The Dream Machine is a design concept that uses AI and interaction design to help people clear their minds. Many people struggle with ruminating on thoughts, especially before falling asleep. Usually, people process their thoughts and impressions from the day during sleep, which can manifest itself through dreaming. Can we use machine learning to help people process their thoughts? Can it help people see new perspectives, and close their days with more ease and mental space?

The Dream Machine makes people’s daily thoughts tangible and collects them through voice. At the end of the day, it is designed for slow and reflective interactions by users inputting their thoughts and “seeing” them being processed by the machine. We wrote a JavaScript program that converted user input into emotional states to guide them through questions reflecting on their thoughts. The program then used Rytr and DALL-E 2 to create a reflection card that held both a story and a picture created by AI based on their daily thoughts. This reflection card could be kept and stored for further reflection. The project was implemented as an interactive exhibition as seen in a sketch below.