Artificial intelligence is having more influence on our digital life. As a powerhouse in AI, the company Microsoft is proposing several guidelines for designers when they are creating an optimal user experience for AI applications.
First, in the initial stage, you will need to understand your users’ expectations, namely, what can your AI do and how likely things will go wrong.
Second, when your users are interacting with the AI application, make sure the AI responds in a way that is scenario-based and reduces social stereotypes. Back in 2016, Microsoft released an artificial intelligence on Twitter, but it took less than a day that it was taught to be a racist.
Third, when something does go wrong with the AI, make sure your users can shut it down quickly. For example, if your Alexa smart assistant stops making sense, you can simply tell it to stop.
Last but not least, your AI system must learn to evolve as time goes by. Intelligent assistants like Cortana and Google Assistant remember your input and refer to your profile when you make a request in the future.