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

71–80 of 160 posts

Re: Facebook deploys AI for early signs of suicide

#71

Post by bdod6 »

austincheney wrote: post #47

I so wish people would stop abusing the term AI . This is an application that employs heuristics upon scanned data and then executes events. By that definition half of everything written in JavaScript is AI.

They are using AI for natural language processing though. Heuristics would be if they just scanned a post for keywords or waiting for a suicide reporting. The article suggests they are using reporting as labeled training data, and eventually they'll be able to detect posts/videos that indicate suicide.

I agree that AI term gets thrown around a lot these days but don't think that's what is happening here.

Re: Facebook deploys AI for early signs of suicide

#72

Post by ProAm »

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.

Re: Facebook deploys AI for early signs of suicide

#73

Post by ProAm »

solaarphunk wrote: post #63
MechEStudent wrote: post #31

It is how to "cook the books" on the cost in blood of "social network". Teenage girls with gross depression are committing suicide at like 60% higher rate when they are active on social media. Facebook is absolutely a contributor there, and in order to disguise its actual complicity, its being a root-cause, it is now searching for clear signal of likelihood of suicide, with an attempt to ... what. What is their follo…

Do you have a source for the 60% higher rate? Curious to read more about this.

This [1] says it's closer to 30% but still significant (I had to look it up, and this is the first article I found)

[1] https://www.nbcnews.com/news/us-news/social-media-contributi...

Re: Facebook deploys AI for early signs of suicide

#74

Post by rosser »

evandev wrote: post #65
rosser wrote: post #56

Earlier quoted context omitted.

If you talk about it outside a privileged context, you're implicitly waiving the privacy protections. For example, if you run into your physician in public, and say, "Hi Doc. Have you gotten my test results back yet?" you're the one pointing out that she's your doctor, not her. If she, unprompted, said something about your test results to you, she'd have violated HIPAA. It's your privacy to waive as you please.

The P in HIPAA doesn't stand for Privacy.

No, it stands for "Portability."

HIPAA does, however, have a section explicitly governing the privacy of patient medical information, under Title II, generally known as "The Privacy Rule".

EDIT: Tone and specificity.

Re: Facebook deploys AI for early signs of suicide

#75

Post by dajohnson89 »

austincheney wrote: post #54
sjg007 wrote: post #23

Earlier quoted context omitted.

Interesting.. I wonder if these profiles fall under HIPAA.

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.

Re: Facebook deploys AI for early signs of suicide

#78

Post by Shank »

There's a lot of pessimism in the comments. I believe this is one of the best uses for an AI that we've come up with. First, let's go into the details:

1. They only monitor and elevate to moderators content from posts and Facebook live.

2. Moderators can then take action, and reports from this system are prioritized over other non-pressing issues.

The actual life saving chances of this system are huge. Just from a pure safety area, it can help detect and alert someone in an authority area that something bad is happening when nobody else might. The bystander effect is a real problem, and I'd imagine that it might be even greater on social media. In many cases, posts that have suicidal tendencies are screaming for help from their friends. They're posting because they have no options left. In these cases, it's a safety net that can stop loss of life. Moderators make the final say, however, so it's not just AI diagnosis. It's AI warning that gets elevated to a human before anyone else. That's more eyes on the problem that can potentially help.

Are there other implications for AI? Yep. Advertising? Yep. Other types of tracking? Certainly.

But if we're going to use AI in those fields anyway, we may as well extend it to trying to save lives.

Re: Facebook deploys AI for early signs of suicide

#79

Post by freedomben »

Communitivity wrote: post #19

This is an unbelievably slippery slope. It has the potential for great good, but also for great harm. Cliched and click-bait phrases, but I believe they are accurate in this instance. I'll try to explain why I believe that. A case has been made that Facebook users are not the customers, they are the product, and that data purchasers and advertisers are the customers. If Facebook can determine whether you are suicidal…

I agree with you. I see this being extended to other sorts of diseases/mental illnesses. If facebook thinks you are "harming" yourself with substance abuse, don't they have a moral obligation to report you and stop you? They'll do so with the opening sentence of the article:

> This is software to save lives.

Literally anything is justifiable with that pretense.

Re: Facebook deploys AI for early signs of suicide

#80

Post by phkahler »

stevehiehn wrote: post #26

Does this mean they have created a training set of posts/behaviour prior to suicide?

I was hoping they used data on people who actually committed suicide, but it looks like they're using activity that users reported and "expert opinion" instead.
Post reply on HN