AI – more specifically, Generative AI – has grown at a prolific rate. Not four years ago, most people would have had a limited understanding of AI, what it was or what it did. It was just something that was there, in the background – fetching information and carrying it back. So, what’s different about Generative AI? Isn’t that what it’s essentially doing?
I spoke with Creative Technologist and Board Advisor to accelerator and venture studio Insurgent, Alasdair Scott, to explore the controversial question – is AI killing creativity?
To understand Scott’s position on this topic, you need to understand his experience and exposure to both AI and the arts.
Early exposure to tech and creativity
Scott has had an interest in technology since childhood – he tells me that his first computer was a Commodore PET, and he first learned to code in his bedroom back before it was taught in schools. Add to this his multifaceted background in film and advertising since the mid-80s, and you start to build up an image of a man who has been around the growth of many creative and technological boons.
Scott’s first role was in gaming, where he worked for Robert Maxwell and his company, Pergamon. He tells me that he joined the company at “one of those brilliant tipping point moments,” when the idea of a CD-ROM to accompany all PCs was being rolled out. Scott’s role was to determine how the company could migrate an existing product, traditionally delivered on a floppy disk, to a CD. Then he had the fun challenge of deciding what to do with the extra storage space this created.
Following his departure from Pergamon, Scott used the experience he’d gained there and, together with his partner, graphic designer and creative director Malcolm Garrett, started “one of the first” digital studios in London, called AMX. Here, they began to merge traditional graphic design with digitisation.
“This was just pre the web kicking off, when people bought singles … so we’d be working with bands and along with their single, we’d say to the shoppers, ‘if you stick this [CD] into your computer, then you also get the video, and you could sing along to it’.”
Over time, album sleeves got smaller, and CDs eventually gave way to online content. By this time, however, Scott was well-positioned to step into the Internet era, and his appetite for seeing what was ahead of the curve was whetted.
“We didn’t learn this through regular academic routes, because the technology moved so quickly that by the time they codified it into a course, it’s out of date. And so you learn on the job by actually doing stuff.”
This hands-on experience allowed Scott to notice the ebb and flow of how technology was advancing and when the novelty was petering out and making way for the next wave.
Where in the cycle of AI are we?
Drawing parallels between historical technological cycles and the current AI boom, Scott reflects on how newly-introduced tools become transformative only after reaching a certain “tipping point”.
At the start of the curve, he says that the technology is a curiosity, and this builds excitement. After this, at the top of the curve, it becomes “ubiquitous and brilliant”. Then, towards the end of the curve, “it becomes commodified and tails off,” after which point, Scott says, the hunt for the next big thing starts again. This cycle lasts, on average, around five to seven years.
Looking at AI, particularly large language models and generative tools, Scott believes that it is too soon to tell what shape its curve is going to take. However, what he has found particularly striking about AI is just how rapid and widespread its adoption has been, from investors to mainstream users and everyone in between.
AI and the creative process
Talking about how AI both challenges and augments humans’ inventive capacities, Scott speaks about how, rather than AI killing creativity, it instead is part of a recurring “cycle of fear and adoption” which has been part of every technological leap – from photography’s impact on painting, to the proliferation of digital audio workstations enabling amateur musicians to compose music. It is this point where Scott believes that amateurs appreciate the skills required to do a job, and professionals raise their standards as a response to the democratisation – the more people who gain access to creative tools, the baseline skill increases, and a new appreciation grows for expert craftmanship.
However, he does distinguish between the “big idea” – that spark of inspiration that is uniquely human – and the “craft” or the skilful execution of creative projects. Talking about AI, he argues that it may commodify certain aspects of craft, but the heart of genuinely resonant, meaningful ideas still appears to be tied to human consciousness and lived experience.
“There’s an immediate excitement with typing something in and getting a short clip back, and that might be exciting from a viral perspective … but in terms of proper storytelling, the proper craft of it – that is something only a trained human can capture.
“If you have never actually shot or edited film before, and so you don’t understand shot composition, storytelling, colour grading, etc., you will end up with something that isn’t really styled the way you wanted it.”
When it comes to AI and music generation, Scott draws on the industry’s overuse of sampling, where repeated iteration on past works begins to diminish originality.
“In the music scene that we currently have, very few things are actually original, because a lot of stuff is sampled from other folks who sample came from other folks, and once you go back about four generations of having that … then the number of unique artists in the music world is incredibly reduced …”
He suggests that because AI has drawn on such vast material, whether ethical or not, the machine has nowhere else to go but to draw on itself.
“I don’t think it takes away the skill of a human. The reason these machines are so good at completing things is that they’ve looked at a base of content generated by humans, and based upon that, they are coming back with some stuff … But what’s the next generation of AI learning coming from? It’s learning from itself.”
However, with that also comes some serious sustainability considerations – especially when so much is going into producing work that could be considered “sloppy”.
“When we look at AI, how much of a percentage of the world’s power output are we going to be using to train the next generation of things that are potentially only 5-10% better than the last one? So diminishing returns on the training, but also the material we’re training it with, how much of that is previously AI-generated, and therefore, kind of sloppy.
AI is a tool, and it can be used for good
“LLMs are very good completion engines,” says Scott. On the one hand, it has scraped “all of human knowledge that it has access to in a digital format,” but “if we park that thought for a second [and look at] text-based feedback rather than content generation.” The difference between the two is crucial because content generation will, as the name suggests, generate content. Text-based AI, however, can be a very useful tool. By way of example, Scott talks about Google Gemini’s NotebookLM. He explains that the system is fed questions, and from those questions it can come back with a podcast scenario where two hosts discuss what you’ve asked it.
Considering the fact that attention spans are now in decline, especially among teenagers, Scott believes that this kind of AI tool is a clever way of allowing people to absorb “a huge amount of information” in an incredibly short amount of time.
The dangers of relying on AI
As much as we have seen that AI can be a fun and useful tool, there is also a danger that there can be an overreliance on it. From a human perspective, concerns have been flagged around diminishing self-trust. From both a personal and business viewpoint, there is a growing temptation to get as much content out as fast as possible; quantity over quality. Indeed, many people have lost jobs to AI, whether or not this has been to the success or detriment of the business.
Scott stresses the need for businesses to approach AI adoption ethically and reflectively, warning that rapid workforce reductions in recent times could have unanticipated negative effects and that long-term value may be lost if human nuance and oversight are discarded in favour of automated processes.
Advice for creativity and looking ahead
When talking about how to approach AI, Scott encourages creators to “play, break, and experiment” with it, and to treat it as an assistant or creative partner rather than a threat. “The last thing we should do is ignore it and hope it goes away,” he says. “It’s not going away. What is happening with machines and compute in two years will look quite different from what we’re doing with it now.”
The goal is not to resist or ignore change, but to engage with it hands-on, and to discover the uniquely human contributions that remain essential to creativity – whether that’s in storytelling, visual composition, editing, or anything else that requires the uniquely human traits of empathy, understanding, and foresight.
Scott also urges the need for vigilance around the ethics and sustainability of AI development, noting that there are substantial energy costs that come along with large-scale model training. He also says that the tool is currently limited to a pocket of humanity, and there is a need for equitable, global access to creative tools.
When it comes to technological evolution and revolution, Scott believes that it needs to be approached from a “humanist perspective”, believing that people – not just technological capability – are at the centre of creative and business innovation.
Scott believes that, over time, one of AI’s greatest contributions may turn out to be outside its core function, which is potentially accelerating humanity’s efforts in sustainable energy due to its massive power requirements. He stresses that while the pace of change is dizzying and the landscape unpredictable, the most resilient and rewarding path for creators and businesses alike is to remain adaptable, ethical, and fundamentally human-centred as they navigate the new era of AI-powered creativity.
