
"the seemingly magical abilities of AI models are built on the back of low-cost labor." - thank you, that's a point that always gets overlooked.
OpenAI reportedly hired workers in Kenya – screening tens of thousands of text samples for sexist, racist, violent and pornographic content – to help make its ChatGPT model less toxic. Released last November, ChatGPT has taken the internet by storm. Its ability to generate text automatically given an input prompt has led to …
If wealthy foreigners are to purchase influence through economic activities,. Let them pay full price. Otherwise soon becomes sacrificial lapdog to apathetic foreign interests.
Leaving everything broken when those wealthy foreigners pull out, to go to a more economically ravaged, cheaper labor area.
Maybe there should be laws that ensure e.g. US company pays US wages regardless where they hire their workers.
Not a bad idea but what about goods that everyone buys that are made in low wage countries? Such as furniture, appliances and cars? Think of all the stuff made in China that were outsourced by major corporations to "save money". Most people would not be able to afford them.
Here in the States food would be become very expensive if the farmers had to pay a livable wage to the seasonal workers.
I'm not saying that the paying a livable wage is bad but there would be consequences.
It wouldn't work. Instead of directly hiring people in a different country, the company in whatever developed country would contract with a subsidiary if allowed, local company if not, to provide a service for a certain amount, and that local company would pay the people who provided it. If this was audited very much, the employing company could provide services like obfuscating how many people did the work and how long it took, because they're providing a service, not the labor. Contracts that pay for tasks completed instead of hours worked are common.
I don't think there's anything Hell has to offer that's as scary as what can be dredged up from the twisted minds of we mere mortals. There's some very dark corners in the minds of some of humanity's worst specimens, the most despicably depraved thoughts, most remain as thoughts, some get written down and the worst are acted out.
For all the amazing things we can do and all the amazing things we will do, the animal is still just an inch below the civilized veneer and when combined without our intellect there's no limit to the depravity we can imagine.
"To compare it to US wages is a false equivalence that mischaracterizes the situation. A comparative Western wage would be between $30 and $45 per hour."
Maybe. If OpenAI were based in Kenya, and all its management and employees were making Kenyan "living wages", then there might truth there. But it isn't, so OpenAI is basically paying a pittance for services that will potentially make them filthy stinking rich, even at American wage scales.
I have lived in both kenya and America. $200 - $300 in Kenya would buy you a lifestyle that would resemble that of perhaps the poorest trailer park in America. If you have many children you might not be able to afford to send them all to school. While most Kenyans would be grateful for a job at all, to compare it to the lifestyle of $45 an hour wage earner is just the creative imagination of the PR department. You would need at least $3 000 or $4 000 a month to emulate such a lifestyle here.
That's the whole problem - it *isn't* intelligent. It's just pattern-seeking, and if you train it on bad patterns it finds them.
[very quick overview - these models start with noise and then amplify the bits that it recognises from the training data. This is repeated until there's very little noise left and the whole thing looks like it 'belongs' with the original training data]
You'd probably need a second model just trained to identify toxic content, but you're still back to "at some point, human judgement on toxicity is needed, and that means people looking at the worst horrors of the internet".