6 ms·

About 500 people were directly endangered by this incident. The Safety Check alert seems to extend well beyond Greater London, so at least 9 million people. Tha

About 500 people were directly endangered by this incident. The Safety Check alert seems to extend well beyond Greater London, so at least 9 million people. That gives us an approximate false positive rate of 99.995%. Is that reasonable or proportionate? Does it actually provide useful information? Is it likely to quell unfounded fears and provide genuine reassurance? As tragic as this incident was, it affected a single building. Safety Check might be a useful feature in some very specific circumstances, but it's abundantly clear that Facebook is using it inappropriately. People hundreds of miles from the incident are having their profiles show "not marked as safe". The article makes a quite reasonable point - that Safety Check strongly implies that "not safe" is the default, contrary to long-established social norms. Unless used very carefully, this feature could cause completely unnecessary fear; it is easy to envision a scenario where Safety Check could prompt a dangerous mass panic. If this is a genuine public safety feature, if Facebook really are using Safety Check out of concern for their users welfare, then they should fully accept the concomitant responsibility. They should expend real resources to liaise with the relevant agencies and ensure that Safety Check is deployed only when appropriate, that it is accurately targeted and that it helps rather than hinders relief efforts. If Facebook wish to rely on their usual excuses of "move fast and break stuff" and "it's not our fault, it's the algorithm", then we should see Safety Check in those terms - as a careless and cynical effort to boost engagement metrics.

btown 9y ago
I have to disagree. "Not safe" became the default as soon as the concept of social media hit; if there's the slightest chance you may have been affected by an incident, anyone who would check your social postings to see if there was any "word from you" would assume the worst in the absence of such a posting. The human mind is evolved for both empathy and "what-if" scenarios, but not for probability; it will snap to "what if my loved one was in that apartment because they were in the country" rather than "it is highly improbable." Given that, Safety Check actually dampens those emotions on a societal scale, by making it easier and more natural for each potentially-endangered poster to assuage the fears of people in their circles.
- jfoutz 9y ago
  That's not an unreasonable stance. I lean the other way. I think, that there is an opportunity, if not a requirement, for fine grained tuning. FB knows a lot about it's users, and can do a better job of predicting who needs the i'm safe button. Clearly there's a geographic component. Someone in the area, around the time, or in the area regularly. There's probably a fancy combination of, how bad the event was, how widely the event was reported, how close you are to friends, are you traveling to the area, and a bunch of other factors. Clearly they can handle it dynamically. If you get the "are you ok?" messages, they can add you to likely events, and offer the i'm safe. They can make it easy to add yourself to a disaster event. There are a lot of people in London. They have the social graph, coupled with location and messaging data. They can do a better job at targeting. Again, i respect your opinion, but your needs, and my needs are different. I think they can do a better job at meeting both of our needs.
  - matt4077 9y ago
    Part of the problem is that peoples' worries depend not on what Facebook knows, but what those people know: I, for example, know several people in London, but have no idea where in London exactly they life. Now I didn't actually worry because that just didn't make sense considering the numbers involved, but it seems like a very understandable, human mistake to still get alarmed. That could mean displaying people safe without their input, based on location data. But that wouldn't have an impact on the problems the author is concerned about. I think it's essentially a good feature, but this incident may just not have been a good fit.
    - aaron-lebo 9y ago
      The odds of any given event impacting someone you know is so small as to be incomprehensible. Do you get worried because you heard there was a shooting in a city and it might have included your friends? You probably don't, and before Facebook you wouldn't have given it a second thought. Your friends are more likely to get hit by a car or have a heart attack and you and Facebook would never know and that's ok. Those are the odds and we don't always have to be connected. Yeah, it's a human mistake, but technology should be moving people away from animalistic fears and desires, not pushing them towards it. The easier thing to do would be to realize those odds and realize you can't control these things instead of relying on a social network as a crutch. That may be cynical, but Facebook isn't doing this because they care about you, they're doing it because it's yet another way to get people attached so Facebook can make advertising dollars. That's cynical.
seanp2k2 9y ago
"If this is a genuine public safety feature, if Facebook really are using Safety Check out of concern for their users welfare, then they should fully accept the concomitant responsibility. They should expend real resources to liaise with the relevant agencies and ensure that Safety Check is deployed only when appropriate, that it is accurately targeted and that it helps rather than hinders relief efforts. If Facebook wish to rely on their usual excuses of "move fast and break stuff" and "it's not our fault, it's the algorithm", then we should see Safety Check in those terms - as a careless and cynical effort to boost engagement metrics. " 100% agree. I think it's a well-intentioned effort gone wrong because the tuning is so off. I know a few engineers at Facebook who work on algorithms + related tooling, I'll make sure they see this.
- jakeogh 9y ago
  Excellent. I appreciate that. From my perspective well-intentioned is charitable for a social engineering platform. I have no doubt that is an accurate description for the vast majority of people involved, including the engineers you know, but this is almost a law of physics, it's too valuable to not be worth using; it's much more useful than old-school media: http://onlinelibrary.wiley.com/doi/10.1111/ajes.2017.76.issue-2/issuetoc http://onlinelibrary.wiley.com/doi/10.1111/ajes.2017.76.issu...
- afro88 9y ago
  Simple solution: have a "ask X if they're safe" button that friends/family can press. If someone presses it, that person gets a push to mark themselves safe. They don't get a push otherwise.
  - eatsfoobars 9y ago
    Actually, this button already exists [1]. [1]: https://www.facebook.com/help/516656825135759 https://www.facebook.com/help/516656825135759
  - donkeyd 9y ago
    That will have the flip side of family members getting extra worried if not marked safe, because Facebook wasn't checked for example.
an_account 9y ago
You could also just post on Facebook, "I'm okay, I survived [the event]" if you would find value in telling all your friends. Why does Facebook need to prompt?
- corobo 9y ago
  I thought it was well known and well complained about that when you post something not everyone sees it. The way my timeline seems to be right now if someone posted they're safe right this moment I wouldn't see it until tomorrow night.
  - jnty 9y ago
    If you were concerned about someone's wellbeing surely you'd check their page directly?
    - corobo 9y ago
      Fair point
- aaron-lebo 9y ago
  Facebook can't make money off that or make people feel like it's necessary to their lives. Similarly, Facebook couldn't take a cut out of people recovering from these events unless they integrated it into their platform. This is especially problematic when you consider Facebook’s business model benefits from increased engagement with its platform. Add to that, it also recently stepped into the personal fundraising space. And today, as chance would have it, Facebook announced that Safety Check will be integrating these personal fundraisers (starting in the US). An FAQ for Facebook’s Fundraisers notes that the company levies a fee for personal donations of 6.9% + $.30, while fees for nonprofit donations range from 5% to 5.75%.
- ClassyJacket 9y ago
  Unfortunately, you can't trust Facebook to actually show your posts to other people. Their "algorithm" - a cancerous wart that has infected Instagram too now - means that they show you what they want to show you, not necessarily what your friends post.
dx034 9y ago
> About 500 people were directly endangered by this incident. The Safety Check alert seems to extend well beyond Greater London, so at least 9 million people. That gives us an approximate false positive rate of 99.995%. I live in London but have friends around the world that haven't visited me yet and don't know where I live. After an event like yesterday I get a lot of messages asking if I'm ok, so using the safety check can help even though there was no remote risk that I could've been affected. I think it's reasonable that the safety check area is that large, it helps for friends/family who don't know exactly where you are.
- alistairSH 9y ago
  As a counter to that, I have friends in London and I didn't think to check in on them. Somebody would have called/texted/emailed if those friends were in trouble. As said above "no news is good news".
Nursie 9y ago
Probably way more than 9 million. I live in a small city 60 miles from London, and facebook asked me to "mark safe" yesterday because of the fire. It's nuts.
mikeash 9y ago
I'm puzzled by this idea that social norms were "safe by default" until social media came along. It's certainly not my experience. When some high-profile event happens in the area, people check in and worry if they don't hear from you. It's been like this since long before Facebook was a thing. This looks like yet another case of nerds assuming the entire world thinks and acts like we do. Yes, worrying about someone just because they live in the same major city as a disaster that affects a couple hundred people is irrational. Yes, your adds of being killed by a bus while crossing the street are far greater. Yes, just about any event which makes the news is unlikely by definition. No, none of this actually influences how normal people think and act.
- willvarfar 9y ago
  It Britain, the default is "no news is good news", which is a proverb meaning "without information to the contrary you can assume that all is well".
  - mikeash 9y ago
    I hear that phrase a lot, but is that how it actually works when things like this happen, or is it just what people say?
    - Crespyl 9y ago
      I think when things like this happen, it usually falls under the category of "bad news is bad news" and people will want to reach out to friends/loved ones who may have been affected. If my friend is going overseas for a month (and there aren't any terror attacks or plane crashes or whatever), I will probably assume they're doing just fine, haven't been mugged or gotten lost in the woods or anything unless I specifically hear otherwise.
      - mikeash 9y ago
        Well put. "No news is good news" is predicated on no news. If nothing noteworthy happens then most people are fine with assuming you're doing fine until and unless they hear otherwise. But if some disaster (even a localized one) happens in the area where they know you are, that doesn't apply anymore.
captainmuon 9y ago
You might have friends on Facebook who are not locals. When a chemical plant blew up spectacularly in Tianjin, China, I was more than an hour away in Beijing. I still got a lot of worried inquiries (from those who heard about it anyway). The same when there was a terrorist attack in Berlin. I was at the other side of the city. There was no Safety Check, but I posted about it, because events like these affect people emotionally. The fire in London was not a catastrophe of the scale of a tsunami or a major terrorist attack, but the pictures were frightening and impressive, and it is still a huge tragedy. So I can imagine many residents of London got worried messages from friends and family abroad.
- alistairSH 9y ago
  Why would people worry that you, in Beijing, was at risk from an explosion in Tianjin (100km distant)? If a bomb goes off in Baltimore, I don't call all my DC friends.
  - captainmuon 9y ago
    Because people have no sense of scale in foreign countries, that is my point. Especially if you are viewing a map of all China trying to find Beijing, 100km is not much :-). On 9/11 I was a kid living in Germany, but had relatives in California. My teachers were incredibly concerned and treated my like I was wrapped in cotton, or as if I had lost somebody personally. It's a weird thing.

btown 9y ago

I have to disagree. "Not safe" became the default as soon as the concept of social media hit; if there's the slightest chance you may have been affected by an incident, anyone who would check your social postings to see if there was any "word from you" would assume the worst in the absence of such a posting. The human mind is evolved for both empathy and "what-if" scenarios, but not for probability; it will snap to "what if my loved one was in that apartment because they were in the country" rather than "it is highly improbable." Given that, Safety Check actually dampens those emotions on a societal scale, by making it easier and more natural for each potentially-endangered poster to assuage the fears of people in their circles.

jfoutz 9y ago
That's not an unreasonable stance. I lean the other way. I think, that there is an opportunity, if not a requirement, for fine grained tuning. FB knows a lot about it's users, and can do a better job of predicting who needs the i'm safe button. Clearly there's a geographic component. Someone in the area, around the time, or in the area regularly. There's probably a fancy combination of, how bad the event was, how widely the event was reported, how close you are to friends, are you traveling to the area, and a bunch of other factors. Clearly they can handle it dynamically. If you get the "are you ok?" messages, they can add you to likely events, and offer the i'm safe. They can make it easy to add yourself to a disaster event. There are a lot of people in London. They have the social graph, coupled with location and messaging data. They can do a better job at targeting. Again, i respect your opinion, but your needs, and my needs are different. I think they can do a better job at meeting both of our needs.
- matt4077 9y ago
  Part of the problem is that peoples' worries depend not on what Facebook knows, but what those people know: I, for example, know several people in London, but have no idea where in London exactly they life. Now I didn't actually worry because that just didn't make sense considering the numbers involved, but it seems like a very understandable, human mistake to still get alarmed. That could mean displaying people safe without their input, based on location data. But that wouldn't have an impact on the problems the author is concerned about. I think it's essentially a good feature, but this incident may just not have been a good fit.
  - aaron-lebo 9y ago
    The odds of any given event impacting someone you know is so small as to be incomprehensible. Do you get worried because you heard there was a shooting in a city and it might have included your friends? You probably don't, and before Facebook you wouldn't have given it a second thought. Your friends are more likely to get hit by a car or have a heart attack and you and Facebook would never know and that's ok. Those are the odds and we don't always have to be connected. Yeah, it's a human mistake, but technology should be moving people away from animalistic fears and desires, not pushing them towards it. The easier thing to do would be to realize those odds and realize you can't control these things instead of relying on a social network as a crutch. That may be cynical, but Facebook isn't doing this because they care about you, they're doing it because it's yet another way to get people attached so Facebook can make advertising dollars. That's cynical.

seanp2k2 9y ago

"If this is a genuine public safety feature, if Facebook really are using Safety Check out of concern for their users welfare, then they should fully accept the concomitant responsibility. They should expend real resources to liaise with the relevant agencies and ensure that Safety Check is deployed only when appropriate, that it is accurately targeted and that it helps rather than hinders relief efforts. If Facebook wish to rely on their usual excuses of "move fast and break stuff" and "it's not our fault, it's the algorithm", then we should see Safety Check in those terms - as a careless and cynical effort to boost engagement metrics. " 100% agree. I think it's a well-intentioned effort gone wrong because the tuning is so off. I know a few engineers at Facebook who work on algorithms + related tooling, I'll make sure they see this.

jakeogh 9y ago
Excellent. I appreciate that. From my perspective well-intentioned is charitable for a social engineering platform. I have no doubt that is an accurate description for the vast majority of people involved, including the engineers you know, but this is almost a law of physics, it's too valuable to not be worth using; it's much more useful than old-school media: http://onlinelibrary.wiley.com/doi/10.1111/ajes.2017.76.issue-2/issuetoc http://onlinelibrary.wiley.com/doi/10.1111/ajes.2017.76.issu...
afro88 9y ago
Simple solution: have a "ask X if they're safe" button that friends/family can press. If someone presses it, that person gets a push to mark themselves safe. They don't get a push otherwise.
- eatsfoobars 9y ago
  Actually, this button already exists [1]. [1]: https://www.facebook.com/help/516656825135759 https://www.facebook.com/help/516656825135759
- donkeyd 9y ago
  That will have the flip side of family members getting extra worried if not marked safe, because Facebook wasn't checked for example.

an_account 9y ago

You could also just post on Facebook, "I'm okay, I survived [the event]" if you would find value in telling all your friends. Why does Facebook need to prompt?

corobo 9y ago
I thought it was well known and well complained about that when you post something not everyone sees it. The way my timeline seems to be right now if someone posted they're safe right this moment I wouldn't see it until tomorrow night.
- jnty 9y ago
  If you were concerned about someone's wellbeing surely you'd check their page directly?
  - corobo 9y ago
    Fair point
aaron-lebo 9y ago
Facebook can't make money off that or make people feel like it's necessary to their lives. Similarly, Facebook couldn't take a cut out of people recovering from these events unless they integrated it into their platform. This is especially problematic when you consider Facebook’s business model benefits from increased engagement with its platform. Add to that, it also recently stepped into the personal fundraising space. And today, as chance would have it, Facebook announced that Safety Check will be integrating these personal fundraisers (starting in the US). An FAQ for Facebook’s Fundraisers notes that the company levies a fee for personal donations of 6.9% + $.30, while fees for nonprofit donations range from 5% to 5.75%.
ClassyJacket 9y ago
Unfortunately, you can't trust Facebook to actually show your posts to other people. Their "algorithm" - a cancerous wart that has infected Instagram too now - means that they show you what they want to show you, not necessarily what your friends post.

dx034 9y ago

> About 500 people were directly endangered by this incident. The Safety Check alert seems to extend well beyond Greater London, so at least 9 million people. That gives us an approximate false positive rate of 99.995%. I live in London but have friends around the world that haven't visited me yet and don't know where I live. After an event like yesterday I get a lot of messages asking if I'm ok, so using the safety check can help even though there was no remote risk that I could've been affected. I think it's reasonable that the safety check area is that large, it helps for friends/family who don't know exactly where you are.

alistairSH 9y ago
As a counter to that, I have friends in London and I didn't think to check in on them. Somebody would have called/texted/emailed if those friends were in trouble. As said above "no news is good news".

Nursie 9y ago

Probably way more than 9 million. I live in a small city 60 miles from London, and facebook asked me to "mark safe" yesterday because of the fire. It's nuts.

mikeash 9y ago

I'm puzzled by this idea that social norms were "safe by default" until social media came along. It's certainly not my experience. When some high-profile event happens in the area, people check in and worry if they don't hear from you. It's been like this since long before Facebook was a thing. This looks like yet another case of nerds assuming the entire world thinks and acts like we do. Yes, worrying about someone just because they live in the same major city as a disaster that affects a couple hundred people is irrational. Yes, your adds of being killed by a bus while crossing the street are far greater. Yes, just about any event which makes the news is unlikely by definition. No, none of this actually influences how normal people think and act.

willvarfar 9y ago
It Britain, the default is "no news is good news", which is a proverb meaning "without information to the contrary you can assume that all is well".
- mikeash 9y ago
  I hear that phrase a lot, but is that how it actually works when things like this happen, or is it just what people say?
  - Crespyl 9y ago
    I think when things like this happen, it usually falls under the category of "bad news is bad news" and people will want to reach out to friends/loved ones who may have been affected. If my friend is going overseas for a month (and there aren't any terror attacks or plane crashes or whatever), I will probably assume they're doing just fine, haven't been mugged or gotten lost in the woods or anything unless I specifically hear otherwise.
    - mikeash 9y ago
      Well put. "No news is good news" is predicated on no news. If nothing noteworthy happens then most people are fine with assuming you're doing fine until and unless they hear otherwise. But if some disaster (even a localized one) happens in the area where they know you are, that doesn't apply anymore.

captainmuon 9y ago

You might have friends on Facebook who are not locals. When a chemical plant blew up spectacularly in Tianjin, China, I was more than an hour away in Beijing. I still got a lot of worried inquiries (from those who heard about it anyway). The same when there was a terrorist attack in Berlin. I was at the other side of the city. There was no Safety Check, but I posted about it, because events like these affect people emotionally. The fire in London was not a catastrophe of the scale of a tsunami or a major terrorist attack, but the pictures were frightening and impressive, and it is still a huge tragedy. So I can imagine many residents of London got worried messages from friends and family abroad.

alistairSH 9y ago
Why would people worry that you, in Beijing, was at risk from an explosion in Tianjin (100km distant)? If a bomb goes off in Baltimore, I don't call all my DC friends.
- captainmuon 9y ago
  Because people have no sense of scale in foreign countries, that is my point. Especially if you are viewing a map of all China trying to find Beijing, 100km is not much :-). On 9/11 I was a kid living in Germany, but had relatives in California. My teachers were incredibly concerned and treated my like I was wrapped in cotton, or as if I had lost somebody personally. It's a weird thing.