed it to action?
Ultimately, moving from attempting thinking machines to attempting creative machines can transform our understanding of ourselves. Seventy years ago, Alan Turing – sometimes described as the father of artificial intelligence – devised a test he called “the imitation game” to measure a machine’s intelligence against cheers. “Turing’s Greatest Insight”, writing Joel Parthemore, a philosopher of technology from the University of Skövde in Sweden, “deceives himself in seeing digital computers as a mirror through which the human mind could view itself in a way that was not possible before”.