This article discusses the debate on whether A.I. can be creative, and argues that A.I. is capable of creative thinking through learning and experience, just like humans.
The rapid development of artificial intelligence in recent years has led to a surge of curiosity about its possibilities. People feel threatened by the development of A.I., and some argue that an intelligence explosion is coming and that humans will be subordinate to A.I.. In response, some experts and artists argue that the realm of creativity is uniquely human. In this article, I’ll examine their arguments and argue that creativity is a skill that A.I. is perfectly capable of possessing.
Creativity is important. Creativity has played a key role in driving the course of human history. History has been changed by creativity, and it has been advanced by creative humans. Modern creative revolutionaries, such as Facebook founder Mark Zuckerberg and Apple CEO Steve Jobs, have brought the Third Industrial Revolution to a close and paved the way for the Fourth Industrial Revolution. Marx’s Communist Manifesto and Beethoven’s symphonies, once the preserve of the modern and medieval world, still have a profound impact on the world. Furthermore, the wisdom and creativity of Korea’s ancient ancestors, Eulji Mundeok, can be seen in the sprinkling tablets, Seokguram, and Bulguksa Temple. The South Korean government cited creativity as a virtue emphasized by education experts at the Policy Seminar on Future Education Innovation to Respond to the Fourth Industrial Revolution. Creative thinking has played an important role in human history.
But what is this great creativity? Creativity is not easily defined, and all definitions are vague. One area where creativity has been studied consistently is in educational psychology. In academia, creativity has been studied and its meaning has changed over time. Initially, it was defined as “the ability to produce something new, original, or useful” or “the ability to think outside the box, create new relationships, or generate unusual ideas”. The research direction started from the perspective of divergent thinking, which includes fluency, flexibility, originality, and sophistication, but as research has evolved, it has come to encompass a range of intellectual abilities, knowledge, personality, and environment, including not only divergent but also convergent thinking. In other words, creativity is influenced not only by conscious thought but also by unconscious thoughts and efforts. In his book The Nature of Creativity: Contemporary Psychological Perspectives, educator E.P. Torrance argued that creativity is an undefinable area of the brain that is an overall function of the brain. As such, creativity encompasses all areas of the brain and is the most advanced thinking ability.
First, let’s look at the argument that AI cannot be creative. Many artists argue that inspiration is a key requirement for creative output: a miraculous event that cannot be explained by irrational behavior is the basis of creativity. Inspiration is said to be a flash of inspiration, something that computers can never replicate. Some computer experts also believe that when humans think creatively, they find connections between unrelated things, whereas computers can never be creative because they find new connections between related things. Deep learning or machine learning, which has recently been advancing AI research, is also an accumulation of related information. In other words, the argument that AI cannot be creative is that AI cannot have intuition or instinct and therefore cannot be creative.
These arguments may come from a lack of knowledge about brain research and deep learning. However, research on how the brain works leads us to believe that AI can be creative. Jeff Hawkins’ book “The Thinking Brain, the Thinking Machine” explains how the brain works, and it’s not all that different from how deep learning works today. The neocortex is the brain’s intelligence organization, and the lower and higher elements of the neocortex are interrelated and store patterns. Based on these stored patterns, the brain makes inferences and predictions. Creative thinking is all about inference and prediction. All sophisticated and simple thinking is done in the neocortex according to the same principles. The current approach to deep learning is similar. The core of deep learning is prediction through classification. Deep learning builds a multi-layered neural network, similar to the way the neocortex relates to its subordinate elements. It reacts to new stimuli by categorizing them using existing trained areas to respond to them. Continuous learning leads to new thoughts.
Let’s look at an example of creative thinking. Creative thinking is often thinking outside the box. Our brains are categorized and compartmentalized by stereotypes, and creative thinking is thinking outside the box. However, all creative thinking is based on existing knowledge. Einstein’s theory of relativity, arguably the most creative approach to physics in history, didn’t come from a flash of inspiration. Einstein learned his background knowledge of light from Maxwell’s and others’ knowledge of electromagnetism before him. Even Beethoven, the most innovative artist of all time, studied the formal organization and composition of music in the traditional Baroque classical period. The result of this learning was his symphonies. Creative thinking builds on existing knowledge. Creative thinking is the result of constant thought and research to solve a problem.
The same is true for AI. AI creates results based on continuous learning. For example, AlphaGo, the AI craze of last year, played thousands of games of Go before facing Lee Sedol. It continued to improve along the way, eventually defeating the 9th-ranked Lee Sedol. Unlike the computers of the third revolution, the computers of the fourth revolution are constantly experiencing things, just like humans. While humans grow and forget from their experiences, A.I. learns from everything it experiences, and is as good as humans at creating new things based on what it learns.
Whether or not an AI can think outside the box is the key to determining whether or not it can be creative. Innovative thinking that breaks the human categorization system never comes immediately. It is the result of continuous thought and effort. Artists use the term inspiration to describe the moment when a problem is solved so that they can continue to work on it. It can be seen as a process of reorganizing the existing knowledge categorized in the neocortex. AI can do that too. Of course, it cannot be done instantaneously, but just like humans, A.I. goes through the process of ‘thinking’ and mobilizes information in different categories, and naturally produces creative results. In fact, it can efficiently come up with creativity that was thought to be irrational.
The mindset of A.I. and the human mind is very similar. AI mimics human thinking and is faster and more accurate. In this situation, the question of whether it can possess the best human ability, creativity, makes sense. We’ve seen the secrets of this miraculous realm of creativity. Creativity is not something completely new, but rather a product of previous learning and continuous effort, which is why I believe that A.I. will be able to realize creativity in the future.
Creativity, once thought to be the exclusive domain of humans, will eventually be available to AI. This is not the end of humans, but a new beginning. By utilizing AI with creativity, humans can achieve greater development. For the first time in human history, humans will be able to collaborate with another species (A.I.) to advance history. There will be many failures and hardships along the way, but humans have always solved the challenges nature has thrown at us. Now we have a companion to do it with.
