Life, death and data. AI’s capacity to support research on human health is well documented. But so are the harms of biased datasets and misdiagnoses. How can AI developers build healthier systems? We take a look at a new dataset for Black skin health, a Covid chatbot in Rwanda, AI diagnostics in rural India, and elusive privacy in mental health apps.
Life, death and data. AI’s capacity to support research on human health is well documented. But so are the harms of biased datasets and misdiagnoses. How can AI developers build healthier systems? We take a look at a new dataset for Black skin health, a Covid chatbot in Rwanda, AI diagnostics in rural India, and elusive privacy in mental health apps.
Avery Smith is a software engineer in Maryland who lost his wife to skin cancer. This inspired him to create the Black Skin Health AI Dataset and the web app, Melalogic.
Remy Muhire works on open source speech recognition software in Rwanda, including a Covid-19 chatbot, Mbaza, which 2 million people have used so far.
Radhika Radhakrishnan is a feminist scholar who studies how AI diagnostic systems are deployed in rural India by tech companies and hospitals, as well as the limits of consent.
Jen Caltrider is the lead investigator on a special edition of Mozilla’s “Privacy Not Included” buyer’s guide that investigated the privacy and security of mental health apps.
IRL is an original podcast from Mozilla, the non-profit behind Firefox. In Season 6, host Bridget Todd shares stories of people who make AI more trustworthy in real life. This season doubles as Mozilla’s 2022 Internet Health Report. Go to the report for show notes, transcripts, and more.