The Collage of Creation: How Inspiration Drives Innovation in Humans and Machines

Anyone preparing to make something new almost always gathers equipment. Musicians pick up guitars, painters open sketchbooks and writers seat themselves in front of keyboards. Tools act as extensions of our physical abilities - necessary technologies that help us to generate outcomes. In all cases, we must of course know how to use them, and this takes time - thousands of hours of sustained practice in many cases. 

But then what? When our tools are at hand, and we know how to use them, what magic occurs that leads to unique creations? Many will tell you acts of innovation stem from the soul; a religious interpretation of self that transcends the body. Others will cite ‘natural talent’ or a ‘creative mind’ as the catalysts that spark successful artistic expression. But in searching for what enables us to create, few focus on the indispensable, core factor that feeds our ideas and shapes our visions. Inspiration.

Inspiration is an interconnected network of contextual memories that act as the architect in almost all acts of creation. It’s a force that works silently (but powerfully) behind the scenes; composing songs, envisioning artworks and building stories by curating and amalgamating things we have seen before. But because people fail to see the natural connection between inspiration and outcome in human artists, a misunderstanding has arisen that accuses AI of breaking the law. 

Breaking the Myth

Allow me to paint a hypothetical scenario. A baby is placed in a room and looked-after by a robot. The robot tends to the needs of the baby as it grows, but does not communicate or ‘raise them’ as a human would. They feed them, wash them and clothe them - purely to keep them alive. Importantly, they teach them nothing. The person does not have anything to play with and sees nothing beyond its immediate environment. They are alive, but entirely understimulated, and consequently a blank slate. 

One day, the robot arrives with a canvas, paints and a brush. They put the brush in the (now fully grown) person’s hand - the first thing it has needed to grasp. The robot guides them, dabbing the tip in paint and wiping it across the canvas to make a mark. The person recognises this - they have moved food on a plate to make shapes before; expressing instinctive behaviour to create. While their skills are limited, the person starts to paint a basic image - similar to what a toddler might make. But what do they paint? The robot? The room they live in? The food they eat? Most likely, one, or all of those things.

But it's what they wouldn’t paint that’s important. And the list for this is extensive - infinite in fact. They wouldn’t paint a mountain because they have never seen one before. They wouldn’t even paint another human, because they would have never seen one of those either. They will only ever try to paint what they know. 

Imagination as Collage

As someone who has taught many hundreds of art students over the years, I know that inspiration is the most important pillar of creation. Regardless of the age of the student, or the level at which they are learning, contextual influence is a key aspect of what they do. Research is the foundation of all art & design projects, be it first-hand observation or secondary inspiration found online. Successful creative outcomes are driven by exposure to material that feeds them, and it is within this process that innovation is born.

Meanwhile, the impact of exposure to influence already rings loud and clear in the students that arrive to begin their studies - a perennial battle when guiding them towards original outcomes. That’s because - what they have seen throughout their lives is usually seen in what they make. Dragons feature prominently, as do long-haired men bearing oversized swords. Space sci-fi leans visually towards Star Wars, while aliens are often disproportionately small with almond-shaped eyes. Harry Potter introduced boarding schools and wizardry into children’s book concepts, and Guardians of the Galaxy placed intergalactic animal characters in vogue. I could go on. And what’s more - none of these influences were themselves original in the first place. The fantasy world of dragons and sword-wielding heroes dates back millennia, Star Wars drew influence from real-world conflicts, little-green men have their origins in 1950’s pulp fiction, JK Rowling was inspired by C.S Lewis & Tolkien and - unless my eyes deceive me - Guardians of the Galaxy is inspired by the 1993 Super Nintendo game Starfox . . .

It’s probable that none of this will come as a surprise to you, and yet because it's a topic few outside creative education consider on a daily basis, the integral role of influence and inspiration must be highlighted in the debate surrounding Generative AI. 

Breaking the Law

That’s because there is a strong movement building among creative professionals - one that frames AI art, video and music generation as illegal - copyright infringement at scale. On a surface level, it does seem as though AI is repurposing the work of others for financial gain. This is something that, if true, violates intellectual property laws in most parts of the world. Freely available models such as Midjourney (AI art), Sora (AI video), and Udio (AI music) are capable of creating professional, commercially usable outcomes with the look and feel of work that has gone before. They have been trained on work accessible by the public - art, films and music - that’s protected by law. This enables the AI to create work inspired by human creations. Not only that, but it does so often without the original creator’s consent or giving them a cut of profit in the process. That can’t be allowed. Can it?

There is a LOT of anger around this, and it’s obvious to see why. Out of nowhere a technology has landed disrupting the creative community in ways that would have been impossible to imagine just a few years ago. It’s immensely unfair and definitely feels illegal. But what we are experiencing is the evolution of a new and highly disruptive technology, played out at lightning speed. 

AI Doesn’t Copy - it Learns

Despite working and teaching in the creative sector myself, I have concluded that AI is not, in the majority of cases at least, infringing copyright laws. And in order to understand my perspective - one that I know places me in a minuscule minority of practitioners - it’s important to reflect on how AI does what it does. Because it’s startlingly similar to how humans learn to make art, only at scale. 

AI is effectively a machine-equivalent of the human brain; a synthetic mind that has the potential to learn patterns and approaches in a comparable way to how we do. In the case of image production, neural networks such as ChatGPT are fed with data to teach them how to make images; they are exposed to a vast range of pictures made by millions of artists, photographers and creatives. In this process, AI learns broad styles, ways of composing images and - importantly - how humans make them. It does not hold a bank of images from which to copy protected works. It doesn’t use images owned by others as reference material during the act of creation. Instead, it has learned the skill of image creation through what it has been exposed to.

What AI is exposed to differs greatly, depending on the model in question. Some, like ChatGPT, have paid for access to large data pools - a publisher’s catalogue of books for example. Others have been exposed to content available on the Internet in their training. The latter is where people take issue, and I understand why. AI is ‘looking’ at the work of others, and using that work to teach itself about how to do similar things. But clearly, and obviously - that’s exactly what humans do as well. 

Fighting an Unwinnable Fight

There will be people reading this who will disagree with my take wholeheartedly, and that’s fine. Some will dismiss my views, and simply move on. Others will rage at the indignation of my perspective, adamant that I am ill-informed or a traitor to the creative community. There will also be those who believe I have been brainwashed; that I am a cold-hearted tech-evangelist who prizes ‘progress’ over consequence. But I am neither a traitor nor a die-hard supporter of Silicon Valley. I am just someone who has examined the facts and reached a different conclusion. I am also a realist. 

Fighting something that cannot be stopped is - in my eyes - a fruitless task. Anyone who believes that AI will be cancelled, or that generative AI will be curtailed would do well to examine history. To date, no new technology has been stopped - not ever. Nuclear weapons are one of the vilest inventions ever to have been created, and yet right now over 20,000 exist - aimed, and ready to fire, on rival nations across the world. To understand why this is, it’s important to remember that if something can be done, it will be done. However much we dislike the fact (and believe me when I say, I dislike it a lot), this is an immutable tenet of human civilisation. Railing against the onslaught of technology is like shouting at the waves in the ocean - it might make us feel better, but it won’t do a damn thing to stop it.   

Further - creatives will not be reimbursed for the impact AI will have on their careers. Was the publishing industry compensated when the Internet decimated their revenue streams? Has the cinema industry been subsidised by Netflix? Tech moves things in new directions, dissolving business models and disrupting economies in its wake. That’s just the way things are. 

Conclusion

Inspiration drives innovation in humans and machines, with both learning how to make original outcomes through what they are exposed to. As a result, everything we see is a cultural stew; a mix of pre-existing ingredients, cooked up to generate something with a new yet familiar taste. Appreciating this not only highlights the reality of how AI does what it does, but also its immense power and potential to drastically alter society at large. The value of skills learned over years is gradually fading, and that’s because AI can learn to do - at lightning speed - many of the things we do.

There are some whose hatred for AI blinds them to the facts. I have read countless posts on social media claiming AI is a dead-end technology and companies like OpenAI are falsifying its potential to profiteer from copyright infringement in the short term. This, in my mind, is a conclusion born from naivety. Either that, or it’s a misguided attempt to distort the truth. What we are experiencing now is AI in its infancy; a nascent technology that will change the world in myriad ways. These changes will be painful, and they will be disruptive. But refusing to face what’s unfolding, or wilfully seeking to misrepresent reality, will only serve to muddy the water and distort the path ahead.

Understanding the similarities between human cognition and machine intelligence is essential if we are to discover how we can fit into a future landscape built by tech. Appreciating the process by which AI learns, and then reproduces human actions, is a great place to start. That’s because doing so helps us to identify things we are capable of that machines are not. Human creativity will always be of value - more so as AI evolves to generate even better instant outcomes that lack the fizz of human emotion. Human creativity isn’t dead and it never will be. But we must adjust and adapt what we do to counter the impact that AI will continue to have on the creative space.

If you’re a creative working in any field, there is something you can do right now. Not rallying against the progress of a train that’s already left the station. Not complaining about the unfairness of the situation we find ourselves in. Instead, you can reflect on the value of what you do. You will find that the act of human creativity is not replicable - emotions are a core driver for all acts of human creativity, and we have them in abundance. My advice is - capitalise on them, use them, and evolve what you do to prepare for the inevitable and fast-approaching future of creation. 

Watch the accompanying YouTube video, and join the conversation now...