This article is only available in Dutch 🇳🇱
Recently our designers met for a Design Lab: an inspirational session with a different organizer and theme every month. Within these sessions there's room to freestyle on self-devised assignments, collaborate on challenges within projects or discuss new Figma features. Trying out new tooling is also an option. Because of the huge rise of (and interest in) Artificial Intelligence within the design field, Creative Strategist Matthijs Klaver and Team Lead Design Thessa Hulsman took out a Midjourney subscription for our design team. Together they organized a Design Lab completely dedicated to this tool.
And then there was ... Midjourney
In July 2022, AI company Midjourney launched their eponymous design tool. The tool is a typical example of a foundation model, a deep learning algorithm trained on massive datasets taken from the public Internet. Based on linguistic descriptions or prompts, the model generates unique images of super good quality. Midjourney isn't the only program that can do this, but within our field it's the most interesting without a doubt. In fact, the tool is even better trained on digital art than programs like DALL-E (from the makers of ChatGPT) or Stable Diffusion. As far as we're concerned, Midjourney delivers the gold standard when it comes to text-to-image generation. And we're not the only ones with that vision. People worldwide are deploying the tool to generate unique illustrations, app icons and stock photos. This is just a selection of the possibilities, as the similarities with and applications to design work are endless.
What can AI bring us as app designers?
Thessa: 'The goal of this Midjourney Design Lab was to experience and find out what AI can bring app designers in this moment. That's why we first paid attention to the technology behind the program and writing good prompts. For those who wanted to add even more AI to the assignment, we also included Adobe's AI tool Firefly and ChatGPT in our introduction. The first for visual editing and the latter for writing good prompts. After the introduction, the designers got to work on a free assignment individually. During the inventory at the end of the session, all kinds of things came up. From app icons, app designs, illustrations and game interfaces to stock photography and seamlessly executed fantasy crossovers, such as Harry Potter of the Caribbean and balloon sneakers.'
Pros and cons
Matthijs: 'The first advantage of AI integration in the design process is the time saving. Using Midjourney during brainstorms, for example, lets you come up with new, fresh and often surprising ideas super fast. In the design process itself, the tool saves time on aspects such as 3D illustration, which we use to richly decorate apps. Another advantage is that the program generates unique, complete images by accessing a super-sized image bank. It effortlessly combines all the images you can think of, in any style you want. As a designer, it enables you to put your creativity into top gear in no time.Â
A downside of Midjourney? The tool doesn't think like a UX / UI Designer, but as an overall image maker. As a result, the output is really only suitable as visual inspiration. It gives you a starting point. And to get to that starting point, you need to be able to write good prompts. You also need to know the limitations and the capabilities of the tool. In short: personal, human frames of reference and experiences remain indispensable when it comes to quality designs. But we will definitely continue using this tool!'