Over here we are led to believe that AI handles soul destroying things like image identification and horrific content moderation, but maybe it's as naive as thinking Nike's sew themselves....
Moderators were confronted with an unending stream of sexual abuse, torture, violence and beheadings. They were trained to watch the first and last 15 seconds of a video and to scroll rapidly through the rest, stopping at potentially problematic parts. An efficiency target, known as an AHT, or “average handling time”, meant dealing with each “ticket” in 55 seconds.
Over here we are led to believe that AI handles soul destroying things like image identification and horrific content moderation, but maybe it's as naive as thinking Nike's sew themselves.
Distressing but important read. Be back later with additional comment but I wanted to insert this excerpt quickly to see if we can get more readers and more insight.
In its quest to make ChatGPT less toxic, OpenAI used outsourced Kenyan laborers earning less than $2 per hour, a TIME investigation has found ...
... OpenAI sent tens of thousands of snippets of text to an outsourcing firm in Kenya, beginning in November 2021. Much of that text appeared to have been pulled from the darkest recesses of the internet. Some of it described situations in graphic detail like child sexual abuse, bestiality, murder, suicide, torture, self harm, and incest.
OpenAI’s outsourcing partner in Kenya was Sama, a San Francisco-based firm that employs workers in Kenya, Uganda and India to label data for Silicon Valley clients like Google, Meta and Microsoft. Sama markets itself as an “ethical AI” company and claims to have helped lift more than 50,000 people out of poverty.
I'll be honest, working with that kind of content sounds absolutely abhorrent, but they're doing a great service for everyone else filtering out all that toxicity. Even if it's kind of fucked, I'm...
I'll be honest, working with that kind of content sounds absolutely abhorrent, but they're doing a great service for everyone else filtering out all that toxicity. Even if it's kind of fucked, I'm okay with workers voluntarily signing up for a job that pays almost double the average hourly income in Kenya (per Hacker News). Quickly filtering out disturbing text descriptions doesn't sound nearly as dangerous as working in a lithium mining pit.
I think it's really screwed up that the job market in Kenya is such that that's considered an acceptable and already "better" option. (I live in a country with similar conditions. Content...
I think it's really screwed up that the job market in Kenya is such that that's considered an acceptable and already "better" option. (I live in a country with similar conditions. Content moderation laborers are commonly outsourced from here too.) I understand why it's a "better" option, but it's still not ok and it should still be improved.
Maybe this is dumb question, but how would you want content moderation to be handled? Ignoring the profit incentives with AI training data, human moderation jobs can still be grueling and...
Maybe this is dumb question, but how would you want content moderation to be handled? Ignoring the profit incentives with AI training data, human moderation jobs can still be grueling and thankless. Fairly compensating workers at double the local market rate who could've freely choosen other jobs seems fine to me.
Fairly compensate moderators in the same country as the user base? Americans moderate Americans, Germans moderate Germans, etc. People might balk at the idea of content moderators being paid...
Fairly compensate moderators in the same country as the user base? Americans moderate Americans, Germans moderate Germans, etc.
People might balk at the idea of content moderators being paid $200k, but I'll bet there's burnout like no other. Maybe comparable to doctors and therapists.
$2/hr seems reasonable for Kenyans moderating Kenyans, not for Kenyans to wade through the mountains of shit the rest of the world churns out courtesy of exploitation of CoL differences.
I think moderators from the same community as they're moderating may have cultural context advantages, but I really don't understand why companies shouldn't take advantage of CoL differences. I...
I think moderators from the same community as they're moderating may have cultural context advantages, but I really don't understand why companies shouldn't take advantage of CoL differences. I don't think that's inherently exploitative. Even within a country, it makes good sense for companies to hire remote employees or set up operations in regions with lower costs of doing business (ignoring tax breaks).
This generally has a positive effect on the region and boosts incomes and standards of living since wages rise (an actual example of a rising tide lifting all boats). If a company moved in near me and offered to double my salary, I probably would jump ship if the work wasn't too bad. Multinational companies investing in poor areas and raising average wages should be a good thing, no?
To address this for you, @ibuprofen, and @HerosJourneyMadness. Because most CoL differences are due to availability of infrastructure and labor rights. And if the promise of 'lift out of poverty'...
why companies shouldn't take advantage of CoL differences.
Because most CoL differences are due to availability of infrastructure and labor rights. And if the promise of 'lift out of poverty' starts happening, companies will probably just pick up shop and find the next country with the lowest wages and fewest restrictions.
I firmly think free global trade of goods and labor is only morally acceptable if wages and labor laws have a reasonable minimum standard worldwide.
I mean we can look at China as a successful example of exactly what I'm talking about as far as international investment goes. The standard of living in China has massively increased in just a...
I mean we can look at China as a successful example of exactly what I'm talking about as far as international investment goes. The standard of living in China has massively increased in just a couple generations due to global trade and industrialization. Partially due to economics, but mostly due to politics, many companies are now "decoupling" and moving labor to countries like India and Vietnam, and now those countries are starting to experience the exact same increase in living standards. The median Chinese citizen is significantly wealthier than two generations ago; that's the primary reason the CCP has such strong support.
There should absolutely be a minimum global tax to prevent tax havens, but I think the international exchange of goods, labor, and people is a universal good that benefits the whole world.
Which is good for China and no others. The generations of "international investment" has left the countries that invested in Chinese trade and industrialization with dearth of the skills and...
Which is good for China and no others. The generations of "international investment" has left the countries that invested in Chinese trade and industrialization with dearth of the skills and infrastructure to be self reliant. It effectively lifted boats in China, and will do the same in other Asian countries, while sinking the boats that did the investment.
The numbers can't always go up, there's a fixed amount of resources and time in the world; the tide doesn't rise, it comes in at one place and it goes out in another.
Look at towns where auto manufacturers moved from the USA to other countries, or even elsewhere within the USA. Many are husks of their former selves, turning into wastelands of poverty and all of...
Look at towns where auto manufacturers moved from the USA to other countries, or even elsewhere within the USA. Many are husks of their former selves, turning into wastelands of poverty and all of the drug abuse and crime that tends to come with it.
And I'll bet those new workers don't have access to nearly the same QoL that the unionized workers of the old plants had.
There should also be a global minimum wage equivalent to 40 USD, regardless of the local average wages. The subsequent inflation will sort things out in short order. Bet a lot less factories and...
There should also be a global minimum wage equivalent to 40 USD, regardless of the local average wages. The subsequent inflation will sort things out in short order.
Bet a lot less factories and call centers move around arbitrarily when that happens.
In agreement with the other reply to an extent. CoL based wages CAN be used I think acceptably and morally- but I don’t think it is very often. The liberal media (I am liberal) I think tries to...
In agreement with the other reply to an extent. CoL based wages CAN be used I think acceptably and morally- but I don’t think it is very often. The liberal media (I am liberal) I think tries to adhere to some narrative that it’s ALWAYS exploitative. Truth is, I have no idea.
I’ll tell you what though- it wouldn’t be too hard to have mental health experts go in and establish some rules around how to humanely do such a job. Limit hours, rotate with other less stressful duties, etc.
I’d also bet there are things that could be done within the workflow to further limit exposure (specifically like using screenshots to sample video).
IMO this is yet another story about needing real, moral rules around employment. Also- offshoring labor itself should be a source of tax revenue for the US so the Midwest can stop losing electricity multiple times every winter.
The problem I see is that it isn't just CoL based wages, but often the lack of the surrounding supporting infrastructure and labor protection. If if the wages or good for the country people can...
The problem I see is that it isn't just CoL based wages, but often the lack of the surrounding supporting infrastructure and labor protection. If if the wages or good for the country people can often be made to look for longer shifts at this stuff and if it does cause mental damage it is unlikely that mental healthcare is available.
The flip side to this is that a job paying double the average wage is probably massively boosting their standard of living. The difference between making $1 and $2 per day in an extremely...
The flip side to this is that a job paying double the average wage is probably massively boosting their standard of living. The difference between making $1 and $2 per day in an extremely impoverished nation is massive. Sure, they don't have mental healthcare or the resources to maintain a first world lifestyle, but the people working these jobs are lifted out of poverty.There aren't many easy, well-paying alternative job widely available in poor countries.
This is how economies are built up and countries become wealthy.
I literally live in a country where this happens all the time and I'm telling you that it doesn't lift people out of poverty (it just makes them a little bit less poor), nor has my country become...
I literally live in a country where this happens all the time and I'm telling you that it doesn't lift people out of poverty (it just makes them a little bit less poor), nor has my country become wealthy.
That's the core of my argument. People are better off and can raise their standard of living because these jobs are available. Countries benefit from international investments that raise wages and...
it just makes them a little bit less poor
That's the core of my argument. People are better off and can raise their standard of living because these jobs are available. Countries benefit from international investments that raise wages and bring in money. Kenya won't become wealthy just because one multinational company offers some jobs at nearly double the local wage, but it's a start.
You're looking at it from a purely financial perspective. But being slightly better off financially at the expense of their mental health from having to regularly view traumatic content, doesn't...
You're looking at it from a purely financial perspective. But being slightly better off financially at the expense of their mental health from having to regularly view traumatic content, doesn't sound to me like they're actually better off overall. I used to work in Computer Forensics, and that field has similar issues due to investigating traumatic incidents and content (like CSAM) also being a major part of the job... which can be really detrimental to your mental health, with lasting negative effects.
So I can't help but feel like it's horribly unethical to offload this sort of moderation work to a country that likely has less worker protections, and significantly less mental health support and resources available.
But that would eliminate almost all of the Kenyan jobs. Think what you will about how the job should be done, the reality on the ground is that this is a very well paying job that Kenyans are...
$2/hr seems reasonable for Kenyans moderating Kenyans, not for Kenyans to wade through the mountains of shit the rest of the world churns out courtesy of exploitation of CoL differences
But that would eliminate almost all of the Kenyan jobs.
Think what you will about how the job should be done, the reality on the ground is that this is a very well paying job that Kenyans are choosing because it's preferable to their other opportunities. Kenyans are better off if Meta is pressured to provide better employee support than they are if those jobs disappear.
It's like asking how industries should handle toxic waste: you begin by making plans to limit the production of it as much as possible. You don't turn the machines on live and then try to find the...
It's like asking how industries should handle toxic waste: you begin by making plans to limit the production of it as much as possible. You don't turn the machines on live and then try to find the most vulnerable humans to wade through it as an afterthought.
Humanity doesnt need Facebook: we would lose very little of value if it disappeared overnight. it's up to them to figure out how to limit harm before they do harm.
Maybe it's text only.
Maybe images and videos must be tied to a legally traceable account: if you upload illegal horror you're going to jail. Currently the penalty is some poor human will have to watch it, and then report it to some other poor human to moderate, and the uploader has in the mean time made 50 more accounts.
This idealism ignored a fundamental reality of the internet: those who can do, have already did. Lots of the harms aren’t easily foreseeable: those warning of them before they happen will again be...
This idealism ignored a fundamental reality of the internet: those who can do, have already did. Lots of the harms aren’t easily foreseeable: those warning of them before they happen will again be dismissed as cranks in the future.
I don't think its idealism to require, by law, billion dollar companies to not spew toxic waste into our planet and poison human beings, or at least to provide personal protective equipment and...
I don't think its idealism to require, by law, billion dollar companies to not spew toxic waste into our planet and poison human beings, or at least to provide personal protective equipment and long term health benefits for personnel handling its clean up.
And just like toxic material mining: if you can't afford proper environmental clean up then you're not allowed to operate a mine.
Will this drive some underground? Yes. Then let them be underground, not out in the open waving to children and posing as healthy family activity and making money as a healthy wholesome site
Apologies. Let me modify my comment: I don't think its idealism to require, by law, billion dollar companies to not spew toxic content online and psychologically poison human beings, or at least...
Apologies. Let me modify my comment:
I don't think its idealism to require, by law, billion dollar companies to not spew toxic content online and psychologically poison human beings, or at least to provide personal protective training, ongoing wellness monitoring and counselling, and long term psychological health benefits for personnel handling its clean up.
And just like toxic material mining: if you can't afford proper psychosocial and societal mental health clean up then you're not allowed to operate a cesspool media empire.
Will this drive illegsl and horrific online content underground? Yes. Then let them be underground, not out in the open waving to children and posing as healthy family activity and making money as a healthy wholesome site.
I actually agree with you when it comes to companies with ≥$1B in annual revenue. At least at a gut level, it seems that damage caused by social media giants is linear with their income—contrast...
I actually agree with you when it comes to companies with ≥$1B in annual revenue. At least at a gut level, it seems that damage caused by social media giants is linear with their income—contrast mine tailings, which are equally destructive whether the mine is a $4B operation or merely $200M.
My skepticism for internet moderation requirements stems from equal application of the law zealots who don't put revenue limits on their regulations. I was part of the 2017 Mastodon migration. One of the notable controversies from that era was an intersection of the way federation caches posts, European (or maybe just UK) laws against cartoon child porn, and a large Japanese art server. There was a load of FUD spread (possibly justified) and several webmasters closed up their small communities over fears of strict liability.
However, $1B in revenue is plenty large enough to afford moderation and legal teams to do compliance work—it's only an undue imposition for hobbyists and seed round startups (who don't have the numbers to cause massive damage in the first place).
The only remaining quibbles I have are:
The companies themselves are the ones spewing the nastiness online. They're more like negligent sewer companies than active perpetrators. It's still their responsibility to fix, but it's not their core reason d'etat.
Kids should be gently booted from the web until they're old enough to passably lie about their age. I agree that the internet should not be advertized as child-friendly, unless you are Yahooligans!.com
Very good points, 100% agree with you on kids. At the end of the day parents need to take responsibility, even if it means kids live like Luddites.. Slight counter on point one: even if they're...
Very good points, 100% agree with you on kids. At the end of the day parents need to take responsibility, even if it means kids live like Luddites..
Slight counter on point one: even if they're not producing toxic tailings themselves, they're practically inviting others to dump them on this newly minted real estate. They're running a night club, if it invites illegal activities, they need to take responsibility. They've opened a tutoring cram school, cheating is happening, that's their job toom etc.
Re: 2017 Mastodon
I've got my tinfoil hat firmly on for my next statement: I would imagine that new up and coming internet communities tend to face extremely stringent monitoring and aggressive media reporting for anything remotely suspicious because incumbents have a billion dollars or more to lose if upstarts succeed.
Unless they're locked into exploitative contracts, what prevents them from freely choosing other jobs? OpenAI isn't creating exploitative company towns or anything. It seems like they're offering...
Unless they're locked into exploitative contracts, what prevents them from freely choosing other jobs? OpenAI isn't creating exploitative company towns or anything. It seems like they're offering fair compensation relative to the local prevailing wages. Surely there's a positive knock-on effect from the jobs and a reduction in poverty.
there are many variables which may be reducing their ability to make a meaningful choice regarding employment opportunities. examples include personal life factors (desperately need to cover...
there are many variables which may be reducing their ability to make a meaningful choice regarding employment opportunities. examples include personal life factors (desperately need to cover medical treatment for family?), lived experience (young people who don't understand the implications/impacts of this work), the ethics of the company (do they provide adequate information, safety planning, and mental health services for staff or are they hiring in this community to avoid worker protections), the availability, quality and compensation of other jobs (not a lot of jobs for significant segments of the population, other viable positions pay less than living wage and involve significant risk), and on and on.
if the company is invested with lifting people out of poverty, wouldnt it make more sense to offer non-horrifying positions that teach transferable skills rather than adding trauma to their burdens?
You make good points. I think it's fair that the companies should probably communicate the risks and requirements of the jobs better. Even still, I don't have an issue with people knowingly...
You make good points. I think it's fair that the companies should probably communicate the risks and requirements of the jobs better. Even still, I don't have an issue with people knowingly consenting to challenging jobs which pay better than average. These jobs introduce competition in the marketplace and inject international wealth into the local economy. It's not perfect, but it's a start.
i think that's where we differ, i don't feel like companies which rely on public resources (including the extraction of natural resources), user made content, etc should be able to take so much...
i think that's where we differ, i don't feel like companies which rely on public resources (including the extraction of natural resources), user made content, etc should be able to take so much and not give back equally or more. i don't think meeting basic needs of their employees is a huge ask considering their profit margins. they want to benefit from our communities and sell user data but say "we're privately owned" when it comes to taking any of the burdens. i just can't find a way to justify it or see that "consent" in those situations is genuine when your choices are between violence and crime living in the slums at a safer job or get ptsd but maybe live somewhere with running water until you burn out and end up back in the slums? like does it matter if people are technically lifted out of poverty if it's not sustainable long term yk? it's like being in a much more mundane SAW movie
My friend, did you read the article? She's not filtering text descriptions, she's having to watch videos of human torture and worse things done to children, all day, because local economic...
My friend, did you read the article?
With job prospects dire in South Africa, where one in two young people is out of work, she took the plunge.
She's not filtering text descriptions, she's having to watch videos of human torture and worse things done to children, all day, because local economic situations don't allow for much better.
And the other thing of why Meta is doing this is to skirt existing labour laws:
Kenyan judge ruled that Meta could be sued in a Kenyan court, contrary to the company’s contention that the court had no jurisdiction. Though Meta was appealing against the decision, moderators regarded it as a great victory. A subsequent ruling determined that Meta was the moderators’ true employer.
It's not enough to just pay people twice going rates to do dangerous work that give them life long PTSD, without them knowing how detrimental it is to begin with. That's why we have labour laws here now, because they're written in the blood of those who were previously exploited. Mining pit workings, railroad explosive workers....they didn't sign up for death and disability, and neither did content moderators sign up for lifelong psychological harm.
Yeah, this stuff is legitmately psychologically scarring, it's not just reading bad words on the internet. Check out what and how Youtube rotates its staff out for this kind of moderation to...
Yeah, this stuff is legitmately psychologically scarring, it's not just reading bad words on the internet.
We worked in shifts, 4 days on, 4 days off, 10 hours a day. There were 3 shifts per day so you were always overlapping with everyone else on the rotation for part of your shift.
The moment your shift ended you spoke to a staff mental health advisor. While you were off, you got two calls a day, usually around lunch time and after dinner time, from a mental health advisor. These advisors were on call 24-7, and they also had their own separate set of advisors.
...Our turnover rate month over month was about 80%. It’s been about 20 years and I still have nightmares.
Or as psychologically damaging as the visceral impact of pictures and videos, especially if they’re not enculturated in American social history. A shit job, but nowhere near the bottom of the ladder.
nearly as dangerous as…
Or as psychologically damaging as the visceral impact of pictures and videos, especially if they’re not enculturated in American social history. A shit job, but nowhere near the bottom of the ladder.
Really good article. And yes this is a huge issue. It really reminds me of hiring railroad workers from overseas who have little recourse and no money for a drawn out legal battle, and then giving...
Really good article. And yes this is a huge issue.
It really reminds me of hiring railroad workers from overseas who have little recourse and no money for a drawn out legal battle, and then giving them dynamite and telling them to cut through the Rockies. "It's freely chosen work" and "they're paid twice what they would have made" arguments basically boil down to "they're lesser humans who deserve to do the dirty work we wouldn't do ourselves"
There is much more bad stuff on the internet than a lot of people do realize. Mostly on account of... well... moderation on most mainstream platforms. It isn't as if the gore you would encounter...
Sure there is bad stuff on the Internet, but most of that is reruns you could easily auto-detect.
There is much more bad stuff on the internet than a lot of people do realize. Mostly on account of... well... moderation on most mainstream platforms. It isn't as if the gore you would encounter on the early internet one or two decades ago is all gone, it already is moderated away largely.
People are also really creative at making it non obvious on the surface making you need to check it out.
There isn't enough murder happening in the world to traumatize that many moderators.
Not just murder, a lot of other stuff as well.
Are companies that bad at automatic filtering?
To some degree yeah. False positives is still a thing, if you want to protect everyone including your own moderation team you will take it for granted that a huge portion of what gets filtered are false positives. You will never know though as you are trying to protect everyone. Which is something that companies also aren't all too happy with. So you then have a few options:
Have a team of people go through the automatically filtered stuff to check for false positives.
Use less automatic filtering and depend on human moderators.
With both options you still have a team of people look at a lot of vile stuff.
And no, even with the advances made in AI good moderation of platforms still requires a lot of manual work.
Somewhat related, what is the goal of your comment? I mean, even if the numbers are inflated it still means that there are people still looking at the type of stuff that is being described here. If people still are getting traumatized over watching this sort of thing it is worthwhile to see what can be done about it.
There's an option to filter the paywall tag if you'd like. Good journalism that people value will require subscriptions to stay afloat, so I would prefer that this content still be allowed. This...
There's an option to filter the paywall tag if you'd like. Good journalism that people value will require subscriptions to stay afloat, so I would prefer that this content still be allowed. This article is also available to anyone who creates a free account.
Unfortunately a lot of really good articles out there are in paywalled publications, which is sort of unsurprising, given that professional journalists have to get paid, paywalled publications...
Unfortunately a lot of really good articles out there are in paywalled publications, which is sort of unsurprising, given that professional journalists have to get paid, paywalled publications have money to pay them and ad-supported models require lots of clicks which more high-brow journalism doesn't always attract.
You can read the full article if you have the Bypass Paywalls Clean extension installed (for Firefox or Chromium). Like uBlock Origin, I consider this an essential nowadays.
You can read the full article if you have the Bypass Paywalls Clean extension installed (for Firefox or Chromium). Like uBlock Origin, I consider this an essential nowadays.
Over here we are led to believe that AI handles soul destroying things like image identification and horrific content moderation, but maybe it's as naive as thinking Nike's sew themselves.
Distressing but important read. Be back later with additional comment but I wanted to insert this excerpt quickly to see if we can get more readers and more insight.
Generative AI also relies on exploited content moderation workers:
https://time.com/6247678/openai-chatgpt-kenya-workers/
I'll be honest, working with that kind of content sounds absolutely abhorrent, but they're doing a great service for everyone else filtering out all that toxicity. Even if it's kind of fucked, I'm okay with workers voluntarily signing up for a job that pays almost double the average hourly income in Kenya (per Hacker News). Quickly filtering out disturbing text descriptions doesn't sound nearly as dangerous as working in a lithium mining pit.
I think it's really screwed up that the job market in Kenya is such that that's considered an acceptable and already "better" option. (I live in a country with similar conditions. Content moderation laborers are commonly outsourced from here too.) I understand why it's a "better" option, but it's still not ok and it should still be improved.
Maybe this is dumb question, but how would you want content moderation to be handled? Ignoring the profit incentives with AI training data, human moderation jobs can still be grueling and thankless. Fairly compensating workers at double the local market rate who could've freely choosen other jobs seems fine to me.
Fairly compensate moderators in the same country as the user base? Americans moderate Americans, Germans moderate Germans, etc.
People might balk at the idea of content moderators being paid $200k, but I'll bet there's burnout like no other. Maybe comparable to doctors and therapists.
$2/hr seems reasonable for Kenyans moderating Kenyans, not for Kenyans to wade through the mountains of shit the rest of the world churns out courtesy of exploitation of CoL differences.
I think moderators from the same community as they're moderating may have cultural context advantages, but I really don't understand why companies shouldn't take advantage of CoL differences. I don't think that's inherently exploitative. Even within a country, it makes good sense for companies to hire remote employees or set up operations in regions with lower costs of doing business (ignoring tax breaks).
This generally has a positive effect on the region and boosts incomes and standards of living since wages rise (an actual example of a rising tide lifting all boats). If a company moved in near me and offered to double my salary, I probably would jump ship if the work wasn't too bad. Multinational companies investing in poor areas and raising average wages should be a good thing, no?
To address this for you, @ibuprofen, and @HerosJourneyMadness.
Because most CoL differences are due to availability of infrastructure and labor rights. And if the promise of 'lift out of poverty' starts happening, companies will probably just pick up shop and find the next country with the lowest wages and fewest restrictions.
I firmly think free global trade of goods and labor is only morally acceptable if wages and labor laws have a reasonable minimum standard worldwide.
I mean we can look at China as a successful example of exactly what I'm talking about as far as international investment goes. The standard of living in China has massively increased in just a couple generations due to global trade and industrialization. Partially due to economics, but mostly due to politics, many companies are now "decoupling" and moving labor to countries like India and Vietnam, and now those countries are starting to experience the exact same increase in living standards. The median Chinese citizen is significantly wealthier than two generations ago; that's the primary reason the CCP has such strong support.
There should absolutely be a minimum global tax to prevent tax havens, but I think the international exchange of goods, labor, and people is a universal good that benefits the whole world.
Which is good for China and no others. The generations of "international investment" has left the countries that invested in Chinese trade and industrialization with dearth of the skills and infrastructure to be self reliant. It effectively lifted boats in China, and will do the same in other Asian countries, while sinking the boats that did the investment.
The numbers can't always go up, there's a fixed amount of resources and time in the world; the tide doesn't rise, it comes in at one place and it goes out in another.
Look at towns where auto manufacturers moved from the USA to other countries, or even elsewhere within the USA. Many are husks of their former selves, turning into wastelands of poverty and all of the drug abuse and crime that tends to come with it.
And I'll bet those new workers don't have access to nearly the same QoL that the unionized workers of the old plants had.
There should also be a global minimum wage equivalent to 40 USD, regardless of the local average wages. The subsequent inflation will sort things out in short order.
Bet a lot less factories and call centers move around arbitrarily when that happens.
In agreement with the other reply to an extent. CoL based wages CAN be used I think acceptably and morally- but I don’t think it is very often. The liberal media (I am liberal) I think tries to adhere to some narrative that it’s ALWAYS exploitative. Truth is, I have no idea.
I’ll tell you what though- it wouldn’t be too hard to have mental health experts go in and establish some rules around how to humanely do such a job. Limit hours, rotate with other less stressful duties, etc.
I’d also bet there are things that could be done within the workflow to further limit exposure (specifically like using screenshots to sample video).
IMO this is yet another story about needing real, moral rules around employment. Also- offshoring labor itself should be a source of tax revenue for the US so the Midwest can stop losing electricity multiple times every winter.
The problem I see is that it isn't just CoL based wages, but often the lack of the surrounding supporting infrastructure and labor protection. If if the wages or good for the country people can often be made to look for longer shifts at this stuff and if it does cause mental damage it is unlikely that mental healthcare is available.
The flip side to this is that a job paying double the average wage is probably massively boosting their standard of living. The difference between making $1 and $2 per day in an extremely impoverished nation is massive. Sure, they don't have mental healthcare or the resources to maintain a first world lifestyle, but the people working these jobs are lifted out of poverty.There aren't many easy, well-paying alternative job widely available in poor countries.
This is how economies are built up and countries become wealthy.
I literally live in a country where this happens all the time and I'm telling you that it doesn't lift people out of poverty (it just makes them a little bit less poor), nor has my country become wealthy.
That's the core of my argument. People are better off and can raise their standard of living because these jobs are available. Countries benefit from international investments that raise wages and bring in money. Kenya won't become wealthy just because one multinational company offers some jobs at nearly double the local wage, but it's a start.
You're looking at it from a purely financial perspective. But being slightly better off financially at the expense of their mental health from having to regularly view traumatic content, doesn't sound to me like they're actually better off overall. I used to work in Computer Forensics, and that field has similar issues due to investigating traumatic incidents and content (like CSAM) also being a major part of the job... which can be really detrimental to your mental health, with lasting negative effects.
See: An exploration of the personal experiences of digital forensics analysts who work with child sexual abuse material on a daily basis: “you cannot unsee the darker side of life”
So I can't help but feel like it's horribly unethical to offload this sort of moderation work to a country that likely has less worker protections, and significantly less mental health support and resources available.
But that would eliminate almost all of the Kenyan jobs.
Think what you will about how the job should be done, the reality on the ground is that this is a very well paying job that Kenyans are choosing because it's preferable to their other opportunities. Kenyans are better off if Meta is pressured to provide better employee support than they are if those jobs disappear.
It's like asking how industries should handle toxic waste: you begin by making plans to limit the production of it as much as possible. You don't turn the machines on live and then try to find the most vulnerable humans to wade through it as an afterthought.
Humanity doesnt need Facebook: we would lose very little of value if it disappeared overnight. it's up to them to figure out how to limit harm before they do harm.
Maybe it's text only.
Maybe images and videos must be tied to a legally traceable account: if you upload illegal horror you're going to jail. Currently the penalty is some poor human will have to watch it, and then report it to some other poor human to moderate, and the uploader has in the mean time made 50 more accounts.
This idealism ignored a fundamental reality of the internet: those who can do, have already did. Lots of the harms aren’t easily foreseeable: those warning of them before they happen will again be dismissed as cranks in the future.
I don't think its idealism to require, by law, billion dollar companies to not spew toxic waste into our planet and poison human beings, or at least to provide personal protective equipment and long term health benefits for personnel handling its clean up.
And just like toxic material mining: if you can't afford proper environmental clean up then you're not allowed to operate a mine.
Will this drive some underground? Yes. Then let them be underground, not out in the open waving to children and posing as healthy family activity and making money as a healthy wholesome site
I'm confused. We were discussing social media sites, not mining waste.
Apologies. Let me modify my comment:
I don't think its idealism to require, by law, billion dollar companies to not spew toxic content online and psychologically poison human beings, or at least to provide personal protective training, ongoing wellness monitoring and counselling, and long term psychological health benefits for personnel handling its clean up.
And just like toxic material mining: if you can't afford proper psychosocial and societal mental health clean up then you're not allowed to operate a cesspool media empire.
Will this drive illegsl and horrific online content underground? Yes. Then let them be underground, not out in the open waving to children and posing as healthy family activity and making money as a healthy wholesome site.
I actually agree with you when it comes to companies with ≥$1B in annual revenue. At least at a gut level, it seems that damage caused by social media giants is linear with their income—contrast mine tailings, which are equally destructive whether the mine is a $4B operation or merely $200M.
My skepticism for internet moderation requirements stems from equal application of the law zealots who don't put revenue limits on their regulations. I was part of the 2017 Mastodon migration. One of the notable controversies from that era was an intersection of the way federation caches posts, European (or maybe just UK) laws against cartoon child porn, and a large Japanese art server. There was a load of FUD spread (possibly justified) and several webmasters closed up their small communities over fears of strict liability.
However, $1B in revenue is plenty large enough to afford moderation and legal teams to do compliance work—it's only an undue imposition for hobbyists and seed round startups (who don't have the numbers to cause massive damage in the first place).
The only remaining quibbles I have are:
Very good points, 100% agree with you on kids. At the end of the day parents need to take responsibility, even if it means kids live like Luddites..
Slight counter on point one: even if they're not producing toxic tailings themselves, they're practically inviting others to dump them on this newly minted real estate. They're running a night club, if it invites illegal activities, they need to take responsibility. They've opened a tutoring cram school, cheating is happening, that's their job toom etc.
Re: 2017 Mastodon
I've got my tinfoil hat firmly on for my next statement: I would imagine that new up and coming internet communities tend to face extremely stringent monitoring and aggressive media reporting for anything remotely suspicious because incumbents have a billion dollars or more to lose if upstarts succeed.
It's not fair compensation and they can't just "freely choose other jobs"
Unless they're locked into exploitative contracts, what prevents them from freely choosing other jobs? OpenAI isn't creating exploitative company towns or anything. It seems like they're offering fair compensation relative to the local prevailing wages. Surely there's a positive knock-on effect from the jobs and a reduction in poverty.
Edit: added a note about company towns.
there are many variables which may be reducing their ability to make a meaningful choice regarding employment opportunities. examples include personal life factors (desperately need to cover medical treatment for family?), lived experience (young people who don't understand the implications/impacts of this work), the ethics of the company (do they provide adequate information, safety planning, and mental health services for staff or are they hiring in this community to avoid worker protections), the availability, quality and compensation of other jobs (not a lot of jobs for significant segments of the population, other viable positions pay less than living wage and involve significant risk), and on and on.
if the company is invested with lifting people out of poverty, wouldnt it make more sense to offer non-horrifying positions that teach transferable skills rather than adding trauma to their burdens?
You make good points. I think it's fair that the companies should probably communicate the risks and requirements of the jobs better. Even still, I don't have an issue with people knowingly consenting to challenging jobs which pay better than average. These jobs introduce competition in the marketplace and inject international wealth into the local economy. It's not perfect, but it's a start.
i think that's where we differ, i don't feel like companies which rely on public resources (including the extraction of natural resources), user made content, etc should be able to take so much and not give back equally or more. i don't think meeting basic needs of their employees is a huge ask considering their profit margins. they want to benefit from our communities and sell user data but say "we're privately owned" when it comes to taking any of the burdens. i just can't find a way to justify it or see that "consent" in those situations is genuine when your choices are between violence and crime living in the slums at a safer job or get ptsd but maybe live somewhere with running water until you burn out and end up back in the slums? like does it matter if people are technically lifted out of poverty if it's not sustainable long term yk? it's like being in a much more mundane SAW movie
My friend, did you read the article?
She's not filtering text descriptions, she's having to watch videos of human torture and worse things done to children, all day, because local economic situations don't allow for much better.
And the other thing of why Meta is doing this is to skirt existing labour laws:
It's not enough to just pay people twice going rates to do dangerous work that give them life long PTSD, without them knowing how detrimental it is to begin with. That's why we have labour laws here now, because they're written in the blood of those who were previously exploited. Mining pit workings, railroad explosive workers....they didn't sign up for death and disability, and neither did content moderators sign up for lifelong psychological harm.
Yeah, this stuff is legitmately psychologically scarring, it's not just reading bad words on the internet.
Check out what and how Youtube rotates its staff out for this kind of moderation to understand how dangerous excessive exposure to this can be (I can't find an accessible official link with details, but this reddit comment reflects what I've heard secondhand): https://old.reddit.com/r/technology/comments/ixbzn7/excontent_moderator_sues_youtube_claims_job_led/g66k0h8/
I doubt Kenya and Korea get as much consideration
Or as psychologically damaging as the visceral impact of pictures and videos, especially if they’re not enculturated in American social history. A shit job, but nowhere near the bottom of the ladder.
Really good article. And yes this is a huge issue.
It really reminds me of hiring railroad workers from overseas who have little recourse and no money for a drawn out legal battle, and then giving them dynamite and telling them to cut through the Rockies. "It's freely chosen work" and "they're paid twice what they would have made" arguments basically boil down to "they're lesser humans who deserve to do the dirty work we wouldn't do ourselves"
There is much more bad stuff on the internet than a lot of people do realize. Mostly on account of... well... moderation on most mainstream platforms. It isn't as if the gore you would encounter on the early internet one or two decades ago is all gone, it already is moderated away largely.
People are also really creative at making it non obvious on the surface making you need to check it out.
Not just murder, a lot of other stuff as well.
To some degree yeah. False positives is still a thing, if you want to protect everyone including your own moderation team you will take it for granted that a huge portion of what gets filtered are false positives. You will never know though as you are trying to protect everyone. Which is something that companies also aren't all too happy with. So you then have a few options:
With both options you still have a team of people look at a lot of vile stuff.
And no, even with the advances made in AI good moderation of platforms still requires a lot of manual work.
Somewhat related, what is the goal of your comment? I mean, even if the numbers are inflated it still means that there are people still looking at the type of stuff that is being described here. If people still are getting traumatized over watching this sort of thing it is worthwhile to see what can be done about it.
Archive link https://archive.is/AT6JK
I think most of the FT's revenue comes from institutional (i.e. corporate) subscriptions, so they're not too fussed.
I wish we'd just ban all paywalled articles.
There's an option to filter the paywall tag if you'd like. Good journalism that people value will require subscriptions to stay afloat, so I would prefer that this content still be allowed. This article is also available to anyone who creates a free account.
I wish we'd ban discussion on paywalls in paywalled articles. It's so incredibly offtopic and tiring.
Unfortunately a lot of really good articles out there are in paywalled publications, which is sort of unsurprising, given that professional journalists have to get paid, paywalled publications have money to pay them and ad-supported models require lots of clicks which more high-brow journalism doesn't always attract.
We'd be missing out on excellent journalism if we did.
Journalism that is paid for with advertising only is not journalism, it’s infotainment.
Either that or those submitting paywalled articles should have the sense to directly link to the archived page in the first place.
This is an archive link.
You can read the full article if you have the Bypass Paywalls Clean extension installed (for Firefox or Chromium). Like uBlock Origin, I consider this an essential nowadays.