Making the web more inclusive with adaptive user interfaces

I build user interface that adapt their structure, appearance and behavior to the goals, abilities, preferences and cultural norms of their users. Prior work in adaptive user interface community has demonstrated that adaptive and adaptable interfaces can improve users' performance and satisfaction. These findings alone should make adaptation a core component of the user interface design practice. But I argue that adaptive interactive systems are even more fundamentally important: they help overcome implicit biases built into most interfaces and they are a scalable approach for democratizing access to digital resources. To convince you of it, I will first present several examples of situations in which the typical one-size-fits-all user interfaces can be a source of unintended, but systematic discrimination causing some groups to be less likely than others to take advantage of a digital resource in the first place, or causing them to have a less efficient or substantially different experience compared to their peers. I will then present examples of several adaptive user interfaces that successfully provided more equitable experiences to broader populations compared to traditional non-adaptive designs. I will conclude by reflecting on the major challenges that stand in the way of broad adoption of adaptive techniques in practice. In particular, I will highlight the mismatch between the abstractions needed to develop effective adaptive user interfaces and the current software engineering practice.