Not too long ago, I was asked to present a tool to some of my clients. It was a simple prototype, where a person would type in a few things (i.e., advertising channel, product and occasion), and in turn, the machine would give a number of sample ads. When I clicked the button, in just a few seconds, the machine spat out several ads complete with images and text. The first comment was, “Wow, that was really fast.” What would take a person a few hours to do, this machine did in but a fraction. There were a lot of other interesting comments, some even pointing out that this machine was really creative. Then one person spoke out, a comment that put the room into an uncomfortable silence, “This thing is going to take my job.”
We are in a time of uncertainty. As AI applications become more visible and popular, many will start wondering how they will impact our society. There are the “doomsayers” who think AI will take over the world. Then there are the more “sane” people who think that AI will never be able to replicate humans. After all, how can a machine copy something so intricate and complex? But then again, day by day, the advancements in AI continue to surprise us, as if to challenge our very humanity.
One trait that we hold dear is human creativity. When a person is being creative, it’s as if they put their entire being into what they’re doing. More than experience, they put their thoughts, emotions and, on a higher level, some even say their soul into the work. This creates art in many forms, such as music and poetry. But then again, AI applications today can write music and poems that compete at the highest level. If machines can copy our creativity, what stops them from taking over every other aspect of our lives?
Will Machines Be Able To Emulate Human Creativity?
The notion of a machine copying human creativity is as convoluted as defining what human creativity is. But as complicated as that is, I believe by exploring what creativity means to us, we also learn if a machine is able to copy it.
In my lectures, I always mention that the human mind is an amazing thing. We are able to take a wealth of information and naturally compress this to make it understandable and useful. This makes us very good at a lot of things.
An example is classification problems. After seeing a dog once, a child would be able to identify many other dogs. This, in machine learning lingo, is what we call “one-shot recognition”—a feat that we are very far from achieving with AI. We do this because we have a small amount of memory, probably a few gigabytes worth. We compensate for that by compressing the information we receive and later expanding it by drawing correlations with other memories. It is a self-preservation mechanism to understand what is important and not. It is natural for us to generalize so that we can draw meaning from our surroundings. This is why it’s so hard to build self-driving cars. Our ability to classify and recognize what that means to us is incomparable.
The awe of the human mind is what started the field of machine learning, but what made it into what it is today is the realization that we do not have to copy how humans think to get results. Machines do not have the problem of having small memories. The internet itself can contain and process petabytes of data, which is easily accessed by a machine at incomprehensible speeds. Instead of discarding irrelevant information, it takes all the information and finds patterns within it. From there it draws out likelihoods, and the most probable answers will be its conclusions. This is the same for its creative output, it finds a definition of success and creates what it thinks is closest to it. It doesn’t contain any generalizations or out-of-scope associations—things that, as mentioned before, are important for us to find meaning. Today, a machine’s output is extremely close to being indistinguishable on a human level, but does that make it creative in a human sense?
So it goes back to the definition: Is human creativity about the end result, or is it about the process? If we believe in just the end result, then the answer is yes. But what is the difference between a Picasso and a copy of it? On the other hand, if we believe that what makes creativity special is the perspective of the creator, then creativeness in a machine sense is vastly different. So different that it could probably be called machine creativity as opposed to human creativity.
Will Machines Take Over Professional Creativity?
When we talk about professional creativity, the easiest industry to talk about is marketing, as it is by far the most performance driven, which makes it the most susceptible to highly efficient machines. But even in the most utilitarian rendition of creativity, it still is about making a connection that requires a human touch. Personally, I believe the human perspective will always be important and will have its place in any human-led industry. That’s not to say that machine creativity will not challenge its importance in our society. It will disrupt our place in the working world, and it will make us question ourselves. But as perspective is important to creativity, we will find that a machine’s perspective only adds to our own creativity.
It’s not a direct conflict; on the contrary, understanding a different way of seeing things will only allow us to create even better things. Maybe so much that we actually learn more about what human creativity actually is.
So for those in the creative field wondering if a machine will take over their job, I believe the answer is no ... at least for now.