HWAI
How Weird Am I?
(how weird are you??)
The idea for this project came from working in this field as a data scientist for the last ~decade. Right now there's a ton of money being thrown around with the goal of developing "biomarkers" to improve treatment of human diseases. And often that money goes to building datasets using cutting-edge biological platforms (VC loves anything you can put "-omics" after), to profile study populations of a few hundred to a few thousand people.
For some things this approach makes sense - if you want to help people with lupus, measuring proteins in people's blood might give a strong enough signal to guide treatment. But for mental health, this won't work. To better understand and treat depression, anxiety, addictions, post-traumatic stress - blood biology is a lot less useful, and the complex, overlapping nature of these things makes it even harder.
For behavioral health, it makes more sense to build a dataset of... behavioral data. Basically just ask people about their habits / personality / mental health over time. Collect data from as many people as you can, to account for noise in the diversity of human brains, and collect it over time, to account for noise from the fact that people's lives vary over time.
The funny thing is, it wouldn't cost millions of dollars to do this - just the hosting fees to run a website like this one (so far, about $15/month).
So yeah, if you took the baseline survey - awesome, you already helped build this thing that could help people someday. If you want to do more, answer more questions or leave some feedback. So far we're at 290 users.
or for a long(er)-winded take on these ideas, see here.
For some things this approach makes sense - if you want to help people with lupus, measuring proteins in people's blood might give a strong enough signal to guide treatment. But for mental health, this won't work. To better understand and treat depression, anxiety, addictions, post-traumatic stress - blood biology is a lot less useful, and the complex, overlapping nature of these things makes it even harder.
For behavioral health, it makes more sense to build a dataset of... behavioral data. Basically just ask people about their habits / personality / mental health over time. Collect data from as many people as you can, to account for noise in the diversity of human brains, and collect it over time, to account for noise from the fact that people's lives vary over time.
The funny thing is, it wouldn't cost millions of dollars to do this - just the hosting fees to run a website like this one (so far, about $15/month).
So yeah, if you took the baseline survey - awesome, you already helped build this thing that could help people someday. If you want to do more, answer more questions or leave some feedback. So far we're at 290 users.