Good point, it's similar to some extent. Although clearly the quality of the work that the people doing RLHF on the major LLMs is rather low in comparison with those volunteering at Wikipedia.
There were no "good" volunteers qualifier used though. Obviously, some RLHF "volunteers" are better than others just like some used by Wiki are better than others. I wonder if there's edit battles between RLHF like we've seen on Wiki?
Good point, it's similar to some extent. Although clearly the quality of the work that the people doing RLHF on the major LLMs is rather low in comparison with those volunteering at Wikipedia.
There were no "good" volunteers qualifier used though. Obviously, some RLHF "volunteers" are better than others just like some used by Wiki are better than others. I wonder if there's edit battles between RLHF like we've seen on Wiki?