Explosive adoption of AI has increased productivity and efficiency, but it has created some new challenges. One of those issues is the need for human users to communicate with AI systems. For now, much of the effort to make AI more accessible is to make them conversational.
Two researchers at IBM have used their knowledge of sociology and art to find new ways to analyze conversations and use their discoveries to make communicating with AI seem more like talking with a person.
One emphasis of conversational UX design is giving AI entities a greater ability to formulate natural sentences. When perfected, users will talk with AI without realizing they are speaking with a machine.
Many commercial AI systems, including consumer-facing apps such as Apple’s Siri and Microsoft’s Cortana, use machine learning to understand what people say. The deficiency in these systems is that they require human designers to create appropriate responses.
As a result, many conversations with current AI’s seem unnatural and produce either unanticipated or confusing results.
UX design has already revolutionized the online world by creating designs from the user’s perspective. As a result websites and software applications have become more usable and users have become more satisfied. Conversational UX design aims to have a similar effect on AI.
Therefore, IBM’s researchers are hoping to help AI designers to achieve two goals. The first is to standardize the design of AI conversations. The second is to make those conversational more productive and efficient for users.
IBM now has a library of UX conversational patterns that may soon help to improve conversations with AI. Additionally, teams of linguists and scientists are working together with sociologists to gain a better understanding of how people communicate.
Despite the progress being made at IBM, the researchers seem to think that standardized conversational UX will take a long time to develop. This primarily is due to the fact that so many different organizations are working on their own set of standards.
Just as UX design has revolutionized traditional interfaces, it will play a major role in making AI more accessible and practical. Ultimately, developers can improve any system, regardless of its nature, by putting the end user first.