The AI-Human Creativity Paradox: Artificial Intelligence as a Spark to Human Creativity
Creativity is one of the holy grails of artificial intelligence(AI). For decades, transcendent cognitive abilities such as creativity or aptitude have been seen as a frontier for AI systems. Nowadays, we are regularly seeing new examples of what can be considered sparts of creativity in AI systems. From the many examples of AI agents creating forms of art to DeepMind’s AlphaGo “creative” strategies in recent matches( the famous move 37 ), AI systems are starting to exhibit behaviors that “simulate” human creativity. However there is another angle of the relationship between AI and human creativity that is not explored often and is the fact that AI has proven to spark human creativity on different disciplines.
The quest for unlocking creativity in AI agents is a fascinating one but there is another aspect of the relationship between AI and human creativity that is equally important and not discussed very often and is the fact that use of AI has proven to spark human creativity in different cognitive disciplines. There are many examples in cognitive studies that show that as AI agents get better in “non creative” tasks such as knowledge gathering or tactical problem resolution, humans will feel more confident to be more creative on specific disciplines. I like to call that relationship the AI-Human Creativity Paradox.
What is Creativity?
Recently, I’ve written extensibly about creativity in AI agents (check out previous posts) but don’t worry that I don’t plan to bother you with neuroscientific definitions ;) A very simple way to define creativity is as the ability to recognize patters despite differences in detail and context.
Creativity is one of the cornerstones of human cognition and an ability that has been essential to the evolution of human knowledge. Some of the most important laws and theories of human’s existence such as Darwin’s, Newton’s or Einstein’s were created in sparts of creativity, Now AI’s capabilities ave the opportunity to broaded the horizons of human creativity.
The AI-Human Creativity Paradox
The AI-Human Creativity Paradox explains that in a collaborative scenario between humans and AI agents, humans tend to become more creative as the knowledge of AI agents becomes more complete. The paradox can be seen as a specialized case of the famous Moravec theory. In 1988, well-know scientists Hans Moravec stated that “computers tend to be good on tasks in which humans aren’t and vice versa”. There have been a lot of debate about the Moravec’s paradox over the years but some examples still hold true.
An Example from Chess
In order to illustrate the AI-Human Creativity Paradox let’s take an example from Chess. For the last 20 years, computer systems have regularly outplayed chess grandmasters. However, most experts agree that chess AI systems are not really creative but rather good on finding complex technical manipulations. For instance, chess software systems have no notion of strategy but they excel on positional tactics.
As a result of the mainstream adoption of chess systems, many players have reduced their strategies to repeat moves learned from computers during training sessions. After all, if chess computers rarely make mistakes, how could you loose? Not surprisingly, the number of draws in chess tournaments have increased exponentially and highly solid openings such as Reti or Caro-Kahn have become a favorite of young grandmasters.
Based on the previous explanation, you might think that competitive chess is getting really boring. Far from it! The fascinating thing is that, while second-tier tournaments are plagued with matches using computer-dictated openings and strategies, top chess tournaments have seen an explosion of theoretical “novelties” or creative strategies that deviate from the mainstream theory. There are two fundamental factors contributing to this phenomenon. firstly, top chess grandmasters assume that their opponents are overly prepared in the main lines of an opening so they try to surprise them using not-very well known strategies. But how are they coming up with those creative strategies? Very simply, are top grandmasters rely on chess computers in their preparations, they feel encouraged to explore new, out-of-the-book strategies because they know that computes will always find the best corresponding tactical position to backup their strategy. In other words, the fact that chess systems excel in non-creative tactical manipulations is pushing grandmasters to be more creative strategically; a clear example of the AI-Human Creativity Paradox.
In an environment obsessed with the human vs. machine scenarios, the AI-Human Creativity Paradox is an example that the collaboration between humans and AI agents can push transcendent cognitive skills such as creativity to the next level.