AI gives rise to a range of entirely new types of interactions that were not possible before.
These capabilities as a new design material offer great potential for crafting unseen user experiences. They also pose new challenges, raise new questions, and require new considerations.
A few of the explicit design and UX-focused questions are:
User feedback - How does the user provide direct feedback to the algorithm about its performance?
Explainable AI - How does the interface communicate a probabilistic outcome and its reasoning to the user?
Data profile - How does the user have access and the ability to manipulate their data?
A few examples of the softer, more humanistic considerations are:
Semi-autonomous agents - Which actions are worth automating vs doing yourself? Which choices are worth outsourcing vs contemplating? What is my relationship to a tool that doesn't enable me but does it for me? At what point does anticipation become a nudge?
Affective - When do I want my technology to be aware of my emotions? How should it respond to for example sadness, anger, or happiness? How universal are displays of emotions? How do I keep myself from growing emotionally attached to my device?
Personalized - What are the implications of an algorithm knowing me better than I know myself? What is the meaning of public life and debate if we all live in our own filter bubble? What's the role of serendipity and discovery in a world of relevance?