Google, in partnership with the US National Alliance on Mental Illness, has recently announced a feature to help diagnose people with depression by presenting them with the PHQ-9 questionnaire if they search for depression. In what is roughly a follow-up to a previous piece about Instagram and diagnosing mental illness, I consider whether or not tech giants like Google should be dealing with this kind of sensitive data, or whether or not they are in some ways part of the problem.
Can and should one’s Instagram feed be used to predict whether or not one has depression? Recent developments in machine learning combine people’s usage of Instagram with existing data on mood and colour to train AI to be able to predict signs of depression amongst users. But this might not necessarily be a good thing.