AI adoption lagging? It may be a poor UI

Think about it: If something is slow, clunky, overly complicated, messy or inconsistent, you’re not going to want to use it, right? 

Well, the same goes for artificial intelligence (AI) platforms. 

In adopting them, organizations can tend to be attracted to looks — the thinking being that a shiny, sleek design will encourage use. 

Really, though, AI is only as good as its user interface (UI), said Petr Baudis, CTO, chief AI architect and cofounder of intelligent document processing (IDP) platform Rossum. Although it may seem counterintuitive, AI’s full potential comes down to functionality and simplicity. 

“For AI technology to be widely adopted at an organization, it must have a strong UI,” said Baudis. “Otherwise, its usage frequency will fall short, and the volume of recurring customers will decline.”

Infinite value — if correctly implemented

At this point, AI is just about everywhere — you’d be hard pressed to find a use case where the technology isn’t applied in some way. 

“AI has completely transformed how people work, what they do and how they spend their time — both from a personal and work perspective,” said Baudis. “The value of AI is infinite, and its usage is increasing day-in and day-out.” 

And it will only become ever more omnipresent: The AI market size is projected to grow from $86.9 billion in 2022 to $407 billion by 2027, representing a compound annual growth rate (CAGR) of more than 36%. 

As users become more appreciative of its ability to reduce human error, AI will become increasingly advanced and powerful — which is fundamental to its very nature, said Baudis. AI needs training, data and regular practice and use to fix problems, learn, improve and become smarter.

“This helps boost the value of AI technology immensely,” said Baudis. 

And “friendly UI” is critical to this — ease of use makes people come back, apply the technology daily and even recommend it to their peers. 

Ultimately, proactive AI use only benefits organizations. According to a recent survey by BCG and MIT Sloan Management Review, when workers derive personal value from the technology, there is a 5.9 times increased likelihood of organizational value. And employees who derive personal value from AI are 3.4 times likely to be more satisfied in their jobs. 

AI worth its salt

As an example, Baudis pointed to a dedicated AI engine that Rossum built for Morton Salt. 

As the leading provider of salt in North America, the company receives a steady stream of purchase orders from a variety of trading and manufacturing customers, he explained. And for a long time, each document set had its own format, making the data entry process tedious and time consuming. 

But since integrating Rossum, Morton saves as much as 95% of time per document; its average time for processing documents is 10 seconds, said Baudis. 

The engine, he said, “was built and trained specifically on their purchase orders, and to this day, is always adapting by human data validation.”

Simplified, streamlined

In assessing AI design elements, organizations should look to their building, layout and streamlining capabilities, Baudis advised. 

For example, if a UI is “overwhelming,” with excessive dropdown options or a cluttered interface, “companies are likely set up to fail.” 

UI designers must ensure that interfaces are, as he put it, “easy to navigate, cohesive, use clear messaging and accommodate users of varying skill levels.” 

It comes down to striking the right balance between looks and functionality. Don’t overcomplicate and make it difficult for users to perform common tasks. At the same time, ensure that brand identity is fluid throughout a UI. 

“Negative UI experiences can harm company reputations, so organizations must keep this top of mind when building out their AI technology’s UI,” said Baudis. 

Regular training, feedback

Encouraging employees to use AI requires a deep understanding of the technology’s value coupled with training, said Baudis. Organizations must be up front about their AI strategy and share how it will positively impact employees’ day-to-day tasks and overall experience. 

Also, companies must be transparent about goals, and illustrate real-life impact by showing the metrics that have improved since implementation. They must also provide continued training when new features are added. Furthermore, an internal feedback system can promote sharing of best practices, said Baudis.

Whether it be regular check-ins, a dedicated Slack channel or a live best practices document, consistent communication about how to use AI most effectively will improve overall performance and adoption rates. 

Also, a customer feedback loop is key. “All customers and their preferences must be kept in mind when working toward building a strong UI,” said Baudis, “especially when the goal of AI is to make lives easier and better.”