AI UX—Serendipitous Design

Learn about the duality of your users, and why you should design systems that have consistent behaviors, while still designing for serendipity.

Keep up to date with the AI UX YouTube Playlist

Subscribe to the Intel Software YouTube Channel


Loi, D., 2018, Intelligent, Affective Systems: People’s Perspective & Implications, Yogyakarta, Jakarta, Malang, Indonesia

Loi, D., Raffa, G. & A Arslan Esme, 2017, ‘Design for Affective Intelligence’, 7th Affective Computing and Intelligent Interaction conference, San Antonio, TX: 

Bostrom, N., & Yudkowsky, E. 2014. The ethics of artificial intelligence. In The Cambridge Handbook of Artificial Intelligence. Cambridge University Press

Sophia Chen. AI Research Is in Desperate Need of an Ethical Watchdog. Retrieved 14 October, 2017

Gershgorn, D. (2017, 30 Aug 2017). The age of AI surveillance is here. Quartz

This is AI:UX, a mini-series focus on 10 guidelines that were created to assist all those that are involved in the design and development of AI-based systems. 

I'm Daria Loi, an Intel researcher, and today I talk about guideline number six, design systems that have consistent behaviors, yet designed for serendipity. 

In my research, people expressed a fascinating duality. On one hand, they asked for consistency and reliability in their AI-based systems. But on the other, they didn't want to feel predictable, and asked for technology that could enrich their lives, while also surprising them. 

One of the research participants shared with me that they don't like the idea of an AI system changing how it acts, based on their behaviors. They said, I need predictability. But added, I want it to have consistent behaviors. 

Imagine the consequences if a system enabled you to achieve a specific goal by engaging in the same two steps daily. But suddenly, it requested steps or provided outcomes completely different from what you got used to over time. It would be confusing and frustrating, and you would feel like you were no longer in charge. 

Routine matters. And a break in routine is aggravating, as users interpret sudden behavior changes as a lack of consistency. On the flip side, another one of my participants shared their interest in an AI-based system they could listen and adjust things accordingly. When they want to make random decision or a quick adjustment, they expect the system to adjust and be able to enrich their lives, by surprising them with the unexpected. As long as it is within their comfort zones, priorities, and established rules. 

Many participants share these sentiments, and this illustrates a duality between wanting AI systems to predict their behaviors, while also being serendipitous. Remember, that human behavior cannot always be predicted. With this duality in mind, it is important to balance these contrasting perspectives in your user experience design. 

Thanks for watching. Don't forget to like this video and subscribe. I will see you next week on Tuesday for more AI:UX. 

Product and Performance Information


Performance varies by use, configuration and other factors. Learn more at