Hacker News Reader

Author here - not an original insight, although it's clichéd enough that I can't point you to where I picked it up from.

I also want to emphasize that I do not view bias as a bad thing in the context of supervised models. In some ways, I think it's the whole point of a supervised model (to inherit the judgment of its creators). If the bias helps filter predictions that are useful for your goals, it's a good thing.

We can walk around bias through self supervision with sampling techniques to select training pairs

Bias is learned behavior, so it seems that "learning models" is precisely the right name for it despite whether we considered the ramifications of learning

You can call them learning models, the same way our kids "learn" in environments that are also biased.

It's why they are called models and not just "how it works" or absolute truth.

I know several industries filled with questionable researchers who might disagree with you. For clarity, I agree with you for most all cases.