AI in Product Design: Revolutionising Creativity and Efficiency

With the ubiquitous nature of AI filtering into every stage of software development (and life in general), the topic of how it fits into the world of Product Design is becoming more and more prevalent. As with a lot of other disciplines, the whispered, fearful question being increasingly asked is: will all designers eventually be replaced by AI tools?

At Amdaris we are tech optimists, and maintain faith in a future where humans aren’t replaceable by technology. What is highly likely is a future where every designer is using AI tools in their workflow to enable them to focus their attention on the more creative aspects of their work, and the nuances of connecting everything they do into one, well-designed output.

There’s so much more innovation to come, but there’s already plenty of areas where AI is changing and improving product design. In this article, we’ve assembled the key areas where we believe product designers should look to incorporate AI into their lives and products.

PERSONALISATION

Algorithms for driving content are nothing new (just look at social media), but AI algorithms take the concept even further. By analysing user preferences and behaviours, they will be able to not only drive content, but tailor the UI to meet every user’s exact needs. The more the AI system learns about users by interacting with the product, the better it will become at predicting their preferences, and creating an enhanced, personalised user experience. I’ve been playing about recently with Dot (new.computer) and it does this on a conversational level, understands and remembers who you are and what you’ve mentioned in the past. In this case it changes the content of the conversation but if this same concept was applied to UI’s/features then products would be ever-evolving and organic, removing the need for over-simplifying to accommodate every user type.

EFFICIENCY/AUTOMATION

As with every industry and every role within them, AI is making product design processes more efficient. There’s plenty of ways that AI can be utilised to help automate routine tasks and streamline different aspects of product design, allowing designers to focus on the more creative aspects of design. This is something Figma have just confirmed with their latest announcements at ConFig, implementing new AI features (some already released, some coming soon) across every touch point of the design process in Figma. Some of the key areas where efficiency can be improved by AI include:

● Analysing Data

AI can analyse user data rapidly, uncovering insights to help understand the target audience better. In fact, user research platforms are already integrating this functionality, such as usertesting.com. They use AI to summarise key findings, identify friction, analyse user sentiment and more.

● Creating Personas

AI can help in developing dynamic and accurate user personas by analysing user interactions and behaviours, making the process faster (and at times more accurate/realistic) and allowing designers to shift their focus elsewhere.

● Copywriting

Copywriting was one of the earliest uses of AI in product design from apps such as Hemingway. Many designers may not be fluent in copywriting, so these tools can be hugely beneficial. Copy generation is one of the features Figma has added, while still at a fairly basic level, as they become more advanced, AI tools can also be utilised to check copy for microaggressions and adjust to be more inclusive (I’m currently using Claude to do this with prompts and it provides some interesting insights).

● Automating Design Workflows

AI can streamline different tasks at many stages in the design process and we’re seeing more and more Figma plugins pop up, including image and copy generation, accessibility improvement suggestions, front-end code generation, colour suggestions and much more. And now some of these plugins and more are being integrated into Figma itself, including UI generation but this is still at an early stage and from first look should only really be used for quick rough and ready visualisations.

DOCUMENTATION

While there are some designers who do a great job of documenting their work, it is often one of the more forgotten (but still essential) stages of the design process. Documenting how components work for developers and how they should be utilised by designers can massively help in streamlining the creation and use of design systems.

Advances in AI tools are unlocking new documentation methods for product designers. Librarian, for example, labels and catalogues Figma designs, and allows designers to interact with it using human language. For example, a designer might ask: “What is the best way to show a table on a dashboard?” and Librarian will return the components that fit exactly what they’ve asked for. Eventually, the documentation process itself will be fully automated, and AI tools will understand the behaviour of a component and be capable of writing full documentation – Figma seems to be leaning towards this a bit with their new feature that can generate prototypes based on static designs, it understands intended behaviour of navigation and links it up (albeit not perfect yet), now if it could also generate behaviour documentation as well this whole documentation process would be automated.

For now though, asking language models such as ChatGPT or Claude to expand on a quick description of a component’s behaviour is a suitable solution.

UI GENERATION

While AI is still a way off from creating UI designs to the standard of any human designer, it can be a great tool for rapidly mocking up UI examples during workshopping sessions. Programmes such as Uizard or Creatie can be fantastic for generating designs to help rapidly visualise concepts. Being able to quickly generate examples will enormously help to provide explanations and improve understanding of how something will work. And as mentioned earlier, this is something that is now natively supported inside of Figma (whether or not it can compete with these other tools is still up for debate).

INSTANT FEEDBACK AND SUGGESTIONS

While designers usually know what they should do to make something look or work better, there are times where creative blocks can get in the way of progress. It’s always better to talk with other designers, but AI tools can offer rapid feedback, minimising the need for smaller designer-to-designer discussions, freeing that time up for more creative discussions.

PREDICTIVE ANALYTICS

AI can’t see into the future, but it can certainly help us predict small parts of it. Companies such as Hume are creating AI models that can predict moods, attention, and emotions, and with this capability attached to a product, there’s no reason why you shouldn’t be able to predict how your users are going to feel at different stages of their journey. Some companies have already started to utilise this. Neurons, for example, is a platform that uses AI to predict where an audience’s attention will be when looking at a design/product, what their emotional response will be, how much of the information will they take in, and what will they remember. It’s unlikely that this technology will replace comprehensive testing with real users, but its applications as an automated process that can provide data instantly (with surprisingly accurate predictions) is not to be overlooked.

CLOSING THOUGHTS

The progression of AI products is fairly unpredictable, so the direction that it continues is mostly guess work. However, the future is bright for the AI-product design relationship and as designers, we get to shape the products that utilise AI so people feel comfortable supported. Innovation will not see a world where designers are replaced by AI, but one where AI technology has enabled designers to push the boundaries of what is possible and minimises the time spent on repetitive tasks. The removal of these tasks will only enable product designers to unleash their inner artists and focus instead on creativity and humanistic, emotional design.