Sunday, June 26, 2016

Misandry bias

The article is Artificial Intelligence Has a ‘Sea of Dudes’ Problem by Jack Clark.

It is, to me, an example of the prejudicial bias that the chattering class has against seeing their own prejudicial biases. It is an empirical fact that while it has become ever easier for women to enter STEM (recent studies indicate a bias of 2 to 1 for the female candidate when presented with two equally qualified candidates), there are fewer and fewer women choosing to go into STEM and into the computer sciences in particular and into artificial intelligence in the very particular.

If women, as citizens with agency, are choosing this course, is there a problem? Not on the face of it. There is much research on the value of diversity which produces a scattershot of results which pretty much come down to what you would expect. In general there is little inherent value in diversity (racial, gender, etc.) except in very particular circumstances. Diversity as a goal in and of itself has no real return. Diversity of experience and accomplishment is valuable in particular contexts.

The article posits that there is a problem arising from the fact that the field is numerically dominated by men. I read the article expecting that there would be some examples of why AI is negatively affected by the gender imbalance.

There is not one such example. The premise of the article is that there is a problem with male gender dominance but they are unable to muster a single example of a problem in AI created by such an imbalance.

All the article boils down to is that the reporter does not like that so many men choose to go into AI and so few women do.

There is no problem that needs fixing because, based on evidence advanced in the article, there is no problem in the first place.

Different people make different choices and the author doesn't like those choices.

Not much there, there.

UPDATE: Nursing, Education, Librarians, Sociology, Psychology are all fields overwhelmingly dominated by women. Other than education, there is, as far as I am aware, not much research indicating that the gender skew has any negative consequence. To capture the blindness to their own prejudice, recast the original headline to get a flavor of just how prejudiced and inappropriate the headline is: "Sociology Has a ‘Sea of Dames’ Problem", "Nursing Has a ‘Sea of Dames’ Problem", "Teaching Has a ‘Sea of Dames’ Problem", etc.

No comments:

Post a Comment