Designing products with artificial intelligence has been an ongoing trend in recent years, as new technologies make more efficient workflows possible and people become accustomed to machine-learning capabilities. Some managers and entrepreneurs might think about employing AI power on the way of process optimization.
Artificial intelligence is developing but still, it is not good for all purposes. The question is how to do it consciously and what tasks are more suitable for automation. Here are tools and approaches that can be useful for a design team.
- enhancing the quality and reducing mistakes
- broaden creative exploration
- greater personalization
- eliminating prejudices while testing
- better design decisions through analysis
Building a user interface (UI)
AI-based programs to help auto-generate visual elements in the sketching process. Example: Auto Draw, one of Google’s AI experiments that auto-completes your sketches and turns them into more polished versions in a few seconds.
A designer draws basic sketches, a few parameters are entered and the interface renders a prototype in alignment with a company design system. In this way, designers will have more time for the strategic decisions of the product. Examples:
Airbnb solution can identify the design sketches and then convert them into coding in real-time.
Microsoft also has a similar tool — Sketch2Code uses computer vision and AI to convert drawings to working HTML prototypes.
AI makes possible an automatically refined layout, while you editing and moving interface elements. Example: DesignScape — an experiment by Adobe and University of Toronto. The tool can also propose an entirely new composition.
With AI, designers can automatically generate documentation, specifications, patterns, and anything else that is tedious and time-consuming. Example: Zeplin allows automatically generates specification details such as font sizing, colors, spacing, and other information vital to implementation. Customizes to chosen platform: Web, iOS, or Android.
Companies just have to answer a few questions and provide information about their business. Tools will generate a lot of ideas and you can choose from them. Examples: Tailor Brands and Turbologo.
Fontjoy helps developers choose font pairings, making sure they are combining fonts in the best way possible.
Personalizing user experience (UX)
AI helps to analyze how users interact with all the elements of a design system, which helps to understand which one is the best for each function. It greatly helps in optimization.
Designers very often face such tedious task as product localization — the creation of the same graphics in multiple languages. An example of a solution we get from Netflix: localization of show banners. A designer just needs to approve or reject, and if necessary manually adjust them.
Hyper-personalization, in this context, could mean using an individual’s device type and region to suggest a different content layout or variant. AI technology is making it possible for websites to take into account user activity data from the page visit before deciding what products should be shown.
Example: Netflix also uses AI for artwork personalization. A member who watches many movies featuring Uma Thurman would likely respond positively to the artwork for Pulp Fiction that contains Uma. Meanwhile, a fan of John Travolta may be more interested in watching Pulp Fiction if the artwork features John.
Research and Analytics
With the ability to collect and process more data deeper analysis of user actions and preferences is available for designers. Teams of UX professionals have the potential to tap into technology that can track and analyze large data sets, which offers excellent value. Data might include user devices, location of users, session time, session length, pages visited, exit pages, user flow. These key metrics help analysts develop a clear picture of user behaviors and interests, which is beneficial when it comes to experimenting with new ideas or running usability tests.
Employing AI for testing speeds up the process and eliminates the need to recruit a single participant. Example: VisualEyes, which simulates eye-tracking studies and preference tests with a 93% accurate predictive technology. You can achieve instant results without leaving your design tool or browser window with the plugin, which is compatible with all popular design tools.
However, there are still not many effective solutions for generating storyboards and wireframes as it needs human touch and creativity.
- let AI do routine tasks that represent more or less mechanical work
- beware of automating the wring things that need the human touch to make a human-centered design effective
- augmenting designers’ abilities, not replacing them
- remember that the goal of harnessing AI in design tools is to improve a process by eliminating tasks deemed unnecessary or of little value.