How Machine Learning is Changing Design
Source – tech.co
Pew Research Center survey in 2015 revealed that almost two-thirds of Americans expect computers to take over much of human work within 50 years. However, there are certain professions that some consider to have more security over others. Disciplines such as the arts require creativity – a uniquely human trait that computers should not be able to grasp.
Unfortunately, this may not be the case for long. Developments in artificial intelligence (AI) and machine learning (ML) are now making headway even in creative work. The unprecedented amount of data humans are generating is enabling AI to generate insights on what makes certain visual designs effective. This, in turn, is giving rise to design automation apps and services to fill common design requirements of businesses and enterprises.
Here are three areas where AI and ML are changing the way design is being done.
Branding defines a business’ identity and visuals such as logos aid greatly in brand recognition. Consistent presentation of branding increases revenue by 23 percent. Because of this, it has become common for marketing teams to have in-house graphic artists and designers in order to apply the brand’s design language into all visuals they produce.
Small to medium enterprises (SMEs), however, may not have the luxury of having a dedicated human resource for such purpose. It is common for SMEs to outsource or find low-cost solutions to their design needs. This is why there is still an abundance of poorly executed brand designs abound today.
Tailor Brands, for instance, provides an alternative to freelance design work by using ML to automate the process. Users simply have to input a company name and answer several questions about the background of the company in Tailor Brand’s logo designer to generate unique logos. The company also offers a full brand design service which includes landing pages, seasonal logos, business cards, and presentation templates which are created using the ML-driven technology.
Yali Saar, CEO of Tailor Brands, believes that by applying design principles and best brand identity practices, an algorithm would be able to select the appropriate typefaces, forms, and colors to create logos that capture the company’s identity. Tailor Brand’s system learns with every input so that it can even factor in emerging design trends.
User experience (UX) design has become an essential part of building a solid digital presence for businesses. UX designers strive to create interfaces with minimal friction so that users are able to intuitively navigate and use websites and apps. Intuitive and frictionless sites and application typically lead to better user engagement and higher conversion rates for businesses.
As such, UX has been the focus of many analytics efforts. Behavioral analytics and heat maps could track a visitor’s journey and record each interaction made on the interface. Data from these analytics efforts reveal insights that could guide designers as to which designs for content, navigation, forms, and buttons lead to better conversions.
Organizations now enjoy easier means to putting up a digital presence. Web design has evolved to a point where services like Wix allow users to create and design their own websites using templates and drag and drop interfaces. These services take into consideration what has been identified as best UX design practices and incorporates them in the elements that site builder users can add to their projects. Users are now able to build effective designs that convention without having to know and write a single line of code
Still, AI and ML-driven services seek to build upon these by further automating the process. Instead of a drag and drop interface, Firedrop users could design web pages through a chatbot. The Grid, also offers a website designer that only needs input such as content and the desired color palette to automatically build a fully-functioning website.
While these advances in ML-driven design automation are picking up steam, there are still certain requirements that need human creativity and intervention. Fortunately, the impact of AI and ML isn’t only confined to automation. These technologies are also helping improve tools designers use.
Design software giant Adobe is building their own AI dubbed as Adobe Sensei. Adobe Sensei is used to power various Creative Cloud software. For instance, it enables features such as face-aware liquify in Adobe Photoshop which allows artists to change the facial expression of a photograph without causing distortion. Adobe Sensei also powers search of stock images to allow in-depth filters based on image characteristics.
Getty Images also has the Pen project which allows users to sketch a concept or idea which then gets digitized, recognized by AI, and used to search for a corresponding image. This greatly helps designers who need a particular image with the particular composition they have in mind for their work. It solves the problem of trying to find a specific image based on its content and composition using a text-based search engine.
Work in Progress
AI- and ML-driven design is a work in progress and is bound to get better in time. While one can argue that these services may still lack the full creativity of human designers and artists, it is also difficult to discount the value automated design services bring depending on the context.
SMEs with minimal resources to spend on branding and UX design could definitely consider using these AI-driven design automation tools. Designs that follow the fundamentals and best practices are still bound to be better than shoddy workmanship.
Still, highly-skilled creative professionals may still rest easy knowing that it may still take some time before robots take over their jobs. For now, they can simply enjoy better tools to help them do their work.