Very true! That's why there are psychologists/therapists to do such stuff - analyse the signs of depression and offer help/medication.
However, in the context of the broad question, I believe there actually exists a 3rd option. Let the device listen, but the learning from it has to be local, i.e. on device, and no personal data should be exported out except for the learned parameters (or whatever your algorithm uses to recognise patterns).
Think of it the following way:
When you visit a mental health professional (psychologist/therapist, etc), you're confiding in them your personal data (opinions, thoughts, etc.) And they're bound to keep client confidentiality by not sharing your data outside.
However, over the course of years of their practice, they themselves are actually a big data store of all their past experiences with other clients and they know what would work and what doesn't, i.e. learned data parameters. Also, they actually have a clause in which they can share client details in a general sense (maintaining client confidentiality) with other psychologists for help with their case or to seek expert opinion - sharing parameters.
The following generally hold true with regard to mental health:
1. In high income countries, mental health is becoming extremely inaccessible for the middle class due to the rising costs of therapy.
2. In low income countries, mental health is almost in a negligence because the priority for wellbeing follows the following order: basic necessities for self > need fulfilment for people closeby > mental health for self. Only when one fulfils the first 2 is one able to think about their mental health. There is also the problem of awareness.
3. In middle income countries, mental health is still in negligence because of awareness. Everyone thinks they can do it themselves when they actually can't. People neglect signs of depression or name it as simply being demotivation or as people from my home country would recognise - बहाना (feign).
In all of these situations, personal devices that can recognise such patterns related to mental health, while preserving your privacy, actually makes a lot of sense. There are many forms of therapy that people can practice without consulting a mental health professional and those are actually known to improve different mental health conditions like depression, anxiety, etc.
Now, if your smart devices could actually see early signs of these conditions and recommend proper techniques for you to follow all the way while implementing local learning, preserving your privacy, promoting awareness, saving you the money, and being accessible to millions - I believe it would be a game changer in terms of well being of the population at mass.
What do you think
@SJBoss @PritishPriyam ?
Click to expand...