We are currently in an AI age, and that has recently began to bleed into creative fields like design. Today, there is a growing fear that AI might completely replace us, our jobs and our passion for design. This has lead to a distain for AI in creative circles, with some refusing to use it out of fear or protest. Despite this, AI can still be a great too to augment the design process without replacing the core of what it means to be a designer. In fact, AI has fundamental weakness that lead to poor designs when people are taken out of the equation completely. Let’s take a look at some of the ways you can use AI to enhance your design creativity.
What is AI good and bad at?
The key to understanding how AI can augment design creativity is to understand what AI is actually good at. With current training methods, AI has inherit strengths and weaknesses. With the right knowledge, you can utilise those strengths without falling victim to the weaknesses.
Strengths
- Pattern recognition. AI’s large amount of training data with their next-word prediction makes them excellent at surfacing information.
- Speed. Due to how AI models are trained, they can perform tasks extremely quickly, especially compared to manual human work.
- Automation. Due to the above two points, AI models are great at completing simple, laborious tasks quickly.
- Early Ideation. AI can use its enourmous amount of training data to pull together many different ideas. These can span a range of topics and concepts, giving you a great jumping off point.
Weaknesses
- Facts. While AI isn’t quite “autocomplete on steroids”, a large component of its generation is prediction. This means AI cannot ‘check’ if something is correct. Instead, it relies on facts being repeated enough times in enough contexts to ‘train’ the AI that it is correct.
- New Ideas. Since AI draws from a training pool to provide answers, there is no way for it to create something ‘new’. Anything it ‘creates’ will have to exist in one form or another for it to draw from. While it might provide you things you didn’t think of, those ideas had to come from somebody.
- Deep Thinking. A core limitation of AI is deep thinking. In fact, Apple found that reasoning models don’t do what they say on the tin, exposing a fundamental limitation with current AI models and their ability to actually ‘think’.
How to use AI to Enhance your Design Creativity
1. Generate ideas to spark your mind, not replace it
Sometimes you would like a bit of help thinking up different ideas. AI-assisted brainstorming can be a great way to reignite your creative mind by generating concepts, themes or ideas. These can be great jumping off point, letting you think about what’s good and what’s bad about each to hone in something that works.
But it’s easy to get into the trap of letting AI be your idea generation. While it might seem like it at first, AI-generated ideas can be boring, basic or just plain wrong. AI creativity, especially in design, relies on things that came before it, and struggles to fit your needs within its trained context. Instead of solely relying on what it generates, analyse what it makes so you can take the interesting parts to explore, adapt and refine into a high-quality work.
2. Placeholder images are fine, but replace them for a final version
Almost all visual designers use placeholders of some sort. Often, these are paragraphs of Lorem Ipsum, a nice photo of a mountain or a service like Lorem Picsum. These can be great time-savers, but you lose out on the relevancy of the content to your design. Instead, AI can generate a wide array of placeholder content for you to use while you design, giving you access to good-enough relevant content extremely quickly.
The most important thing to remember when use AI for placeholders is just that: use them for placeholders and not final content. Whether it’s AI-generated images, text, or anything else, you will almost always be able to find better and higher quality and more applicable assets if real human intuition and thought went into it.
3. Let it find facts, but don’t believe them until you’ve checked
Even though design is generally thought of as a visually creative field, there is a lot of theory, context and concepts throughout the process. Researching and understanding these can be especially difficult in niche market segments or with very specific user needs. AI can help by acting as a research partner, helping surface information you would otherwise struggle to find.
Although, it’s key to remember that AI struggles with facts - it has no way of knowing whether something is true or not, and will often insert its own opinions. When using AI as a research partner, you must never rely on the interpretations it generates. Instead, you can best use AI as a research finder. Leave the comprehension and analysis of the research up to you, or an expert in the field.
4. Creating conceptual images is great, final works is not
Throughout very early exploration phases of design, getting threads of ideas into a communicable form can take a lot of time. In these low-stakes environments, you could choose to use AI to create conceptual images that are ‘close enough’ to your idea. This can drastically reduce the amount of time to go from idea to presentation.
Before choosing to use AI to generate mock images, consider the purpose of the image. If the aim to provoke thought, be a visual support to an idea or indicative of what something might look like, AI generation could work. But if you’re looking to create a higher-quality mock-up or visualisations adhering to a brief, you should instead create these yourself. AI struggles with creating understanding specifics of what we want, especially when concepts are outside its training data. Any crucial image, illustration or graphic should always be created with at least substantial human input.
AI and Design Creativity: The Common Thread
It can be tempting to see the output from AI models as ‘free work’ and take it as-is, but doing this isn’t a good idea. AI might be able to generate something flashy and impressive, but hidden underneath is a disaster waiting to happen. AI is not a replacement for your creativity, but an augmentation. It won’t remove you from the process, but refocus your efforts onto the things that you do best: analysis, critical thinking, and precise design work. At the end of the day, you are designing for real users with real human needs - something AI is yet to master.
A Warning for a Generated Future
One of the fundamental weaknesses of AI is its inability to generate genuinely new ideas. What this means is AI relies on human creations for it to create anything of its own. If AI models are instead trained on generations from other AI models, it can lead to “model collapse”, with output qualities spiralling. AI can only be as good as its training data, and as it turns out, AI generation is horrible for that.
“If we allow generative AI to destroy human creative industries, by displacing jobs and discouraging aspiring artists, we are heading towards a future where art and music styles are fixed and static, and we are doomed to listen and see the same styles forever.” - Ben Zhao
So in reality, while it might seem that AI is crushing the creative industries, it needs us to survive. Unless we want a future like Ben describes, our use must be AI-enhanced creativity, not AI-replaced work.