Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Facebook deploys AI for early signs of suicide

techcrunch.com

Post reply on HN
Print view View original

141–150 of 160 posts

Re: Facebook deploys AI for early signs of suicide

#141

Post by sjg007 »

arkades wrote: post #108
sjg007 wrote: post #57

Earlier quoted context omitted.

Probably.. any disclosure could potentially be regulated. This is a company creating a health care profile of you (possibly without your consent) and potentially sharing that information of you (without your consent or release) to other entities.. so maybe? I bet there will be several lawsuits to clarify things.

Unequivocally not. Posting to Facebook is a public disclosure Health data: no one, including your own physician, is obligated to protect that which you opt to publicly disclose.

We are talking about a Facebook AI program reading your posts and deciding if you are suicidal or not. I'm not even sure how you would clinically validate that and if it would require FDA clearance, but if you are forming a mental health opinion on a stream of data and contacting the government about it that is probably something that may need to be protected.

Re: Facebook deploys AI for early signs of suicide

#142

Post by rtpg »

flashman wrote: post #100
asdgkknio wrote: post #15

This may or may not be a useful technology, but the fact that Facebook thinks they have the ability and the right to diagnose their users with mental illnesses is disturbing. They have more information about their users than psychologists have about their patients. They can (and do) build a psychological profile and diagnose mental illness. Yet rather than keeping this information in the closest confidence, they sell…

> Facebook thinks they have the ability and the right to diagnose their users with mental illnesses Maybe it's more that Facebook thinks it's horrible when people kill themselves, and that if Facebook's tech could help prevent suicide, inaction would be immoral. Most of us accept some level of State interference in our lives for our own good. Maybe we are going to have to have that discussion about involuntary interv…

I think the core idea here is that if FB decides to work in this space, they should follow similar codes of ethics as, say, your local psychologist.

Its' _absolutely not_ OK for any of this information to get anywhere near the "targetted ads" part of their service. Imagine if you could advertise your alcohol products directly to alcoholics!

It's not OK for people to run explicit "psychological manipulation" experiments without the user's consent and without very specific rules with regards to the danger to the users. There's a reason why you don't replicate the prison experiment.

Maybe FB is already doing this in context of their algorithmic feed improvements. In that case they should already be asking for user consent when running these experiments.

It's not even that hard! "We're trying new changes to feeds to help show you content that is more relevant to you/more positive/etc. Click here if you want to opt-in. See here for details"

They can do right here, and I bet enough people within FB would like to do things "right", but might simply not be informed enough.

Re: Facebook deploys AI for early signs of suicide

#143

Post by alangpierce »

asdgkknio wrote: post #15

This may or may not be a useful technology, but the fact that Facebook thinks they have the ability and the right to diagnose their users with mental illnesses is disturbing. They have more information about their users than psychologists have about their patients. They can (and do) build a psychological profile and diagnose mental illness. Yet rather than keeping this information in the closest confidence, they sell…

> Facebook thinks they have the ability and the right to diagnose their users with mental illnesses

Was this ever stated in the article? There's a big difference between diagnosing someone with mental illness and recognizing when someone is in a suicidal crisis. If someone is publicly saying things like "I wish I was dead" and "I'm going to kill myself" on Facebook, it seems highly ethical to automatically escalate to a moderator ASAP rather than waiting for someone to report it.

If someone has a heart attack in the street, would you say "I'm not a doctor, so I don't have the ability or the right to diagnose what just happened"? No, you'd recognize that it's a serious emergency and find people who are able to help. This really isn't much different.

Re: Facebook deploys AI for early signs of suicide

#144

Post by Steeeve »

The fight that people are putting up is ludicrous. If you can do _anything_ to help someone who is in trouble like this, you should do it.

My hope would be that they open source whatever it is that they've come up with so that other sites can implement similar tools.

Re: Facebook deploys AI for early signs of suicide

#145

Post by sjg007 »

dajohnson89 wrote: post #75
austincheney wrote: post #54

Earlier quoted context omitted.

Absolutely not. HIPAA regulates the medical and insurance industries, which Facebook is certainly not. If you broadcast medical data to the public or put it in a newspaper HIPAA does not apply, much like this case. Best case scenario, and this is still a stretch, is that Facebook could be guilty of slander if they describe you has suicidal and that description becomes known to a third party without your consent.

>known to a third party without your consent. Sneaky suspicion: consent is given upon account registration.

Maybe, but I think health information disclosure has a lot of rules to it (at least based on my experience) where authorizations expire after 1 year etc... So I don't think a blanket exception applies. And although Facebook is a public forum not all communications are public per se so it's not clear what a disclosure is... maybe it is sufficient for Facebook to say that the information was "disclosed to us" and we are free to do what we want with it, including using an AI bot to alert emergency services. It seems like a big gray area.

Re: Facebook deploys AI for early signs of suicide

#146

Post by alangpierce »

ProAm wrote: post #72
asdgkknio wrote: post #15

This may or may not be a useful technology, but the fact that Facebook thinks they have the ability and the right to diagnose their users with mental illnesses is disturbing. They have more information about their users than psychologists have about their patients. They can (and do) build a psychological profile and diagnose mental illness. Yet rather than keeping this information in the closest confidence, they sell…

Great points, I wonder if FB actually employees psychologists and if so how many and what are their accreditations and level of professional experience.

Not sure about full-time employees, but according to the video in https://www.facebook.com/safety/wellbeing/suicideprevention, they work with suicide prevention groups and researchers.

Re: Facebook deploys AI for early signs of suicide

#147

Post by arkades »

sjg007 wrote: post #141
arkades wrote: post #108

Earlier quoted context omitted.

Unequivocally not. Posting to Facebook is a public disclosure Health data: no one, including your own physician, is obligated to protect that which you opt to publicly disclose.

We are talking about a Facebook AI program reading your posts and deciding if you are suicidal or not. I'm not even sure how you would clinically validate that and if it would require FDA clearance, but if you are forming a mental health opinion on a stream of data and contacting the government about it that is probably something that may need to be protected.

Whether or not one thinks it ought to be protected, there is no contortion of existing law that does so.

Re: Facebook deploys AI for early signs of suicide

#148

Post by jacquesm »

ProAm wrote: post #140
jacquesm wrote: post #116

Earlier quoted context omitted.

Yes they do. Main mission: to make Facebook as sticky as possible.

Is there proof though? Id assume so, Im not sure psychologists follow a Hippocratic oath if they did this seems like it would cross a bit of a line, no?

> Is there proof though?

Just one google search away. Keywords such as Facebook, psychology, experiments, ethics will generate lots of interesting reading, some of it with good sourcing and links for further reading.

It does for me, but there is always some excuse. Usually the mortgage or something to that effect.

Re: Facebook deploys AI for early signs of suicide

#150

Post by sjg007 »

arkades wrote: post #147
sjg007 wrote: post #141

Earlier quoted context omitted.

We are talking about a Facebook AI program reading your posts and deciding if you are suicidal or not. I'm not even sure how you would clinically validate that and if it would require FDA clearance, but if you are forming a mental health opinion on a stream of data and contacting the government about it that is probably something that may need to be protected.

Whether or not one thinks it ought to be protected, there is no contortion of existing law that does so.

Well take a look at this: https://nakedsecurity.sophos.com/2013/07/05/can-facebook-upd...
Post reply on HN