There is nothing I enjoy more than having a spirited debate about philosophical ideas infused with the emerging state of technology and design. I love how opposing ideas, when shared with mutual respect, can provoke deeper thought and challenge one's own perceptions.
Recently, at an industry leaders event, I had one of those moments in a discussion about whether AI could ever truly be as creative as humans. During this discussion, I became more convinced than ever that the gap between human and artificial creativity is not as vast as many believe. Our creativity, often hailed as uniquely human, is increasingly mirrored by AI through its multi-modal data processing capabilities. With that in mind, I thought I'd jot down some thoughts on how AI's methods of generating creative outputs bear striking similarities to our own, challenging the myth of human uniqueness.
Before making comparisons, we must first understand human creativity-or, more specifically, my personal view of what makes us creative. Human creativity is a complex interplay of experiences, sensory inputs, and cognitive processes. Every conversation we have, every opinion we share, everything we write, draw, or build is based on our past experiences. These experiences might involve formal education or knowledge-building activities through reading, watching, or listening, or simply living our lives and absorbing multi-sensory stimuli from our surroundings. Our senses-sight, hearing, taste, smell, and touch-feed our brains with a constant stream of information, which we process, integrate, and use to form new ideas and concepts.
Take, for example, a chef creating a new dish. The chef's experiences with different cuisines, their sense of taste and smell, and their knowledge of ingredients all come together to inspire a unique culinary creation. Similarly, an artist's work is influenced by what they see, hear, and feel, and the techniques they've learned over the years. All of that input translates into their ability to create art. Even writing this article right now and the opinions I have on the subject are based on my own acquired knowledge and the conclusions I've drawn from it.
"The chef's experiences with different cuisines, their sense of taste and smell, and their knowledge of ingredients all come together to inspire a unique culinary creation."
Of course, the 'nature versus nurture' debate suggests that much of what humans do is down to our genetics rather than purely learned behaviour. Factors like what we refer to as 'natural talent,' where a child is particularly gifted in a subject or skill at a very young age compared to their peers, certainly support an argument for human ingenuity.
Having established (my opinion on) how humans think and draw on creativity, let's explore how modern AI systems, particularly those equipped with multi-modal input capabilities, function in a surprisingly similar manner.
At the most basic level, an AI system works by ingesting vast quantities of data, identifying patterns, and building correlations before establishing certain if-this-then-that type rules. By having billions of data points and associated rules, and having access to Natural Language Processing (NLP)-the ability for a computer to translate the nuances of how humans talk into specific instructions and vice versa-AI can emulate 'intelligence' by understanding the context of a question and responding in a meaningful way.
"The multi-modal approach allows AI to integrate different modalities of data to learn, analyse, inform, and produce outputs that can be remarkably creative."
The latest iterations of 'multi-modal' AI models can not only understand text but also process visual data from images and videos, understand and generate audio data, and analyse textual information. This multi-modal approach allows AI to integrate different modalities of data to learn, analyse, inform, and produce outputs that can be remarkably creative-much like how our senses work together to inform our own creativity.
What's particularly interesting about recent developments in multi-modal AI is its ability to see and experience the human world. Just like humans use their senses to see, hear, touch, and feel their surroundings, AI models can take multiple modalities of input to create a more well-rounded picture of the context of a situation and consider that when responding to a query. The parallels between human senses and AI's data inputs are clear:
Sight vs. Visual Data:
Just as humans use sight to gather visual information, AI uses computer vision to analyse images and videos.
Hearing vs. Audio Data:
Humans rely on hearing to process sounds and speech, while AI systems can take audio input, analyse it, and understand its meaning.
Reading vs. Textual Data:
Both humans and AI read and comprehend text, using it as a foundation for generating new ideas and content.
Touch/Smell/Feel vs. Sensor Data:
There is significant development in artificial sensors and their data being fed into AI models. Through a combination of textural, olfactory, and environmental sensors, AI can emulate similar data to what humans perceive as their senses, providing a form of spatial and tactile awareness.
Both humans and AI use their respective inputs to recognise patterns and generate novel outputs. Human creativity often stems from recognising and reinterpreting patterns in our sensory experiences. For example, a musician might draw inspiration from the sounds of nature, integrating these patterns into a new composition. A doctor might draw upon their experience of recognising certain patterns of symptoms in past patients to arrive at a diagnosis. In each of these cases, humans rely on recognising patterns from past experiences to deduce a possible outcome in a current situation.
Similarly, AI excels at pattern recognition. Machine learning algorithms analyse data to identify patterns, which can then be used to generate new and innovative outputs. This ability to learn from data and create something new is at the heart of AI creativity. There have been various examples in recent years where AI models have been trained with the works of specific artists to produce entirely new pieces that are indistinguishable to the average person.
Of course, with any new technology, there will be questions of abuse. Since the recent explosion of AI growth and usage, some of the most creative industries like art, music, and film are all grappling with the conflict between AI tech start-ups wanting to push boundaries and the incumbents pushing back.
The recent lawsuits by Scarlett Johansson against OpenAI for unlawfully using her voice to train the ChatGPT voice assistant, the writers' tussle with Book3/Pile, and more recently, the lawsuit by three of the largest music labels against Suno AI and Uncharted Labs for copyright infringement-all underline the same issue: AI models are being trained with copyrighted material without prior consent or any commercial returns to the original artists.
As we acknowledge AI's creative capabilities, it's essential to consider the ethical implications. Questions about authorship, originality, and the value of human creativity versus AI-generated works are crucial. Moreover, as AI continues to evolve, we must explore how it can complement human creativity rather than replace it.
The myth that human creativity is unique and inimitable by machines is being steadily debunked. AI, with its ability to process multi-modal inputs and recognise patterns, mirrors human creativity in profound ways. As we continue to develop and integrate AI into creative fields, we may find that the line between human and artificial creativity becomes increasingly blurred.
So, the next time someone argues that AI can never be as creative as humans, consider how both systems-human and artificial-rely on similar processes to turn inputs into innovative, creative outputs. The future of creativity is not just human; it's a harmonious blend of human ingenuity and artificial intelligence. Personally, that's the future I'm rooting for-whether it's Jarvis from Iron Man or Samantha from Her, I'm looking forward to a (very near) future where I can work with AI as a genuine, smart, and capable personal assistant that knows and understands me, my personality, my likes, dislikes, and preferences, and truly helps me be better and more efficient – just in a more human way.
Usman is a digital veteran and a renowned expert in human-centric innovation and product design. As the founder of Pathfinders, Usman works directly with businesses to uncover disruptive opportunities and helps create tangible business value through design and innovation.
Book an initial consultation with Usman »