That’s why we have trials and regulations to assess this risk. Is 1000:1 a good ratio? I don’t know. But I know it is not up to us to make that call.
> But the risk of this won't stop me from looking for better answers to these problems because the answers we have right now aren't good enough.
Current answers we have for covid is not good enough either but it doesn’t mean we get to experiment willy-nilly in the public space.
I admire your self-experimentation and I actually like your tool but you can’t make money out of it while advertising in association with the claim that it fixed your depression.
You’re a small enough operation right now that neither FDA or FTC would care much about it, but the very least if you are sincere with your intentions you should put out a bold banner saying “I am not a professional, this is not intended to diagnose cure any problem, if you need help see a professional” everywhere on the site.
Also you should be careful with collecting psychometric data, not that it would put you in HIPAA land because you’re not a licensed medical professional, but it is still highly sensitive data so you should be explicit about your data collection.