A few months ago I joined yet another chat debating the implications of generative AI technologies entering the workplace.
A former colleague, and seasoned creative director, mentioned after seeing some of the visual outputs of Google’s new Veo 3 and OpenAI’s Sora, that he finds gen AI tech “simultaneously exhilarating and terrifying”, and that he’s “humbled by the force of what’s coming”, going on to ask if we’re heading towards a creative renaissance, or the final erosion of consensual reality.
This got me thinking about the growing polarization of views surrounding AI tools, particularly in the creative field. Depending on who you talk to and where they’re at in their career, it feels like equal parts excitement and optimism, but also fear and anxiety.
Like many creative professionals, I’ve experimented a lot with emerging AI tools to see what’s possible, and to also understand the current limitations.
At work, we don’t have a formal mandate to use AI. So I’ve decided, mainly out of curiosity, to take a proactive approach and learn how I can integrate AI into my workflow.
What I’ve found is that some of these tools are incredibly useful for both preliminary ideation and execution, while others have proven a time-consuming proposition in terms of managing the so-called “hallucinations”. For instance:
- image-based tools: creating people with extra fingers, distorted anatomy or unnatural facial expressions; inconsistent elements (e.g. a horse with cat ears and a fish tail), rendering objects or environments that are physically impossible.
- text-based tools: citing scientific papers or authors that do not exist, making up historical events or legal precedents, fabricating quotes (the Megalopolis trailer fiasco is a great example!), statistics, or describing nonexistent features in a real product.
Will the future of creative expression be relegated to text prompts and how many tokens you’re willing to spend to develop an idea?
Taste and judgement, it seems, are more important than ever as AI tools get smarter and better —but also error-prone, automating higher-level tasks that were previously done by hand. The devil is in the details, of course, but machines (LLMs) lack empathy and a genuine worldview, so outputs will need a human sensibility —the physical, lived experiences we all take for granted— to weed out the incomprehensible slop.
Oversight by experienced people will be crucial, if the goal is to create believable outputs. Art directors and designers will need to strive for visual outputs that are invisible and go unnoticed, in the same way the best cinematic special effects composites blend seamlessly into a story’s visual narrative without raising so much as an eyebrow among audiences.
Text outputs too, need to be rigorously scrutinized and fact-checked for accuracy —again, the trailer fiasco involving AI-fabricated pull quotes used to promote Francis Ford Coppola’s Megalopolis is a superb cautionary tale of failing to execute due diligence.
This brings me back to the polarizing question of AI ushering in a creative renaissance, or contributing to the erosion of consensual reality. I will say, without hesitation, the former of the two: renaissance 100%. The barriers to expression have diminished while the possibilities have clearly broadened, and will continue to do so, empowering storytellers and creators.
But at the same time, many people are uncomfortable with change and may tightly hold on to the past. Others (myself included) see opportunities and have chosen to embrace these new tools as a way to amplify and enhance creative output. I believe many of the current AI fears are misguided. They’re “tools” to be leveraged by humans. Nothing more.
We’re really just seeing the beginning phase of the erosion of traditional workflows and methodologies as these new tools gain momentum. Designer Alexander Kohlhofer believes we’ll see more roles shift from technical execution to strategic judgement:
When everyone can generate content, code, or designs, the real value lies in:
Knowing what to create: Understanding what’s worth making in the first place
Making meaningful choices: Selecting the right approach from countless possibilities
Evaluating quality: Distinguishing between good and great outputs
Understanding context: Applying the right solution to the right problemThe most valuable professionals will be those who can:
- Ask the right questions
- Frame problems effectively
- Make sound decisions
- Provide meaningful direction to AI tools
Yes but, when everyone can, for example, create code, with apparently 92% of developers now using AI coding tools we see a lucrative and growing market for vibe coding cleanup services, because as Donado Labs put it in a recent post: “The harsh reality nobody wants to admit: most AI-generated code is production-unready, and companies are desperately hiring specialists to fix it before their technical debt spirals out of control.”