We take a look at the possible implications of replacing a human designer with AI.
Out of all the AI questions around design, 'Can AI assist in the idea-generation phase of designing an app?' strikes me as one of the hardest to consider. Helping, guiding or, in some cases, removing the ideation stage from a designer feels like the most significant sacrifice.
I see the value in using AI to help with generating accessible colour palettes, filling in placeholder text and implementing alignment consistency. In these cases, a designer would have reached similar outputs to the AI anyway, just slower. Assisting with ideation, though, feels different.
How far can it go? When does the design only belong to the designer because they realised an AI idea? Did they genuinely design it? What can this process mean for designers in the long term? Can there be a world where AI is effectively and responsibly used to support the ideation process rather than veer into taking it over?
Many of these questions depend on various other parameters, considerations and author perspectives. So for this blog, I’m choosing to focus on ‘simply’ AI aiding the ideation process.
Sometimes, ideation is challenging. Most of the time, a designer just has to start and trust that ideas will come. Iteration of ideas will lead to a new idea, and eventually, a designer will land on their chosen design - in whatever capacity that might be. This part of the ideation process is interesting because ideas generally come from context, past experience, personal preferences (which can be hard to ignore at times), intuition, and often parameters of a brief that, in some cases, are hard to define. Can AI replicate this?
In this context, we’re thinking only about this initial ideation phase and not considering ideation in other areas such as content generation, automated prototyping - this helps exist initial design ideas, image and style recognition for existing prototypes, collaborative tools for idea sharing.
There are 3 main areas and points in a designer's journey of ideation where AI can assist in forming ideas:
Generative design - tools that generate designs based on parameters and constraints set by the designer. This can help explore a wide range of possibilities and spark creative ideas.
Pattern recognition - AI can analyse existing design patterns and trends to identify what works well in similar apps. This can provide valuable insights and inspiration for designers.
User behaviour analysis - understand how users interact with similar apps. This information can be used to make informed design decisions and create user-friendly interfaces.
We’ll look at each of these in more detail throughout this blog.
Generative design tools leverage algorithms and artificial intelligence to automatically create and iterate on design solutions. They can help designers in the very early stages of design processes. They work by generating solutions based on a set of parameters and constraints the designer provides. These constraints include:
Next, algorithmic exploration calls upon a vast design space based on these parameters. The algorithms can generate numerous design variations, often more than a human designer could quickly produce manually.
This is about providing designers with inspirational designs, finished interfaces and complete solutions based on their input requirements. Is this really ideation? It is more comfortable to conclude this as refined inspiration, which poses the question - is it any different from scouring across Dribbl?
Ultimately, the design or solution won't necessarily provide the designer with enough UX consideration to take a complete interface design and run with it. However, if resources allow, it could be a helpful stage in the ideation process to replace or add to a designer's set of tools to seek inspiration.
AI can analyse vast amounts of design data, including trends, styles, and successful patterns in similar projects. By recognising patterns, AI can provide designers with insights and inspiration, helping them stay informed about current design practices and user preferences.
A designer could feed this AI tool with a previously generated design and focus on its ability to analyse successful UI and UX patterns and emerging new trends. At this point, has the designer been assisted or overtaken?
In another case, the designer could use their ideas against a pattern recognition AI tool and see what comes from it. However, that veers slightly more into the refinement of a design or developing a pre-existing idea. Let's remember, too, that the interpretation and application of these patterns still rely on human designers' creative judgement and expertise.
Research into User behaviour analysis tools suggests perhaps unsurprisingly that, essentially, it aids a designer in analysing available data and making informed recommendations based on their parameters and constraints, as we have mentioned.
AI can also compare user behaviour across different apps in the same category. This competitive benchmarking helps designers identify industry standards, innovative features, and areas where their app can differentiate itself during the ideation phase.
Finally, it can identify common pain points across certain features or determine which features are used most and least. This latter point is interesting because it can give prominence and structure to the approach of a design. If a designer knows that the most attention should be given to a particular feature, sometimes a design can surround it, sparking inspiration and originality with a more focused perspective.
Should we not do that anyway? Assuming User Research has taken place and there is analysis to draw from, AI, in this case, can’t always help to move a designer beyond this point of analysis and understanding of how the design can fulfil the requirements.
AI can assist a designer in many areas of the design process. With careful consideration, these tools can be adopted to help push the boundaries of the discipline and help produce engaging user-centred design.
In the case of ideation, though, the jury is out. AI can help a designer arrive at the ideation process with perhaps more refinement or where they might have done alone just a lot faster. The generative design would spark the most inspiration, a supercharged dribble. Designers are always looking for inspiration, especially if they are pressed for time.
For me, the only real benefit of using AI at the ideation stage is how quickly inspiration is provided with closer matches to a designer's requirement. In time, frequent use of such tools would likely be helpful, but at first use, there would be a high set-up cost including:
Finding the right tools
Learning how to feed them the correct information
Checking through with human contextual understanding on how valuable the generations actually are
Purchase cost. The most recommended generative design tool starts at £1,700 for an extension. So it’s an investment.
The role of the human designer remains critical in interpreting and refining the generated designs. Consider contextual factors and make subjective judgments beyond the metrics AI has interpreted. Does this add more stages to a designer's process and take back the time AI has potentially saved and does this ask a designer to work backwards, unpicking a suggested design to check its utility?
Unless a designer has sufficient time, budget, and capacity to fine-tune AI tools to arrive at enhanced inspiration boards, sticking to traditional approaches might still prevail.
Search over 400 blog posts from our team
Subscribe to our monthly digest of blogs to stay in the loop and come with us on our journey to make things better!