Facebook's African Sweatshop
time.com
time.com
I ended up canceling my following meeting and talking through things with him.
He was close to suicide, had cut contact with most people in his life and said that he hadn't been able to sleep or think straight since he started the work. Unfortunately I couldn't offer him any work, only a shoulder to cry on and a promise that there was beauty in the world as well.
I never thought a job could do so much mental damage to someone, but these jobs are horrible.
Any company, large or small, that allows uploads of image content without strict identity verification of the uploader is going to be a vehicle for this junk. I had a super popular website in this space, as soon as I figured out what was going on I shut it down, I am of the opinion that it is not possible to run a service like that in a responsible way without causing damage somewhere.
I always thought that a better model for moderation would be to distribute the effort. There will be people in my social bubble with similar reactions to the same content and once an algorithm understands these connections, it can distribute the effort of moderation amongst these like-minded people.
Another moderation tool would be to simply allow people to trade reputation (which has some value in terms of audience reach) for correctly tagging content.
On a different note, if you outsource your censorship, you will undoubtedly end up being censored based on the political ideology of the person doing the censoring.
[1] https://jobsfactor.eu/job/49057/dutch-content-moderator/
I don't want reactionary right-wingers on Facebook moderating other reactionary right-wingers' content on Facebook.
I agree with you, but for different reasons. A ""sweatshop"" is much better that the agricultural and industrial protectionism used by the US and the EU to prevent African crops from reaching their markets.
If you think about it on a systemic level, these subsidies are quite evil: we subsidize agriculture so that we don't need to buy their crops, then even worse: we dump the leftovers under names like "international aid" which limits the opportunity for a local industry to grow (same with clothes "donations": the local industry can't compete with a price of 0)
In an ideal world, we would not do such things, crops would be grown where the weather and the labor market provide a cost advantage (Africa is warmer than Europe!), the consumers would benefit from cheaper prices, and other countries would industrialize, grow their middle class and GDP.
It's not good for ones country to be dependent on another for survival (food) which is why they subsidize local production. This even occurs on the state level in the US where (I only lived in costal states) costal states enforce a minimum percentage of land towards farm usage.
International aid is just not letting a disaster go to waste. Obviously you'd allocate federal money to go to some corporation to administer food in disaster area otherwise they'd go under and not donate to you. Any second order effect of wrecking economies and etc won't be attributed to you so there is only upside.
Unless you're arguing the Sahara is a better environment to grow crops than the French countryside, that's not an advantage...
As for the Sahara itself, all it needs is water, and I don't see any factor that would make the cost of drop-by-drop irrigation not competitive or produce more CO2 than keeping giant plastic tents warmed up in Europe (and irrigated, harvested...), especially considering the availability and abundance of solar power in the Sahara of all places!
The African issues are many and multi-faceted, they won't be solved by growing more stuff than they already do.
There was alot of very insightful writing during the New Deal era, where federal price supports were introduced. Later, in the late 60s/early 70s, stabilizing prices was a policy position taken to blunt the wartime inflation from Vietnam and the Cold War.
That’s not to say there aren’t any downsides. The destruction of Mexican corn agriculture post NAFTA is an example. The world could drown in the bounty of cheap American corn.
Africa doesn't grow all that much so simply being warmer is a moot point. and yes, africa is a vast continent but the regions that permit agriculture are few. Southern africa and the Mediterranean to name just two.
They do not need to be done.
Facebook needs them to be done for its business model to work.
You may have meant (surely did mean?) the latter, but I think the distinction is important.
None of these people need to be subjected to psychologically harmful material on an industrial scale. We could always just not let Facebook do that, and they'd have to live or die as a company without doing those things. Possibly it would mean broadcast-type social media couldn't exist anymore. If so, then the cost of keeping that seems to be hurting people with this kind of work. But it is something we could stop doing if we wanted to.
Unless you're suggesting that we return to an era in which the only way to publish content is with the approval of someone like a newspaper editor or book publisher, these jobs will need to be done.
It'd be significantly tougher for advertisers and PR agencies, however.
Good. We shouldn’t allow habitual liars to determine what’s possible, or make their lives easier. I’d happily grind the social media, advertising, and public relations “industries” into dust for the good of us all.
"Yet you participate in society. Curious!"
Your stated view isn't that many-to-many moderated social media website should do better, it is that many-to-many moderated social media websites shouldn't exist. Because this is a recreational activity, and because the alternatives you suggested are massively popular, 'Yet you participate in society. Curious!' does not apply.
Message received loud and clear.
It's not a good message, to put it about as nicely as I possibly can, but I've received it.
Nooooo. I mean, yes, content moderation in the modern sense of "trying to mitigate the insanity that is randos posting a firehose of crap that's by-default instantly published to the world and which our stupid algorithms may well choose to shove in lots of people's faces" but no in the sense of selecting what one chooses to publish from among some submissions, which pre-dates the Internet sense of "content moderation" and doesn't seem to be a problem unless you have a publish-by-default-without-review policy that lets some amount of the crap get through, and barely vet your submitters at all.
> None of that is reasonable or desirable by most humans.
OK. So we choose—choose—to have a job category that will cause significant, lasting harm to some percentage of the workers doing it.
But let's stop also insulting them by claiming we're doing it because we "have to".
It's so weird that defenders of algorithmic hate machines accept them as some sort of natural law, forgetting that humanity somehow muddled along for millennial without the benefit of dopamine hijacking.
Now every god damn idiot can have a Twitter page or a Facebook echo chamber group and easily spread their bullshit there. Yeah, I wouldn't mind going back to that era you mention.
usenet and web forums have existed for a very long time before blogs. It didn't take much to spam newsgroups or online bulletin pages.
You have surely never operated a website that publishes user generated content. Every available textbox and file uploader will be filled with garbage so revolting that, without moderation, will quickly kill any chance of respectability for your site.
If you have advertisers (or just standards) you'll also need to check it for horrible stuff that's not illegal. And by "you'll", in the case of a place like Facebook, I mean "some other people'll".
But you could... not do that.
But you could... not do that."
Hacker news and Facebook:
1. Have advertisers (and just standards)
2. Check it for horrible stuff that's "not illegal"
3. Using "some other people'll" that are paid with money.
Now, here's the thing. Facebook and Hacker news are not equivalent in quality or quantity when it comes to how much illegal or horrible content is uploaded to them. But they both have advertisers, a risk of bad content, and use at least one paid moderator to prevent bad content from existing. Brimble is saying that Facebook and other hypothetical services could "not do that", which either means that:
1. The website / service will not exist or
2. Paid moderation on that service will not exist or
3. Moderation on that service will not exist or
4. User created content on that service will not exist.
That would affect Facebook and hacker news. No, I do not believe that hacker news is exposed to as much illegal and horrible content as Facebook. However, it is exposed to some, and it seems like the remedy in both cases is for the website to not exist. I'm pretty sure that if someone asked dang whether they had ever, in their years of moderating this site, seen something that disturbed them, he'd answer yes. I'm certain of it.
If you want to know my opinion though, I don't personally get enough value out of facebook, instagram, etc to use them at all. So if you were to ask me if the price of psychological damage to content moderation/review workers is worth it, I'd say no.
Let me spell it out for you. * The peasant doesn't get to choose whether he participates in society. * The peasant wants to improve society.
Here's you: * You get to choose whether you use hacker news, and you have decided to use hacker news. * You think that websites like hacker news should not exist.
Do you see why your argument isn't convincing? Your riposte of "yet you participate in society" could be used for literally anything. You're saying that hacker news shouldn't exist, but you don't actually believe that. If you argue things you don't believe, it's not really a good faith discussion.
If we're gonna cite the comic, three out of the four panels of the comic do not rely on this, and are precisely about using something by choice even while saying they should be different. The last panel is even more extreme because it's funnier that way, presumably, but the core point does not rely on lack of choice.
The other two examples are: posting from an iPhone that Apple ought to treat its workers better, and buying a car without seatbelts while saying cars ought to have seatbelts.
> but you don't actually believe that.
Nice mind reading. Shall I baselessly speculate about your posts while accusing you of being a bad actor, too? No, no I shall not.
This is more akin to posting from an iPhone and saying phones shouldn't exist.
You may also be surprised to find out I'd prefer much more bike- and walking-friendly city planning, even to the point of making it harder to drive places, while in fact I drive almost everywhere.
Dealing with the current reality does not prevent one from advocating something else. Am I the one taking crazy pills, thinking that's super-obvious?
The fact that you supposedly hold such a strong position as saying 'this platform shouldn't exist' when the cost to you for not using this platform is minimal makes very little sense to me. For the example of a phone, it's not entirely incongruent for someone who's already bought an iPhone to be upset with Apple, and if they generally wish to use a phone, other companies might not be much better. But if somehow they get the notion that phones or smartphones shouldn't exist at all and continued using them, that doesn't make sense. I know people who get around fine without smartphones. If somebody really thought smartphones shouldn't exist yet continued using them, I question the strength of their convictions, and am more inclined to believe they hold their position more for the value of controversy than genuine belief.
I object to the idea that we have to psychologically wreck some workers. No, Facebook (and Twitter, et c.) do, to stay in business, evidently. We don't have to. We could say they aren't allowed to do that anymore, as we have with other worker safety issues in the past. We could even do that while exempting very small low-harm operations and cases in which that kind of work really is something resembling necessary (police work, say). There's no "gotcha" in "but what about police investigators?" or "but what about HN?" (the two I've seen in this thread) unless one ignores the reality that similar worker protection regulations deal with those sorts of edge cases, routinely. Even if the regulations couldn't have any nuance (why not? It's not been a problem any other time), I'd personally be OK losing HN over that, sure.
You seem to be advocating a position that content moderation, and services that rely upon it, should not exist. I'm not sure why Hacker News would somehow be an exception to this, so you seem to be advocating this site should not exist. If you're not advocating this, perhaps you should explain how this site, which consists entirely and exclusively of user-generated content, would get around without content moderation, or why that content moderation would be acceptable.
> I'm not advocating that position because I don't like using HN.
I'm not sure what that has to do with it. If I heard someone advocating that, for instance, meat should be abolished because it's murder (or something along those lines), I would be perplexed if the person advocating that continued to eat meat. It wouldn't do anything to lessen my confusion if they told me they're not advocating that because they don't enjoy eating meat; that part makes sense to me. What doesn't make sense is both having that point of view, and continuing to to engage in that behaviour.
Now, if this individual instead had the position that meat should be taxed, or that people should cut back on their meat consumption because of its impact on the environment, I don't see an incongruence there. But if they hold the position that it should be banned entirely, that doesn't strike me as a deeply-held conviction if they continue to eat meat.
Likewise, for our NYT op-ed writer, if they, e.g., once read newspapers and then decided they were garbage that needed to be abolished, then perhaps there's no incongruence there. If I see him next week with the Sunday Times in hand, I would then gather that perhaps the position he advocated he wasn't serious about.
This is covered in the post you're responding to, directly. And others in this thread. I don't think it's particularly important or helpful that HN be outlawed, but I might be willing to accept that for fixing worker abuse in social media companies generally. I doubt, however, that'd be necessary, for reasons that are, again, covered plenty well in other posts of mine here. The TL;DR is that it's totally normal to have regulations, including for worker protection, that kick in only at a certain scale (say, employee head count) and that could probably work here.
I'm not set on or advocating outlawing HN. My entire original point was that harmful social media moderation jobs do not have to be done. Facebook (among others) choose to operate in a way that makes them necessary for their bottom line. Further, we implicitly choose to allow these jobs to exist by not making jobs that are psychologically abusive illegal without an excellent justification, as we do with jobs that are physically abusive or dangerous without a great reason to be so. This is close enough to regulations we already have for worker safety, that I don't think it's wildly outside the bounds of behavior we (at least in the US) already regulate.
I doubt HN in particular is harming moderators very much, for a bunch of reasons including its size and the form it takes. Probably not a ton worse than your average customer service job. And maybe better, since no-one's yelling abuse at them on the phone.
There is no "gotcha" here, for multiple reasons.
OK, then that's at least a substantial comment on why you feel that HN is different, rather than the evasive ones you've been giving throughout the thread; you could have just led with that instead.
Couple posts up. Cut the passive aggression and read.
You're also really avoiding the fact that you're not describing an improvement, you're describing abolishment. Hacker news only has one single feature, and that's user input. The comments are submitted by users. The articles are submitted by users. If you get rid of user input, then this website is an opening and closing HTML tag. The comics describe a scenario where something is improved. What would improvement of hacker news be to you?
Let's ask this:
1. Do you think that hacker news should exist in its current form?
2. If not, what would you change?
3. Do you use it?
4. Do you think that it should exist?
5. In what ways does Hacker news differ from facebook, such that a positive answer to 4. would have a negative answer to 5?
It's not, though? Else we are all constantly hypocrites in this regard. We may think the electoral system is broken and should be reformed, yet rarely vote for a candidate who supports that (because the current system provides few such candidates), et c. So calling it out seems pointless. We all deal with imperfect information and with the choices that the current system presents, but not others that might exist (or, rather, represent more-practical alternatives than they do now) instead, if the system were different. Prefacing criticism with public confessions when it's not very relevant is just a waste of space, and is stating the obvious.
> You're also really avoiding the fact that you're not describing an improvement, you're describing abolishment. Hacker news only has one single feature, and that's user input. The comments are submitted by users. The articles are submitted by users. If you get rid of user input, then this website is an opening and closing HTML tag. The comics describe a scenario where something is improved. What would improvement of hacker news be to you?
The "something" being improved would be worker protections, essentially. Publishing other people's stuff would probably require human review of it all. Given that AFAIK publishing houses aren't inundated with tsunamis of abusive content, I don't think that level of review is a problem in practice, as far as being abusive to the people doing the work—rather, it's the fact that abuse gets through with some regularity, that draws it in the first place. What's done now with review of user-posted online content clearly is hurting people, and it's the way certain businesses operate that are causing the problem.
If you asked me to draft legislation for this, I'd at least consider a carve-out for very small (low head count) operations. That's already super common in all kinds of business and, specifically, worker protection, regulations.
> 1. Do you think that hacker news should exist in its current form?
It doesn't seem to be a significant part of the problem, no, but I'd happily lose it in the name of reforming social media (=publishers with terribly lax submission and go-to-print standards) generally.
> 2. If not, what would you change?
Oh man. Nothing related to this, actually. Probably put in better poster-identity-tracking capabilities so it's easier to remember when you've already been down an unproductive garbage-thread with someone, or watched someone else do it (I do not have present company in mind here, to be clear) and over time people prone to creating those stop drawing replies at all and hopefully go away. There's a lot of inside-the-rules shitty posting that can't easily be user-policed right now, as it is on some other sites. I get the motivation behind weakening poster identity across threads, but I think it's beyond clear that it does more harm than good, now. Posters I read elsewhere today on HN, who were wondering how some subreddits can see HN as a cesspit: this is a big part of why.
The main downside is that meta-discussion about who's what kind of poster is low-value noise, but norms around here do an OK job of policing worse than that (provided the content being policed follows a clear pattern), so I think it'd work out OK. Just allowing personal tags for users, maybe with color-coding (visible only to the tagger) for posts, would go a long way. Some kind of hide-user functionality might help too, though it'd take some thought to make it work OK with HN's threading. Probably have to hide entire sub-trees they're at the root of, but I've definitely seen users for whom doing that would strictly improve my reading experience and I wouldn't miss a thing worth reading.
> 3. Do you use it?
LOL.
> 4. Do you think that it should exist?
Again, I'm not sure the entire category of thing is, fundamentally, not a crazy thing to bring into the world. It's like having an open email relay then trying to police messages passing through. Clearly insane. But, also again, most of the harm seems to kick in at scale. So, I dunno, maybe.
> 5. In what ways does Hacker news differ from facebook, such that a positive answer to 4. would have a negative answer to 5?
Chiefly in size. Fewer users, fewer employees, way less visual or auditory media to evaluate as a percentage of posts (I'm counting links, or it'd be zero). Arguments in favor of one and not the other are purely the practical sort—what's doing harm in practice, and how do you narrow in on that as specifically as possible—not the iron-clad-pure-reason sort; but, again, that's a pretty common approach with this kind of thing, for good reason.
That passive-voice "social media couldn't exist anymore" implies an Orwellian bureaucracy that meticulously controls who is allowed to talk to whom. Someone has to enforce it.
No one is forcing Facebook to harm people like this. They choose to, in order to make money. We choose to allow the creation of jobs that are guaranteed to cause that kind of harm. My objection is to framing this as work that "has" to be done, as if there's no choice.
> Human beings enjoy many <-> many communications and will invent something to fill that void.
Right. We have Email and messaging apps. Websites would still exist and are dirt-cheap to start up, if someone wants to write for the public. Possibly social media could survive somehow, IDK, but they're (and their supporters) the ones saying they "have to" work this way. If so, maybe they should "have to" stop being in business.
There's a lot of variety in human brains. What will traumatize one person is comedy for another. There are (or at least used to be) websites you can seek out to look at pictures of grizzly death. It's simply not the case that everyone is damaged by looking at terrible things, and - to cut off that line of thought - the people who aren't are not monsters or sociopaths. They're just made of different stuff. I can't comment on whether Facebook et al are careful enough in their hiring practices, but I'm certain that there are plenty of people in the world that can do this job.
Email lists were popular before Facebook took over that niche. Nobody more than I would rather see the comeback of mailing lists, but that doesn't solve the problem. You can assume that if mailing lists got the kind of uptake that Facebook has, gmail/hotmail/etc would have the same content curation issues.
I'm still not sure exactly what you are proposing. Facebook gives me a way to communicate with my friends. It sounds like you want to take that away.
I want us to stop allowing that business models that rely on harming workers must exist. We may choose to allow it, but it's a choice. They do not "have to" do it, and we do not "have to" let them do it.
I hate discussion of harm to these workers that frames it as "well, it sucks, but we have to". We. Do. Not. Have. To. We may choose to, but we don't have to. We don't even have to if we want efficient mass communication. We have to if we want a few publishers to be able to corner and monetize the attention economy.
I'm sure the Linux kernel mailing list gets tons of vile things sent to it all the time which are filtered out.
I don't see why we have to involve business models at all. Any large all to all communication medium is going to have this issue and the only way to prevent it is to devolve back to the cable tv/newspaper era where people only consume content.
Obviously though a competitor with even less scruples will rise up and gobble up their marketshare if they did something that drastic. So the net result makes no difference.
To actually prevent child pornography from getting shared would require a dictatorship censoring the entire internet. Sure it would be the government instead of multiple big private companies, but that seems even worse than the current situation?
And if you go the other way and have millions of little networks, how do you plan on stopping all the bad stuff from being shared without an even more powerful and centralized dictatorship deciding who can form networks or not?
If you mean to get rid of all censorship and let free sharing of child pornography, etc., that will not pass any democratic legislature, at least in the West, since the vast majority of people want to limit that.
So in the end all alternatives would lead to even less freedom. Have you actually thought through any of this?
Facebook's a magazine publisher at massive scale, that relies on worst-in-class submission and acceptance processes to make money. They're not in the file storage business.
I managed a web forum dedicated to one particular hobby. It was quite popular. Yet simply from spam management we disallowed auto activation of new account. We asked people to reply to a few questions in order to know a bit more about them and their intents. Yes we probably lost a large number of potential users but we actually got the users that were really motivated and not thousands of shadow accounts waiting for activations by bots as before.
I guess it wouldn't be out of question for a platform that allow anyone to put content to ask for a mandatory copy of the ID of the user as well as a receipt for a electrical bill or something proving where they live in order to be able to report them to their local authorities if they post bad content. All these measures can be defeated but the amount of content to moderate would be in a much lower scale than when you allow accounts with random names and no validation.
So, as something somewhere between a newspaper and a telecom.
I suppose it could have been carried out in the late 80s through the telephone system if people left messages by calling 1-800-FACE-BOOK, and maybe touch-tone dialed text for the title, and then to check what your friends wrote you called 1-800-FACE-BOOK again, and after combining your friends' submissions it would call you back. It would let you select which you wanted to play by operating on the digital titles, again by dialing, and when you found one you wanted to hear it would play you the recording your friend wrote. And not only would it be less viable, in the 80s it would cost money to use, and have no advertising.
Just not viable. These days, it's viable, and there has just been too much in capitalist subsidies to return to a time before images. There was a time when condensing the sight before your eyes into two dimensions meant buying rolls of photos, taking a dedicated device that could imprint on them the light of reality, giving it time to use the material well until there was enough to justify a trip to get it developed--first dropping it off, writing down in a paper agenda to return in a few days, then returning and paying--and only then would you witness what the camera witnessed. And then cataloging the photographs in an album, and keeping that album on a shelf with all the other books--there were all sorts of different books for all sorts of things--and finally, years in the future, reviewing the photographs to relive the past.
But so much has been spent into making this process trivial. I would estimate trillions in R&D and subsidizing the semiconductors that made all the machinery that came before it a thing of the past. Not to mention bandwidth, without which sharing images on the web was still possible over the course of hours.
On the other hand, you're doing the right thing, right now, in choosing a social network that doesn't use images in almost any sense, has two moderators, and is broadcast-based nonetheless. Your post was just broadcast, brimble.
Mine is an argument against "it's different because it's on the Internet", which as far as I can tell is the whole of the argument for this practice. Put it in the real world and it's instantly clear that they should just stop.
> On the other hand, you're doing the right thing, right now, in choosing a social network that doesn't use images in almost any sense, has two moderators, and is broadcast-based nonetheless. Your post was just broadcast, brimble.
"Yet you participate in society. Curious!" :-)
Any kind of large scale communications platform that allows file/photo sharing will need to do something like this. In fact the requirement to do it is about to get stronger if the UK has it's way.
The alternative isn't particularly fun, letting bad actors monopolise a communications channel to broadcast literal shit at me, my children and friends.
Possibly, accepting submissions from anyone on Earth and then publishing most of them automatically is fundamentally a bad idea and should not be done.
> The alternative isn't particularly fun, letting bad actors monopolise a communications channel to broadcast literal shit at me, my children and friends.
I don't see how this follows. If anything, the centralization of attention driven by the current system is what does that.
There would still be things to moderate, and people beating the validation system but the traffic of content to moderate would be much lower.
I dunno, one day (years ago, now?) I was yet again reading a post like "well it's a shame these workers suffer horribly and maybe we can do some stuff to help, but the work has to be done" and it occurred to me that I was reading someone defending billions-of-dollars corporations with exactly the same reasoning that a kid uses when they say another kid "made them" punch them, and then we were all hand-wringing over which color of medical tape to put on the victim's nose (I mean, clearly the kid had to get punched, everyone agrees on that and you must be crazy if you don't, but we should try to help!) instead of stepping back and going "what if the problem is that there's punching happening in the first place?"
The whole framing of the problem, as it's most commonly expressed, is myopic and deflects from whole categories of solutions (that are very inconvenient to a few Internet giants that dominate the attention economy).
A company can't outsource that kind of responsibility to a bunch of algorithmically selected users.
This can work in some scenarios, or when the junk is basically randomly distributed — e.g. flagging people spamming random business links — but it's a lot harder when you have coordinated attempts with resources. That's been heavily used to game things like Facebook or Twitter and it's not just bots — someone saying “go here and retweet it” can generate a lot of activity from humans who are decidedly not neutral. That can be done for profit, political reasons, or griefing and the latter two can be quite hard to detect.
I'm not sure we have a great solution for this other than destroying the model those companies currently have of surfacing popular/trending content based on the activity of people you don't follow.
That may be "the important part", but the article is very clear about the fact that, in you know, actual reality -- the workers at this company definitely are not receiving meaningful support.
Protecting the local population from immoral companies is a priority/duty and some large companies need to be given ultimatums immediately.
Importing the sickness of the Western countries is not a requirement for Africa to continue rapid development.
In an actual free-market system, Facebook would want to reduce costs because it wants to reduce prices to stay ahead of competition. But Facebook doesn't have much competition, or monetary prices. And so Facebook doesn't act like a competitive company, it acts like a monopoly maximizing its exploitation of a resource it controls.
If Facebook were acting competitively, it would be reducing its price (fewer ads) when it reduces costs. But it actually does the opposite.
Which takes me back to the original question: When Facebook is in this market-power position, is it really it's prerogative just to make as much money as possible?
This could be prevented by changing the law so that it treats any large enough social platform as a communication service provider, requiring it to provide open API with complete feature coverage. This would facilitate native feature-complete third-party cross-platform clients that don’t show any ads, and platforms would have to charge users because no one will be stupid enough not to use a client like that.
And once service is no longer free and cross-platform clients exist, the space will finally start resembling an actual free market—honest competitors will stand a chance, and actual end users will be able to vote with their wallet.
> allow third parties to inter-operate with the gatekeeper’s own services in certain specific situations
With “certain specific situations” presumably they are trying to make it viable for a monopoly to innovate, but as is it’s open to interpretation and may just result in platforms innovating specifically to maintain monopoly and the double-market scenario where they sell to advertisers and only care about users insofar as their eyeballs don’t leave en masse.
IMO a good law would specifically aim to make it infeasible to operate as a double-sided market and pretend to offer a free product to the users while actually being supported by advertisers. I suspect this can only be achieved by an all-encompassing interoperation requirement (a.k.a. open API). In this document, however, “interoperability” seems to mean hardware and OS (?).
Twitter allows exactly that (or used to, unsure if the API is still up): an API for third parties to use it's content and make alternative front-ends, like Nitter. How many people use them? I don't know, but I'd bet my liver (non-binding offer) it's a minuscule fraction of it's users.
Browsers allow to purge just about every ad from your Internet in a handful of clicks. How many people use them? A mere 27% in the US [0]
Non-Chrome mobile browsers allow extensions to purge ads too (savvy Google killed the Chrome store as they couldn't just kill adblockers, yet managed to stay dominant.. power of brands). How many people use them on Android? A laughable 14% [1]. Of which a fraction will be ad-blocking.
People are not idiots, but it's just not as simple as "offer an ad-free/superior alternative and people will flock to it". People are creatures of habits, if they learned to go to twitter.com or facebook.com, it doesn't matter how much better your alt-front-end is; even if you don't have to compete with the network effect as you'd share the backbone, it will be a steep, steep uphill battle to gain traction.
0: https://www.statista.com/statistics/804008/ad-blocking-reach...
And if you think the arms race to hide ads on the Web as a whole is somehow similar you are missing the point. Yes sure, the more people use the extension the less ad money, but guess what—platforms find ways to work around that, lobby for new browser specs, etc.
If it remains the main revenue stream, money will find a way, so that root cause should be eliminated.
And the outcome could be a unified single interface that can tie together your social communications across different platforms—if it were possible, would you not use it?
If "I" would use it is entirely irrelevant, all I meant to point out is people don't flock to better alternatives just because they exist. I think that most of them not even be bothered (or knowing to) de-clutter their Internet browsers is a good example of that point.
Because this is the kind of thing the open API would enable.
This doesn't track for me. Let's assume there are 2-3 other competitive social networks that FB has to compete against. FB figures out how to do $thing that reduces its costs by 10%.
Why does it follow that FB is going to show fewer ads now?
The end winner is the consumer in such a scenario as prices keep dropping.
But here, the price we pay is ads, and it's hard to even evaluate where you saw more ads. People don't pick a social network by the number of ads (short of there being so many it gets in the way), they pick social networks by where their friends already are. The bigger a network is, the more likely someone is to use it.
Facebook's economics exploit this network effect. They don't need to make the 'price' better (ads) because that isn't what drives consumers to it.
Flip that question around.
Given that we know from various leakers, that facebook doesn't have very much capacity for anything that isn't english, do you think they have the expertise to recruit, train and verify people at scale in kenya?
I mean _they should_ but they clearly dont.
https://africacenter.org/publication/islamist-extremism-east...
It's all social media.
I guess the only ones that get away with a moderation made by unpaid volunteers at scale is reddit, since it's their entire business model. The real question is, what kind of individual is "moderating" that place...
I suspect that Facebook does ban folks but that the people who do it attempt to get around the bans. If this could be solved by banning the people who do this, I imagine Facebook would have done that already and pocketed the 5 Billion dollars.
There are a lot of shitty people out there doing shitty things to get attention.
Considering the size of the platform ~$130B/year; $5B per year spent on platform safety, while substantial, seems low as a percentage of revenue.
There can’t be that many suicide vids etc floating around the internet. Catalog that sht once and use AI similar to YouTube and copyright detection.
Sure there will remain a continued need for 2nd line humans but this really shouldn’t be this much of a mental sweatshop operation.
Vids especially.
Written text is a bit harder because anyone can easily produce unique content
Also FB being front and center on this issue seems to be a recurring theme…yet they aren’t even the big video destination
And video is somehow less unique, and easier for AI to categorise and identify?
> There can’t be that many suicide vids etc floating around the internet.
So very little training material for the AI, then?
> they aren’t even the big video destination
More than 4 billion video views take place on Facebook every day. [0]
I really despise FB. But you're trivializing the problem.
This type of video - yes. You and I are both writing unique content right now. Redditors produce thousands of unique posts daily.
There are only so many suicide videos on the internet though and the amount a person can produce per day is inherently limited.
It follows that fingerprinting said videos should be feasible.
>So very little training material for the AI, then?
And? Youtube does this with a single example for copyright purposes. Not all algos have to be backprop DL
In general, it's easier for a computer vision system to recognize and filter a video that has already been banned--though, there is a constant arms race here as well--than it is for it to judge the content of a completely new video. That means that, for a huge number of cases, a human being will have to see the footage at least once.
EDIT: To be clear, I'm not taking up for FB in this situation. I'm specifically clarifying the difficulty of using ML/DL in moderation systems.
It critiques working conditions without providing any context as to how these jobs compare to other low pay -or more likely even average pay- jobs in those countries.
Formal economy jobs paying $250-$300/mo with a steady employment outlook for entry/low-skill work can be hard to come by in many -if not most- countries of the world, never mind Africa alone.
Can you please cite a source for this claim? While it's true that Sama has partnerships with companies such as, according to the article and to Sama's website[1], Google, Microsoft and Wal-Mart, it's not entirely clear that they're also doing content moderation for all of them. They seem to offer a wide range of services.
>Formal economy jobs paying $250-$300/mo with a steady employment outlook for entry/low-skill work can be hard to come by in many -if not most- countries of the world, never mind Africa alone
Ah, right, because if you're lucky enough to be making more than your peers, you should shut up and deal with doing so in a horrible environment.
Not precisely, but a person's choices should generally be considered in light of the alternatives reasonably available to them. If nothing else, it's often a great way to better understand the choices people make. Not every person has fast, easy, low-risk access to a six-figure USD SWE job with a YCombinator startup.
These are the best jobs these people can get. If the "sweatshop" closes down, they're be doing somewhere worse.
Maybe you throw out this shitty provocation as a way to avoid thinking about that fact?
Edit: We can - and should - be working to improve conditions for everyone, but just because "these people" have it better relative to everyone else in their immediate vicinity doesn't mean they need to shut up, be thankful and deal with it.
So we don't seem to be in the same conversation.
>These are the best jobs these people can get. If the "sweatshop" closes down, they're be doing somewhere worse.
The implication that I read here is that you're saying "these people" should be grateful for having to work under the conditions they're working under because life could always be worse and this is the best they will have it.
Is that an accurate read of your comment?
I just meant what I wrote.
The implication, to spell it out, is that if you succeed in closing down these jobs, you'll make life worse for the workers you claim to care about.
BTW, why do you quote "these people"?
Please find where I said we should close down these jobs.
Have we maybe run out of disagreements?
To me, if you can get qualified workers for $2/hour, that's all you should pay. End of discussion.
If people accept the job offers, you are, by definition, not underpaying. It clearly helps the local economy, lifts people somewhat out of poverty, and is a classic win-win situation.
I dont want to seem to defend Facebook or the company mentioned in the article. However there are unintended consequences when the labor market is abruptly distorted by foreign companies.
When the first outsourced English speaking call centers opened in my city, they paid more than the average for English speakers. The foreign call centers paid about 50% more than the average salary for English speakers.
So many English teachers left their jobs and joined the call centers, including some of my teachers in high school. So we were left some months without much to do at school or with definitely underqualified temporary teachers who spoke English worse than we did. I think about 40% of our school-year was affected for this.
The next school year, the school had to increase tuition, and hired new teachers with higher salaries.
On one hand teachers salaries increased, but also school became more expensive for my parents to pay while my faimily's income remained the same.
> It critiques working conditions without providing any context as to how these jobs compare to other low pay -or more likely even average pay- jobs in those countries.
What context could justify the global power imbalance that makes this economic relationship possible? The article is critiquing working conditions and power imbalances. I'm sorry it also doesn't take the time to explain and rationalize the economic injustice that created them.
You seem to using a liberal or neoliberal framework to understand this relation in terms of its individual actors and the economic incentives and consequences of them. The missing context here isn't more details about those actors and consequences, it's an exploration of why these factors are distributed the way they are and how to engage with that system.
I guess you completely glossed over the main part of the article:
> They perform the brutal task of viewing and removing illegal or banned content
Look, I'm a capitalist just like the next person, but subjecting ANY human being to the disgust of banned content on Facebook, Microsoft, Google, etc. HAS to take an immeasurable mental toll.
But hey, I guess you just focused on the "but how much are these guys making"...
Does it not nag you to be employed by one of - if not the most - ethically disgusting companies in the world?
Do you guys just click "hide" on all these articles - on average 1-2 a day - where Facebook has committed yet another absolutely disgusting transgression?
So either he rationalised it away, or, like anything political, half of what you read is blatantly false, and the other half is mostly false.
Lots of people decide that they just want to make a living and can't always find a job that is ponies and rainbows, I think. I've myself thought about working for a local surveillance-related company but it disgusted me a bit and I was in a fortunate position to be able to find something else.
2. If you job depends on your not understanding something, most people won't understand even if it is common sense. The economic incentives don't align
3. It is easy to blame the institution vaguely like all the people who didn't directly kill/torture anyone in Nazi Germany nevertheless enabled the oppressive government to function. [2]
---
[1] More extreme forms we see today all the time with vaccine and other science denialism enabled/ fueled by platforms like Facebook
[2] Comparison with Nazis is not to equate with Facebook and offend anyone, it is to illustrate that even in the extreme case like that unless you are directly involved easy to ignore and support/enable atrocities at that level let alone what companies like Facebook / Google do
> Does it not nag you to be employed by one of - if not the most - ethically disgusting companies in the world?
I'm sorry - I don't work for a "defense" contractor[1]. That said, a lot of societal problems are pinned on FB. IIRC, YouTube moderators also suffer from PTSD, I suspect the same goes for Twitter, and Vimeo[2] - and I believe more should be done - saying this is a FB-specific problem is just dishonest, but convenient.
Getting rid of Facebook will not fix the deterioration of social fabric and degraded, that is a convenient lie people tell themselves. Go ahead and look at YouTube comments, misinformation, or Twitter threads, and and ragebait purveyed by - lets see - everyone. Society is fucked and cannot be fixed, at best, it can only be mitigated.
This is not to excuse Facebook, but just setting the tone on where we are as a society. The reason it's hard to legislate a fix for this because the required changes are going to affect everyone in ways they don't like, therefore we can't even come to a consensus on what the solution ought be.
Getting to your question, FB is a big company, and there are a lot of orgs that do great work that benefits humanity, those don't get as much media attention. That said - I really was "radicalized" on HN: I am a minority, and for years, I've witnessed the cavalier attitude or outright disdain shown by HN towards the more problematic areas of our industry (racism, sexism, etc). What can I say, the "Fuck you, got mine" attitude is infectious[3].
1. I'm being snarky, but I think writing software to take lives is the most disgusting thing fellow SWE's do.
2. Does Vimeo they even moderate? All the gore/"watch someone die" links I have not visited were exclusively Vimeo
3. It's also the best solution to iterated prisoners dilemma scenario
This company has the image of helping lift people out of poverty - I think they are doing good work. The article shows though that if nothing else there is a danger of trying to go with a "we're different and better" corporate image because the truth is often murkier
Also surprised there aren’t more content moderators. FB (just FB alone for now due to them being a monopoly) should be required, IMHO, to review every piece of content before it is shown in a feed using human review. We don’t let movies get shown before they get reviewed (e.g. PG13).
Should they pay that much in a really poor country. What about a really rich one (may be Switzerland)?
End of the day quality of living is totally dependent on local conditions (barring gadgets like ipricy)
You seem to be wildly underestimating the complexity of the task and the long tail of content creation. There are 1.9 billion daily users.
You’re asking to REQUIRE a human content moderation system that reviews 100% of the continent that 1.9 billion users can create a day? Before that content can get shown? I don’t see a feasible way of creating a system like that which will serve the needs of those 1.9 billion users in a way that will seem remotely tolerable to them, that can timely share their content.
Have we actually decided that "scale" is a good a justification for behaving badly, or are we simply accepting it without critical thought because we once respected those companies?
IMO, companies do not have some inherent right to profitability, or even to remain in business, if they can't responsibly conduct their business.