Why UX/UI Should Be Considered Throughout The AI Life Cycle

 SVP and chief information and digital officer, responsible for IT and security, data and analytics, and digital programs globally

For enterprises, it’s become widely accepted that the customer must be considered at all phases in the business cycle in order to achieve success. We’ve all heard the mantra, “The customer comes first,” time and time again. So, when it comes to artificial intelligence (AI), why is it that the customer (or user) is so often an afterthought?

Both user experience (UX) and user interface (UI) design play crucial roles in successful AI product development and implementations and should be considered throughout the entirety of the AI life cycle.

Research And Discovery

At the onset of a project, UX plays a pertinent part in determining what to develop. Are you going to be building something that users actually need? Extensive UX research must be completed before any product development can begin to achieve a solid understanding of how the user is currently accomplishing the task and if an AI model can make the task more efficient for them.

It’s also important to partner with the user to understand what level of accuracy is needed to both see an improvement from the previous process and ensure confidence in the AI model. It’s not uncommon for this to be a challenging measurement for users to articulate — especially if they do not have a firm grasp on their current level of data accuracy. Including an individual with a UX background in these conversations can often be very useful to help bridge the gap between the user and the AI product developers.

Product Development

Once you understand what the user wants and needs, it’s time to design the AI model with the learnings from your UX research in mind. This is when UI design comes into play and should also be considered during the development process. In fact, UX and UI design should be considered congruently during development to ensure the product meets the user’s expectations in terms of functionality, logic and visual appeal. Remember, the goal is to make the task at hand more efficient for the user, so the end product should be intuitive and something that the user actually wants to adopt.

There are also UI design opportunities within the model itself. For instance, a pop-up that communicates what the model is doing or why it is making a certain decision or something as simple as a clearly labeled button are UI design choices that contribute to a positive UX. Ultimately, communication with the user will lead to positive sentiment about the AI as well as build trust with the model — a fundamental milestone for success.

Testing And Iteration

Garnering feedback from the user at every step of the process is also a vital part of the UX process. Once the AI model has been created, it’s easy to see the quantitative data, but it’s equally as important — if not more — to place it back into the hands of the users and allow them to test it out. Ask them for feedback on how it feels. Has it improved their lives? Is it easy to use? If it’s missing the mark, what will it take to get them to a level of satisfaction where they will actually use the AI? With this direct user feedback in mind, product designers and AI product developers can make informed decisions together about what changes or improvements to implement.

Asking for feedback early and often will eliminate the dreaded result of creating a product that becomes frustrating for the user or worse — gets scrapped altogether. Including the user in the feedback loop is another way to help build trust with the AI model.

Implementation And Communication

By default, many users are fearful or hesitant of AI. They don’t fully trust the technology, aren’t aware when it is present or don’t understand what is happening, which further leads to mistrust. For these reasons, once the AI model is ready to be implemented, it is imperative that the changes made to the task or process are communicated effectively. This can be done in many different ways, depending on the situation: an email, a recorded video, a training call, pop-ups built into the UI design. The important takeaway is to communicate, for the consequences of not communicating will leave a lasting impact on the user’s trust and could ultimately cause them to walk out the door on the product altogether.

When People Are Affected

Occasionally, when asking users to be involved in the AI process, we are asking them to automate themselves out of their role. This can be a difficult situation for any business to address, and one that does not have a clear answer.

However, this is not new, as processes such as continuous improvement and Lean Six Sigma have faced similar challenges. From a UX perspective, it is important to be upfront and honest from the beginning. Working with your human resources team to find opportunities to either cross-train individuals in other skills or functions (when the situation allows) or incentivize and create retention plans to complete the AI project are great places to start. The focus should be on transparency and maintaining a positive user experience.

In a world where technology and automation have become the norm, it’s important to not let UX, with a strong UI design to accompany it, fall to the wayside. Keeping the user at the forefront of all decisions, from the beginning to the end of the AI product development life cycle, will result in a product that is trusted, well-liked and, most importantly, used.

Source Story: https: //www.forbes.com/sites/forbestechcouncil/2021/10/01/why-uxui-should-be-considered-throughout-the-ai-life-cycle/?sh=4db8c8a52fe3