Artificial intelligence is slowly replacing non-creative jobs, but can AI be creative enough to replace logo designers?
In this article, we’re going to be comparing two different workflows: working with a traditional designer VS an automated, AI-powered interface.
Let’s look at each stage of the logo design process individually. Logo design can generally be defined in these 6 steps:
- The creative brief
Let’s dive into each one.
1. The Creative Brief
Who is better at accepting and defining a creative brief? We usually define the creative brief as a set of guidelines to be used when designing the logo. Typically a creative brief will include:
- Inspiration (general aesthetics)
- Values that the logo should convey (ie. trust)
- Display requirements (ie. small sizes / black and white)
For example, the requirements might be a) use the name “Acme Co”, b) invoke a feeling of trust, and c) look clean and modern.
Designers generally form a creative brief by asking the client a series of questions and taking notes. A designer is currently better at accepting specific instructions such as ‘I want the T in the logo to look like a paddle’ – AI-powered logo makers can’t accept that kind of input (yet).
Using a DIY interface powered by AI, the user is able to choose from a selection of logos as their inspiration. We can manually teach the AI system which logos have which values (ie. a blue logo might convey trust) providing a more streamlined approach, which is better suited for simple logos.
A key step in logo design is doing research on the industry. Looking at competitors logos and brand values can help guide you to create a unique brand, but one that customers can still feel comfortable with.
Traditional designers will often spend an hour or so looking at 10-15 competitors’ logos and that’s generally as far as they’ll go. They can sometimes spot trends and make actionable insights, but the human brain is not the best tool to make these insights.
With AI, we can crawl a database of thousands of competitors logos, using those as guides for our designs.
AI is able to better decipher and identify trends within the client-provided inspiration and the discovered competitors logos. It’s able to do this by analyzing hundreds of traits for thousands of logos in the industry, whereas a designer often only looks at a few.
Otherwise known as the first set of mockups – this one is a doozy. This is the stage when the designer actually creates the designs based on the creative brief.
Let’s take a look at how both traditional designers design the first set of mockups vs. AI powered systems.
A traditional designer will generally take a week or so to sketch and illustrate 4-5 designs. This is incredibly time consuming but a good designer can create great designs.
Logos are generally just a combination of fonts, colors and shapes – all of which software can easily create and modify. The big challenge for AI is learning to be creative.
Let’s define creativity in the context of logo design: the ability to combine fonts, shapes, and colors in a novel way to satisfy a specific set of guidelines. The more creative one is – the better they are at designing a solution that satisfies a given set of guidelines (defined in a creative brief).
Rarely will a solution 100% satisfy every guideline (especially because different people interpret designs differently), but it’s up to the creativity of the designer to try and get as close as possible.
Designing a logo can be broken down into two parts: assessing the creative brief (ie. convey trust) and then picking a combination of ‘ingredients’ that make the logo (ie. Helvetica Bold, dark blue, text-based layout, etc.)
At Logojoy, we have the idea of ‘ingredients’. A logo is 100% defined by it’s set of ingredients. That’s all a logo is in our system – a set of ingredients. Fonts, colors, layouts, text effects.. these are all ingredients. Our AI is constantly learning which ingredients work well together and which represent certain values (ie. trust).
So, when a client asks our AI interface to ‘design’ a logo, our algorithm is really just picking an assortment of ingredients that it thinks will work well together.
AI can actually be exceptionally good at this. Over time, it can spot trends in users favorites and use those insights to create better initial mockups.
This is a serious bonus for a DIY interface – the ability to see all mockups and revisions instantly in real-life scenarios. Designers will rarely present designs and minor revisions in real-life mockups. With a DIY interface, you can always see your design in multiple scenes.
Generally the client will pick one or two favourites to see revisions of.
Human designers are limited by time here. It takes minutes if not hours to generate and send over even the most basic requests (ie. can we try a different font). However, designers are currently capable of handling more complex requests such as ‘can we replace the T with a paddle).
DIY interfaces have the luxury of instantly generated logos. So seeing multiple revisions is where it really shines. The client is able to ask for and instantly get a batch of revisions. Different fonts, colors, layouts, symbols, etc. are an ease to browse.
AI is also able to recommend certain combinations in real time based on certain updates. For example, if the client changes their font, AI can suggest a new layout based on the new font.
While good designers are very capable of adding that extra layer of polish, it is a time consuming and expensive process that can feel tedious for both parties.
A DIY interface allows the client to directly polish their logo by changing everything from font size to colors.
It’s going to be a tough battle, but I would predict that AI replaces the lower end of designers offering customers simple logos. Beginner to intermediate designers likely won’t be able to compete with AI logo designers, due to the sheer learning capability of AI. Great designers, however, will still reign supreme for at least a few years, as they can provide a more novel solution than AI.