> the definition of “the algorithm” is the training set

Or the trained model itself. There are people looking for intentional bias. But the insidiousness of the problem likely arises from unintentional bias. Letting researchers brute force the ranking models with hypotheticals could be a win win.

Is there any reason to not think that like most human-created systems it has a mixture of intentional and unintentional bias?

Twitter itself is probably the training set for twitter.

In that case, the algorithm is the labels.