Ethics in Voice Design

The Future is Spoken presents Renée Cummings as this week’s guest. Among many roles, Renée is a criminologist, criminal psychologist, AI ethicist, data activist, urban technologist, and international consultant. Renée specializes in therapeutic jurisprudence, urban AI, ethical AI adoption, diversity, equity, and inclusion in AI. She is also the Data Activist in Residence at the University of Virginia. In this episode, Renée examines the importance of understanding and applying ethics in voice design.She discusses how her diverse background led her to working in AI. “I started to look at the risk assessment tools that were being used in the criminal justice system, and how these algorithmic decision-making systems are really misbehaving when it came to the administration of justice.”Renée explains that new and emerging technologies should be given “robust ethical guardrails” to prevent the potential harm it could cause. For Renée, ethical design starts with a fundamental understanding of what good design is, as well as considering whether that design is good for all communities. Diversity, equity, and inclusion should be an integral part of the design process. “Voice has the ability to motivate, to inspire, but it also has the ability to harm and to disenfranchise—and you have got to understand that as a designer of voice technology, you have a role.”Some questions Renée asks are: “Is that technology accessible to someone with a voice impediment or if someone is in a particular state of trauma; if someone has been harmed and they bring more emotion to their voice, or if someone has a different life experience?” Voice technology can also be problematic if misused in hiring situations. This year in particular, employers have relied more on technology in the recruitment process. Tools for hiring that are not designed to be culturally diverse or appropriate run the significant risk of discriminating against some candidates. Individuals with heavy accents, for instance, may be denied employment opportunities.To create more ethical, human-centred AI, it’s crucial that “all voices are amplified, heard, respected, appreciated, celebrated.”Renée discusses considerations when designing for sensitive conversations. Due to differing cultural backgrounds and experiences, what may be a sensitive conversation for one individual may not be for another. Even variations in pronunciation and enunciation can change the understood meaning behind a word. To ensure that no harm is done through words, designers must “bring that level of diversity to that whole concept of what is sensitive conversation.”Renée wants students who are learning design for voice interfaces or chatbots to focus on ethics and to understand that everything created will be part of a long-lasting, socio-technical “legacy.” “As designers, we've got to be bold enough as individuals working in new and emerging technologies, working in artificial intelligence: bold enough to do the right thing, bold enough to understand it is our own responsibility to educate ourselves on things like ethics.”Renée further discusses the importance of understanding and applying ethics to voice technology in this powerful and insightful podcast episode!Find Renée on LinkedIn This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit shyamalaprayaga.substack.com