Re:GPT trainer says he's traumatized from the RLHF work
by klyrs on Friday May 19, 2023 @01:41PM
(#36004711)
Agreed, it doesn't seem like it should be that hard to find employment records and get back pay into the hands of the raters. According to several friends of mine who have done charity work in various African nations, the hardest part is that poverty is so stark and corruption is so entrenched that you can't outsource benevolence. If you want to get money into the hands of a specific individual, you need to go there yourself, do the legwork, and pay them directly. If you pay somebody else to do it, the money evaporates.
Re:GPT trainer says he's traumatized from the RLHF work
by jstarfish on Friday May 19, 2023 @08:19PM
(#36008233)
> But, according to several friends of mine who have done charity work in various African nations, the hardest part is that poverty is so stark and corruption is so entrenched that you can't outsource benevolence. If you want to get money into the hands of a specific individual, you need to go there yourself, do the legwork, and pay them directly. If you pay somebody else to do it, the money evaporates.
This isn't an Africa thing, it's typical third-world grift. What they're saying holds true anywhere, even with immigrant communities in America.
You offer X Solutions $100 to do a job (what a steal!). X pockets $60 and solicits Y Recruitment for $40. Y pockets $30 and hires 10x Z Workers for $1 each. It's middlemen all the way down.
We got GPT 3.5 in part by training models on RLHF data, not just to answer questions, but to rate questions like the humans were.
And in turn we've seen those models used to fine tune models like LLAMA without requiring anyone to sift through that.
Still this is a horrible injustice, OpenAI is going to take on unimaginable wealth and even if that wasn't guaranteed when these people were being paid $1 an hour, it's not too late to try and fix some of what they did.