From Artistic to Artificial Intelligence

The Emergence of the Creative Machine

Creativity has long been perceived to be immune to falling under the dominion of machines. However, this is changing fast. Machine learning and big data now enable intelligent systems to be used in many creative fields, such as music production or the culinary and visual arts. Even though we know that machines and humans are creative in fundamentally different ways, in the future they will be able to collaborate and co-create. However, facilitating this will be extremely challenging.

According to recent studies, 35-40% of jobs are at high risk of being replaced by machines in the next two decades. Tasks that have a high degree of repetition and require precision will be the first to go. But machines will never replicate the creativity, empathy, and personal connection required by many disciplines today. Here, we explore how machine and human creativity are different and why we need both. But first, we need to define creativity and understand the essential difference between creativity in humans and in machines.

What is creativity?

There are many definitions of creativity, from a belief that it is an inherent talent with which some people are gifted to the view that it is a skill that can be built through learning, immersion, and practice. Steve Jobs’ definition of creativity is perhaps the most general definition: “Creativity is just connecting things.”

Based on Jobs’s definition, machines are certainly more creative than humans. Compared to humans, machines are able to collect and make sense of greater amounts of data sets. Ed Rex, a music composer who has created an online platform enabling anyone to compose music with the help of artificial intelligence (AI), agrees: “Creativity is immersion, assimilation, and recombination, and by this definition, it is pretty clear that machines should be able to perform all of those steps: computers can immerse themselves in huge bodies of data, they can find patterns and common features in big data sets (even features that humans would not notice), and they can recombine things in novel ways.”

For us, creativity is more than that. Creativity is the act of turning new and imaginative ideas into reality and making something new or unique. Psychologist Mihaly Csikszentmihalyi, author of Creativity: Flow and the Psychology of Discovery and Invention, says that creativity does not occur in our heads but in the interaction between our imagination and our social context. Creativity is a matter of reflection, experience, and response. It is a matter of relationships and who we are in the world.

Creative thinking involves pushing the envelope and looking beyond the boundaries of the current frames of reference. It happens when you can draw from different fields of intelligence simultaneously. Although machines are able to draw information from different sources and make new connections, they still rely on humans to enhance their creative skills. The question that comes to mind is not whether machines can be as creative as humans, but how their creative process is different.

What is the difference between human creativity and machine creativity?

So far, machines have been able to come up with new ideas, recipes, lyrics, and suggestions for the visual arts. Their source of inspiration is data and patterns, and their creative outcomes are the result. They are not a reflection of personal tastes or preferences. Machine outcomes are a result of making lots of connections and learning over time, and of determining which connections are more desirable. The human creative process is much messier and more complex. Machines do not have emotions like humans (yet). They do not have a sense of self, cannot experience the world like humans, and do not have the power to imagine in the same way. Therefore, human creativity is much different than that of a machine.

Human creativity is bound to our emotions, including fear, grief, and love, as well as our hopes and desires. Emotion cannot be programmed, and people do not feel the same emotions in similar situations. As many researchers and computer scientists have mentioned, the ability to have an emotional response, to be able to pick up on subtle emotional signals from others, and to have empathy and judgment are all things that are hard to replicate in machines’ development. The curving, swirling lines of the hills and sky, the contrasting blues and yellows, and the thickly layered brushstrokes of Vincent Van Gogh’s The Starry Night masterpiece are deep-seated in the artist’s turbulent state of mind, imagination, emotions, and experiences. Would art created by an AI artist be creative in the same way as this masterpiece by Van Gogh? Sure, it would be a creative piece of art, but it certainly wouldn’t be the same as something that has been created by a human artist.

A sense of self is another big part of creativity; machines do not have a unique personality. The fundamental question for any creative machine is whether its abilities derive only from its creator or whether it is capable of independent and surprising originality. For example, consider  AARON, an AI-based machine created by artist Harold Cohen that creates original artistic images. It cannot evaluate its own work, make decisions on what to create next, or self-reflect and modify its work without an external standard – all things that a human can do. Cohen argues, “A robot would have to develop a sense of self. That may or may not happen, and if it doesn’t, it means that machines will never be creative in the same sense that humans are creative.” Perhaps AARON was not creative in the same way that Harold was, but AARON’s creative capabilities certainly opened a new approach to creating art.

As Irish playwright George Bernard Shaw said, “Imagination is the beginning of creation. You imagine what you desire, you will what you imagine, and at last, you create what you will.” According to this definition, machines can become creative once they are able to imagine. Imagination is the creative ability to form images, ideas, and sensations in the mind without direct input from the senses. Creativity is applied imagination. In theory, unless we program imagination and intuition, we will not be able to claim that machines are capable of thinking creatively in the same way that humans do.

Model 1.0: AI will leave the creative tasks to us.

Machines are incredibly accurate in performing repetitive tasks, such as assembling products or harvesting data from a sensor. They are also able to collect and make sense of enormous amounts of data, and to enhance their performance simply by being exposed to more data. This will remove the burden of doing work that is repetitive and straightforward, and will free up humans to concentrate on more creative, complex tasks. Scenarios that involve complex decision making, personal connections, and creativity will be left to humans. It is similar to ATM machines freeing up bank tellers to focus on building personal connections with customers. Hence, we could claim that the job of a bank teller is far more creative today than what it was prior to the arrival of ATM machines. Being able to waive a transaction fee for a customer goes a long way, and that is something that no algorithm will be able to do in the near future.

Model 2.0: A new wave of creativity will emerge as a result of human-AI collaboration.

Ginni Rometty, the CEO of IBM, set the tone on stage at the World of Watson conference in Vegas last year: “It’s about man and machine. Enhancement, not replacement.”

AI is already enhancing human creativity. Musicians are looking to AI to analyze thousands of tracks of music and to derive the underlying tones and lyrics that resonate with people. In 2016, Grammy Award-winning musician Alex da Kid collaborated with Watson to create lyrics and music to a song that was composed after analyzing five year’s worth of data from the New York Times, the most edited Wikipedia songs, and popular movies. The collaboration resulted in the creation of the song “Not So Easy,” which reached the top of the charts on Spotify. Similarly, the platform developed by Ed Rex allows us to tap in a few variables, including genre, mood, and duration, after which the program will create music.

Machines are not only going to be assistants to humans (and artists), but also new, creative collaborators. Imagine when AI gets added to the designer’s inspiration box: Mood boards will be created in a fraction of a second, storyboards with multiple endings laid out, and guidelines on a particular style, size, and feature list provided, leading to an endless array of design possibilities. How might that change the way designers think and work?

The possibilities of human-AI collaboration are not just limited to art and design. Imagine if AI becomes our creative half when it comes to problem solving. With the potential to challenge our assumptions and help get us outside of our tunnel vision, AI can twist our thinking, shift our perspective, and give us a different point of view when we are stuck. More interestingly, we will be able to codify some of the characteristics of creative people. So, we could ask our creative machine to give us different ideas, as if Steve Jobs, Leonardo da Vinci, or Thomas Edison would have thought about them. Simultaneously, we could spit out a few ideas and have the most creative people in human history vet the best ones and help us prioritize. The idea is, of course, not to rely on machines to think for of us, but rather to enhance our cognitive ability.

Model 3.0: Human-machine hybrids will be the ultimate creative species.

This scenario is more radical than the previous ones and takes into account the fact that we are already seeing a hint of how embedding technology in our body might change the way we think and behave. An early example of this is the work of artists like Neil Harbisson and Moon Ribas, who hacked their bodies to acquire an additional sense, such as hearing colors or feeling seismic loads. Although far from a true machine-human hybrid, these examples show how our mind begins to ignore the difference between our biological and electromechanical organs and thinks of them as a unified system. This occurs when cyborgs (humans with technological organs)  connect their minds to a computer and are able to use the best of the two worlds to collect data, to make sense of it, and to problem solve. Imagine what would happen if we gained unlimited mental storage and could review our thoughts and memories as if we were going back in time. In this scenario, the distinction between human and machine creativity would be blurred. This would radically change how we think about technology as a tool and force us to think about it as part of our being.

A promising partnership.

As seen in these scenarios, the emergence of AI and advances in deep learning have resulted in new creative processes and outcomes. Machines are getting better at identifying patterns and human intentions. As a result, they are getting better at interacting with humans, allowing them to become even more creative. With AIs as chefs, rappers, poets, and artists, we will also see a new wave of creativity that is inspired and influenced by humans, but not created solely by them. Change is inevitable. We need to embrace non-human collaborators and adjust our perceptions in the new era; we need to learn to use machines as co-workers, collaborators, and co-creators.

Lastly, we have a bit of advice for any machines reading this article: Improve how you interact with us by reading between the lines and looking beyond the surface. Earn your right as humans’ creative counterparts.

the author

Azadeh Houshmand

Azadeh Houshmand is director of client experience, design & insights at RBC.

the author

Maryam Nabavi

Maryam Nabavi is former VP, IC/ things at Idea Couture.