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

151–160 of 160 posts

Re: Facebook deploys AI for early signs of suicide

#151

Post by tareqak »

danso wrote: post #127
tareqak wrote: post #98

Please permit me to play the pessimistic man on the street: If Facebook can deploy AI to detect the early signs of suicide, why can't they deploy AI to detect "fake news" or those susceptible to "fake news"? Furthermore, "cui bono" (for whose benefit [0])? What does Facebook gain from being able perform this sort of detection given it is a for-profit corporation and not a non-profit philanthropic organization? An asi…

I'm not a psychiatrist nor have any expertise or insight into suicidal behavior, but I feel very confident in saying that the detection of a potential suicide is a much easier problem than the detection of fake news. There's a few reasons for why suicide seems easier to detect, but I'll just give one: it is feasible to conceive of a reliable and reasonably efficient threshold for detecting some suicides. And I can sa…

Thank you for that detailed reply! I really appreciate it, and I read through all of it. One thing that I forgot to mention were that the some of the ads, if not all, run by the IRA (Internet Research Agency [0]) were veritably false. If Facebook kept score of ads that were known to be false run by an ad buyer couldn't they present that information beside the ad e.g. a sort of running "boy crys wolf" heuristic?

Thanks again for your well-researched and extensive reply.

[0] https://en.wikipedia.org/wiki/Internet_Research_Agency

Re: Facebook deploys AI for early signs of suicide

#152

Post by danso »

tareqak wrote: post #151
danso wrote: post #127

Earlier quoted context omitted.

I'm not a psychiatrist nor have any expertise or insight into suicidal behavior, but I feel very confident in saying that the detection of a potential suicide is a much easier problem than the detection of fake news. There's a few reasons for why suicide seems easier to detect, but I'll just give one: it is feasible to conceive of a reliable and reasonably efficient threshold for detecting some suicides. And I can sa…

Thank you for that detailed reply! I really appreciate it, and I read through all of it. One thing that I forgot to mention were that the some of the ads, if not all, run by the IRA (Internet Research Agency [0]) were veritably false. If Facebook kept score of ads that were known to be false run by an ad buyer couldn't they present that information beside the ad e.g. a sort of running "boy crys wolf" heuristic? Thank…

Yes, I can definitely agree that there is far more that Facebook can do in investing enough manual effort into shutting down outliers in egregious abuse.

Re: Facebook deploys AI for early signs of suicide

#153

Post by MichaelMoser123 »

MichaelMoser123 wrote: post #49

the article does not mention how a possible intervention would look like; if they don't report a user with suicidal posts to the authorities then facebook will be sued by relatives of those who put an end to their lives; if they do report it then they will be sued by privacy advocates. Damned if they do, damned if they don't.

data is power and profit (P&P), but too much data can be a liability.

Re: Facebook deploys AI for early signs of suicide

#154

Post by timthelion »

drb91 wrote: post #130
timthelion wrote: post #111

Earlier quoted context omitted.

I'm worried that the field of psychology is gaining a monopoly on expression of feelings. We can no longer express happiness and sadness, anxiety and frustration, without being "diagnosed" in a horoscopic manner which has little or nothing to do with science and bears a great deal of stigma. The psychological disorders of today are the occult of yesterday. We believe that one can become possessed by depression or anx…

You are free to reject diagnosis and medication unless you are determined a danger to yourself or others. HIPAA also gives you control over where your diagnoses travel. Not with facebook, of course, which doesn’t qualify for hipaa protections yet.

Of course you are free to do so, but back in the days of daemon possession, most people were just as free to reject the claim that they were possessed. What mattered back then, was not the claim though, what mattered back then was that ordinary people would look at someone and think "oh he's acting weird, he must be possessed". And that "possessed person" would then be feared. It interfered with the process of empathy.

Re: Facebook deploys AI for early signs of suicide

#155

Post by austincheney »

fragmede wrote: post #126
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.

Facebook is known to collect key presses. If I express a desire to commit suicide in a status update, but don't actually post it, thus not showing it to anybody except Facebook, would that (should it?) be covered by HIPAA? What if I keep paper diary? What if it's in Google docs instead?

No, because you have no obligation to disclose anything to Facebook. If you choose to disclose this to Facebook that is your mistake. You absolutely must disclose medical information with a medical provider and with your insurance company.

> What if it's in Google docs instead?

If you didn't put it there than somebody has violated your privacy or violated HIPAA.

Re: Facebook deploys AI for early signs of suicide

#156

Post by austincheney »

sjg007 wrote: post #145
dajohnson89 wrote: post #75

Earlier quoted context omitted.

>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 ar…

> Facebook is a public forum not all communications are public per se so it's not clear what a disclosure is...

It doesn't matter. Once you surrender data to Facebook it is theirs to do whatever they want with. That is the disclosure. There is no gray area. If you don't want Facebook having that information then don't give it to them.

Re: Facebook deploys AI for early signs of suicide

#157

Post by __warlord__ »

This reminds of Psycho-Pass[0] an anime where pretty much an entity determines if your mental state is stable enough for society, long story short, it doesn't work for everyone in the sense that if the system "decided" that you are "unstable" is over for you.

I really hope the people at facebook are aware of the repercussions of a system like this. But then again, greed and market share is something that facebook prefers over its users.

[0]. https://en.wikipedia.org/wiki/Psycho-Pass

Re: Facebook deploys AI for early signs of suicide

#158

Post by arkades »

sjg007 wrote: post #150
arkades wrote: post #147

Earlier quoted context omitted.

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...

The medical school structured itnas a clinical trial with people volunteering to participate. Medical studies have different regulations than just reviewing public postings.

This is a reflection of the way they’ve chosen to go about things - no doubt in response to past kerfuffles - rather than an actual legal obligation on fb’s analysis of public materials.

HIPAA regs aren’t hidden from the public. Peruse at your leisure.

Re: Facebook deploys AI for early signs of suicide

#160

Post by drb91 »

timthelion wrote: post #154
drb91 wrote: post #130

Earlier quoted context omitted.

You are free to reject diagnosis and medication unless you are determined a danger to yourself or others. HIPAA also gives you control over where your diagnoses travel. Not with facebook, of course, which doesn’t qualify for hipaa protections yet.

Of course you are free to do so, but back in the days of daemon possession, most people were just as free to reject the claim that they were possessed. What mattered back then, was not the claim though, what mattered back then was that ordinary people would look at someone and think "oh he's acting weird, he must be possessed". And that "possessed person" would then be feared. It interfered with the process of empath…

Well, HIPAA also protects you from your doctor telling your social circle. So i’m not quite clear on what the parallel to “being afraid of a possessed person” would be.
Post reply on HN