Making SAD Home: An exploration into developing an Alexa with depression

Abstract

While chatbots are a space that has been researched and worked on for the past few decades, a renewed industry interest in artificial intelligence (AI) and the popularity of devices like Amazon Alexa and Google Home has pushed them back into the spotlight. According to Edison Research and National Public Media, an estimated 21 per cent of US households now use a voice-enabled smart device in some capacity. Similarly, the popularity of texting, technology-mediated communication and social media has laid the groundwork for the return of chatbots. Chatbots are even making inroads into areas like mental health, where they are being used to address the growing mental health concerns of wellness and loneliness. While this is an interesting development, the conversation of what is considered useful in a mental health chatbot is still very much driven by commercial applications. This article considers using natural language processing and networking technologies to explore a more DIY approach to mental-health-based chatbots, by documenting the development of an Alexa that experiences depression.