AI: tool or tyrant?
How AI is changing creative practice
It’s difficult to have a conversation about creative practice these days without AI cropping up. Running a small business, I’ve found various AI tools – from ChatGPT to Adobe Firefly – enormously helpful, whether that’s for project planning, previously arduous Photoshopping or simply acting as a soundboard when I’m working through the rationale behind a creative route.
Sometimes, I experiment with AI as part of the creative process itself – so far this has taken the form of generating imagery where a project wouldn’t otherwise have the budget for bespoke photography or illustration. Elsewhere, I use it more as a production tool – for example, removing unwanted elements from existing photography to help visualise design concepts in real-world environments. Both still need careful art direction, curation and editing.
For me, the question is perhaps less whether we as creatives use AI, but more how we use it, and which parts of our process we’re prepared to outsource to it.
So I was interested in chatting with other creatives about where they stand on AI, how they’ve integrated it into their working lives, and where they draw the line.
Thank you to Ed Cornish, Alexander Adegbite, Syd Hausmann and Ricardo Eversley for sharing their perspectives.
Ed Cornish
Ed is Design Director at Greenspace, a creative company that creates legacy ideas for forward-thinking brands. Recently he’s been leading the design of several projects in the luxury property and architecture sectors.
Previously he worked for himself under the name Studio Brick, and has freelanced at companies including Magpie Studio, Johnson Banks and Sunshine. He’s also worked full-time at Burgess Studio and Mother Design, taught graphic design, and been a mentor for undergraduate and postgraduate students.
He studied at Central Saint Martins and the Royal College of Art.
More on Ed: Greenspace website; Greenspace Instagram; LinkedIn
From a purely creative standpoint I think AI creates a lot that is ‘good enough’ but very rarely great.
Q1: What excites you about the use of AI in creative practice?
Ed: I go through phases of being excited and pessimistic about AI, and I think currently I’m at a bit of a plateau with it. Maybe it’s too much time on LinkedIn, but I’m very fatigued with all the hyperbolic hot-takes from the Huel-chugging optimisation bros.
Having said that…AI has definitely opened up some exciting possibilities that previously were beyond me, such as vibe-coding to make things in minutes that would have been impossible for me to do before. It’s very useful when I get stuck on something technical, and can get a solution quickly, rather than trawling YouTube tutorials.
I’ve not tried using voice commands with AI, but I can see that being something very exciting. Being able to brief in or instruct certain tasks while you’re on the move would be genuinely useful…although I am usually desk-bound so maybe more exciting for someone else!
My current role does not involve much admin, but if all the AI-enhanced accountancy and admin tools were available during my past-life as a freelancer, I’m sure I would have been an evangelist for those too.
The best AI-usage is when the AI is invisible.
Q2: What scares you about the use of AI in creative practice?
Ed: I have all the common concerns with the usual big issues surrounding AI. But from a purely creative standpoint I think AI creates a lot that is ‘good enough’ but very rarely great – so I do worry there will be a lowering of the creative bar and a widespread homogenisation if AI becomes over-adopted. I definitely don’t think everything created with AI is ‘slop’ – there’s good AI-usage and bad AI-usage, but the best AI-usage is when the AI is invisible.
I have seen some frankly terrifying videos of people being interviewed for jobs by AI avatars. I worry that within the creative industries, a process like that could prevent people with less conventional backgrounds from entering and bringing new ways of thinking with them. AI will always contain whatever biases are present in its training data, and outsourcing creative judgement to a majority opinion feels risky.
We have had a few instances over the last 18 months that show how AI is changing things client-side. We’ve had our own work edited with AI and sent back to us as reference, or clients brain-storming potential logos (in meetings) with it and asking our opinions. While part of me finds this scary, I think it’s going to increasingly become the new-normal. Collaborative clients are going to use it as a tool and we can’t hide from it. It’s down to us to be the voice of reason and prove our worth as creative and strategic thinkers, not just pixel-pushers.
Q3: How do you currently integrate AI into your creative practice?
Ed: In several ways:
Vibe Coding in Chat GPT: We’ve used this to create pattern generating design tools. Any pattern-based identity can be laborious to test and develop, so this has allowed us to iterate in compressed timelines. I have also used it to write AfterEffects expressions which has opened up a lot of motion possibilities that were beyond me previously.
Website audits with ChatGPT: I also use ChatGPT to do website audits for digital jobs, which is a big time-saver. It’s great for identifying ways a site could convert better, but I think one needs to take its ‘advice’ with a pinch of salt – it’s pushing an agenda that would make all sites look exactly the same.
Mockup Generation with Runway/Gemini: Great to show clients really specific visuals, especially if location is important or an existing mockup doesn’t exist.
Footage generation with Runway: We use Runway to turn images into moving image footage – which we’ve used for a property developer client who could no longer access homes they had previously sold to shoot new footage, giving their old imagery a new lease of life.
Q4: What would you never use AI for?
Ed: I don’t think I would currently ask AI for an opinion on creative work, it’s only ever going to echo the opinions of the datasets it’s trained on. It’s also obliged to give you an answer and make suggestions. It’s never going to say ‘it’s fine’ or ‘this is great, don’t change a thing’, which you would hope a human would say once in a while.
I am also a total Luddite when it comes to communication, so I can’t see myself using it to write my emails or feedback for me. I really cannot stand ChatGPT’s tone of voice, and would not want that to come across in my own words!
There are some tasks I’ve tried to outsource to AI, with very poor results. Things like ‘extrapolate these few letters into a full A–Z typeface’, a task that you would expect to be relatively straight-forward, have come back totally unusable. It struggles to create anything new, it just regurgitates and splices what’s already out there, so I wouldn’t rely on it for anything that you would want to feel original – I just don’t think it’s capable of that kind of thing, yet.
Maybe starting an identity project by asking ChatGPT to create you some logos with the client’s name is a good way to start a project, and see exactly what not to do?
Q5: Your verdict – AI: tool or tyrant?
Ed: Tool, for now…
Alexander Adegbite
Alexander is a freelance illustrator. He has worked in advertising and animation, with experience in art direction, concept design and background design. Born and raised in Croydon, he studied Digital Art at The BRIT School from 2019-2021 and graduated from The University of Hertfordshire with a degree in Comics and Concept Art in September 2025.
Q1: What excites you about the use of AI in creative practice?
Alexander: In all honesty I can’t say I’m excited about the use of modern generative AI in any creative practices. When the text to image technology first came about, those initial abstract pictures had a surreal quality to them that I quite liked. It was intriguing. I would say I was most excited about it then, when it felt like an interesting, experimental tool.
Q2: What scares you about the use of AI in creative practice?
Alexander: Mainly, the loss of opportunities. A lot of the mundane stuff that could get you paid is being done by AI now, which, sure, efficiency. But even the cool stuff is going too and that was already quite difficult to get gainful work in.
Q3: How do you currently integrate AI into your creative practice?
Alexander: Right now I avoid using Generative AI in my art. Although I can’t deny its usefulness in things like generative fill in Photoshop. I have experimented with what it could do at various stages of its life – I’m sure it could be a good tool if paired with a strong foundation in design and draughtsmanship.
The ‘generation’ of ideas… I think we need to keep that part human, even in the most commercial of products.
Q4: What would you never use AI for?
Alexander: The ‘generation’ of ideas. I think the conceptual part of art is the hardest part. All the problem solving that comes with iteration and working out the ‘why’ of a design. I think we need to keep that part human, even in the most commercial of products.
It’s become an encompassing buzzword and controversial topic, and I believe we cannot let that take away from the good other AI tools can bring to production, accessibility and research.
Q5: Your verdict – AI: tool or tyrant?
Alexander: Overall I think GenAI has brought more negatives than positives. Even ignoring the environmental issues it comes with, it's built off the backs of artists who did not consent to having their work taken, who are now being replaced by it. Taking the commercialism out, I still think GenAI has brought more harm than good. It has made people think less about their artistic decisions, and it cannot give you anything that hasn’t been done before.
However that is not to say there isn’t a place for AI. Artificial Intelligence has been around for a long time and I think it's important to make a distinction between this modern generative AI, and other AI tools being developed. It’s become an encompassing buzzword and controversial topic, and I believe we cannot let that take away from the good other AI tools can bring to production, accessibility and research.
Syd Hausmann
Syd is a Hove-based graphic designer, illustrator and artworker. Over the past 20 years she has worked as an in-house designer, agency creative and freelancer, collaborating with a wide range of clients from small local charities to large global brands.
Recent clients include Ambition Institute, Museum of Croydon, The National Archives, Greater London Authority and Design Council.
More on Syd: LinkedIn
Q1. What excites you about the use of AI in creative practice?
Syd: I think one way is how it can get ideas moving. I have a client – with an existing brand – who I’ve worked with for over a decade to the point that they now often request deliverables with no design direction. I used to spend quite a bit of time sketching out mock-ups but AI is a useful tool for more easily initiating ideas and exploring possibilities with them.
Additionally, I do a fair amount of PDF remediation – making PDFs accessible/WCAG-compliant – which can be incredibly time-consuming, and the software needed to do it properly is expensive. AI is starting to make some of that process easier (though it always gets some things wrong so it’s not all there yet!).
Photo enhancement and enlargement is another big one for me. What used to be time-consuming is now a breeze. I can also see huge benefits in the more administrative side of running a creative business, particularly accounting and all those jobs that need doing but aren’t necessarily the reason you became a designer in the first place.
As the tools become more accessible, more clients are going to want to have a go themselves, and that potentially lowers the barriers to entry – but not necessarily in a good way.
Q2. What scares you about the use of AI in creative practice?
Syd: There are several things. One is the sheer amount of dodgy content it can produce – and the fact that clients are increasingly using it, sometimes not very well! I recently had a client ask me to create a robust PowerPoint template for them because their team was using AI to generate presentations. The presentations actually looked quite nice – if you didn’t want them to be on-brand.
The other big concern for me is how often AI simply gets things wrong. In one of my recent jobs of making a PDF WCAG-compliant, I was getting a lot of errors on how the tables were scoped. In this instance, AI made things worse rather than better. You can’t just assume that because something looks right, it is right – particularly when accessibility is involved. There still needs to be someone who understands what they’re looking at and can spot the mistakes.
I’m also concerned about what it means for the creative industry more broadly. As the tools become more accessible, more clients are going to want to have a go themselves, and that potentially lowers the barriers to entry – but not necessarily in a good way.
And there’s a danger that creativity becomes more bland and more similar. If we’re all using the same tools, trained on broadly the same material, and asking them to produce broadly the same things, there’s a risk of stagnation. Part of the value of creative work is the slightly odd, unexpected or very human bit that you wouldn’t necessarily get from an algorithm.
Q3. How do you currently integrate AI into your creative practice?
Syd: It’s actually become quite embedded in what I do, although I tend to use it as a tool rather than handing over the whole creative process.
I use it to help with my accounts, for example. It’s created some much more robust spreadsheets for me, with integrated formulas that I probably should have been using years ago.
In Photoshop I use Firefly and Flux.2 Pro all the time for image upscaling, enhancing and enlarging. One particularly useful thing is being able to enlarge an image to create additional space around it – for example, creating an area where I can then put text. That can be incredibly useful when you’ve been given a photo that simply doesn’t have the right dimensions or composition.
I also use Midjourney as a sort of launchpad for illustrative ideas. I’m not necessarily looking for a finished image; it’s more about seeing what directions something could go in and using that to kick-start my own thinking.
Q4. What would you never use AI for?
Syd: I wouldn’t use it to get from the starting point to the finishing point of a design. There needs to be some of my own creativity, ideas or inspiration in there. That, to me, is the joy of being a designer – the thinking, the problem solving and the creating. That’s the bit I don’t want to outsource.
It’s not that the process can’t change but I still want to be involved in the thinking, making and directing. The process might now be slightly different, but that doesn’t mean the designer needs to disappear from it.
It’s a tool – but one that can very quickly start dictating the pace and expectations around the work.
Q5. Your verdict – AI: tool or tyrant?
Syd: Both, really. It’s definitely part of my toolkit as it speeds up some of the processes and helps with others. But it’s also created more work and expectations. I’m now finding myself fixing clients’ AI-generated design attempts, tweaking poor-quality photos into something usable, or being expected to produce a finished result in a timeframe that wouldn’t have been realistic before.
It’s a tool – but one that can very quickly start dictating the pace and expectations around the work. The trick, for me, is making sure I’m using it to make my work better or easier, rather than simply using it because I can.
Ricardo Eversley
Ricardo is a London-based creative practitioner, researcher and Senior Lecturer in Digital Innovation and Course Leader for Graphic Design BA (Hons) at London Metropolitan University. His professional and academic practice spans graphic design, visual communication, digital innovation and higher education.
His research explores creative agency, design process and the evolving role of artificial intelligence within visual communication education, with wider interests including AI literacy, authorship and the relationship between traditional and computational approaches to making.
AI poses interesting interventions in the critical expansion of original research and ideas that may have previously remained quite narrow or within a given lens.
Q1. What excites you about the use of AI in creative practice?
Ricardo: My excitement of AI is currently situated around accessibility in higher education. When we look at the awarding gap as well as disparities around the student journey, there are some tools that can positively impact these groups of students if used in a critical and responsible manner. In terms of my own practice AI poses interesting interventions in the critical expansion of original research and ideas that may have previously remained quite narrow or within a given lens. It allows me to rapidly visualise things that might previously have remained abstract, test possibilities, interrogate the ideas and move between different forms of making. The latter development and output stages do however remain void of AI presence.
Q2. What scares you about the use of AI in creative practice?
Ricardo: Creative practice seems to be splintering into a plethora of directions around AI. This is also having a negative impact on design education. Whilst there has been and still is a major emphasis on critical thinking and human-intelligence idea generation and thinking, what we are seeing in some forums seems to be the opposite. The emphasis on speed and output over a deep understanding of the problem and solution is creating a generation that may struggle if the AI switch is turned off. The critical literacy around what makes visual communication a human first method feels like it’s losing ground to hyper perfection.
There are also real issues around biases and inaccuracies within LLMs (large language models). If this idea of productivity and speed first persists, students risk producing work without really understanding what they are trying to say and equally remove their own agency in the process.
The emphasis on speed and output over a deep understanding of the problem and solution is creating a generation that may struggle if the AI switch is turned off.
Q3. How do you currently integrate AI into your creative practice?
Ricardo: My current position is to use AI as much as possible. The core reason here is based on exploring how it can impact pedagogy. It’s imperative that I am able to hold a critical position when discussing AI with my students. So, in terms of development and literacy it’s increasingly present in my day-to-day activities.
I also use it to critically analyse my own output and pedagogical methods which I have found useful in part. A key method here however is to ensure that what I get back I am able to research further outside of LLMs so my prompting is very particular in that it always requires me to explore outside of the LLM and back into human outputs such as journals and academic reading. Using it to narrow results has also been quite useful.
In terms of research work I use AI at different points in the process rather than treating it as a single production tool. I’ve found it helpful in terms of testing structures for academic writing and initial brainstorming. I have also used it as an accessibility aid for my lesson material alongside traditional checkers. As previously mentioned, the prompting always includes steps that take me back to literature surrounding the topic I am exploring.
Q4. What would you never use AI for?
Ricardo: I wouldn’t use AI as an output tool either for my academic work or creative practice work. In the last year I have been extremely intentional in including new-analogue processes back into my research and work to negate my pursuit of AI literacy. The more you use AI the more you lose. This is something that feeds into the previous question about fear. It’s important in my opinion that being intentional in respect to human-intelligence activity sits alongside artificial intelligence use.
It’s a tool that has huge intersectional value if you can retain agency.
Q5. Your verdict – AI: tool or tyrant?
Ricardo: I would say it's a tool that has huge intersectional value if you can retain agency.
I think framing questions around where you used AI versus did you use AI keeps an important level of critical conversation going around this new tool.
Thank you to Ed, Alexander, Syd and Ricardo for chatting with me.
Comments