Users today want content that’s customized to their interests with features that create an enhanced experience. User experience (UX) has never been more critical to engagement, but unlocking the internet will require more than 5G or HTML5. It will be artificial intelligence (AI) that truly shows off the potential of what a website can really do.
At the same time, there are so many open-ended questions surrounding AI that it’s hard to know where to begin. To be sure, AI’s impact on websites is more hype than anything else, but it does offer to solve problems that humans can’t, especially when working at scale. What’s more, there’s a lot of design that AI is actually well attuned to: much of design is programmable because it involves principles, rules, and best practices.
Where is AI? How far does it have to go? And what problems still need to be solved? Let’s dive in.
What AI offers to websites (and their designers)
AI technologies such as machine learning algorithms are currently used in features like chatbots and product suggestions, but what can they offer web design and development? From the user side, AI can help speed up search, provide increasingly personalized experiences, and help marketing teams better promote their sites. At the same time, the development and design perspective offers something else entirely.
One of the dream applications for designers and developers could leave the manual coding behind. The initial creative drawing you do could be taken through an AI-powered program to generate a prototype that’s free from human error. In fact, it already exists in the form of Sketch2Code, which translates user interface (UI) design into HTML markup using AI.
AI can also apply some of the same solutions to website development that it offers almost any other application: the ability to accelerate or automate the most tedious processes so that you can focus on the big picture ideas. Adobe Sensei is a fledgling program that already offers this for creatives, marketers, analysts, and more.
But what about the design element? A machine can learn about white space and font size, but can it learn creativity? The answer so far is a solid “maybe.”
It’s called artificial design intelligence and it’s on the way
Artificial design intelligence or ADI is the branch of AI that they say will impact the design process itself. At its core, ADI uses machine learning to identify the latest and best web design trends and put it together to create a website. In essence, you enter a series of parameters, and the ADI generates a unique or personalized design.
ADI is a bit more sophisticated than programs like Sketch2Code, which still requires the designer to provide the design itself — even if it does leave coding behind. Sketch2Code is more an automated process rather than a fully intelligent one.
The Grid is probably the most popular of the potential ADI programs: the team raised $4.6 million in 2014. However, other giants in DIY web design are also playing around with AI: both Wix and GoDaddy have introduced a basic website generator. These programs are the beginning of the idea that a machine learning program that “learns” design principles can use existing web pages to create a whole new site from scratch — not a template.
Before you begin to panic, it’s worth noting that the technology has in no way caught up with the broader idea of fully intelligent creativity from the brain of a machine. What’s more, the idea of artificial creativity will only extend so far: It might take over small projects for customers who aren’t fully sure of what they want. However, brands with a clear vision of what they require an intimate knowledge of their customer base are as unlikely to rely on these automated systems as the systems are to successfully cater to them.
E-Commerce sites will lead the way to smart design
So, where are you most likely to first see AI as a designer or developer? The most likely players to entertain a greater application of AI will probably be those in the e-commerce sector. Why? Because e-commerce giants are already leading the way. Amazon has paved the way for greater AI use with its machine-learning-capable recommendation engine, which generates 35 percent of Amazon’s revenue. Amazon’s program uses the data from each individual’s preferences, purchases, searches, browsing history and aggregate data from other customer’s purchases to create a hyper-accurate list of products customers are also interested in.
Amazon’s AI application hasn’t just generated billions in revenue. It’s also changed the way people shop online: They prefer looking for goods using Amazon’s search function rather than a standard search engine. The largest e-commerce players (Amazon, Alibaba, etc.) will all have voice- and potentially visual product search by next year (2021). That means designers for SMEs in the e-commerce sector will soon need to catch up.
That’s an engineering problem, not a design issue, you say. And you’re correct. Designers won’t be on the front lines of implementing increasingly complex AI into the structure of the site, per se. However, it will dramatically impact all the classical aspects of web design and development. From layout to font to color selection, you’ll need to design these websites for a UX experience that’s led by machine learning, which will create a completely new customer journey.
AI, UX, and designers
UX is another intersection of AI and web design, and it crosses at multiple points. AI will better allow designers to keep up with changing behaviors and trends while putting usability first. After all, UX will be at the core of both customer demands and users’ journeys.
One of the ways AI can do this is by identifying the correct visuals for the site. (Remember that visual search could be ruling the world by the end of 2021.) AI allows you to skip the image name and categorize images based on pattern structures. In theory, AI can then help you customize a landing page by choosing an image that perfectly matches users’ previous purchases or other behaviors. All of a sudden, you’re more likely to show the user an image that’s truly useful to them rather than an image that you think looks good strictly from a design perspective.
This practice is where current software like Adobe’s Sensei (mentioned above) comes in. The assistant currently provides recommendations for the best layouts and images based on the site’s function or category. As the process develops away from the human-augmented design (a category that Sensei currently falls into), you could be using a very sophisticated tool that’s hyper-personal.
AI will also serve design/developer adjacent purposes
AI is likely to have impacts on website development that fall somewhat outside the scope of the typical designer-developer, particularly in terms of cybersecurity. At present, the use of AI for attacks is largely limited to ethical hacking and “white hat” experiments, in part because the programs take an enormous amount of resources to build. Frankly, plenty of criminals have just as much luck with established, “low brow” measures like phishing and social engineering, so there’s less need to try to implement AI, just yet. Even still it’s something that developers should be keeping an eye on.
Other AI technologies are also infiltrating the project management sphere, which impacts the administrative side of design and development. You’ll soon find that AI will help you collaborate with team members and clients, assist with providing estimates and budget tracking, and help you tackle data analysis when required.
Ultimately, AI is more likely to impact all the other facets of running a web design or development business before it has any meaningful impact on design itself. What’s more, there are still fundamental problems to be solved: AI can only design based on the data it has, and it’s a fundamental challenge because human creativity and inspiration can outwit it every time.
What’s next in AI? Designers and developers will need to wait and see, but we can sit confidently in the knowledge that so far, AI is here to help not hurt.