Affirmative action doesn't work, except for the short-term career prospects of those individuals lucky enough to profit from it, and the virtue-signalers who propose it.
GitHub is not at the "bottom of the funnel". They're in the middle: people who work there today may go on to wildly successful careers elsewhere. And people who have bad experiences there may not.
So please don't make the pipeline argument. Just because other earlier stages have failed to be welcoming to minorities doesn't mean GitHub shouldn't try, and just because GitHub is a corporate employer doesn't mean their employees aren't still living developing human beings.
What evidence is there to believe this?
Because you think wrong.
>Given the starting point of "all these groups of people have the same capacity for talent and skill"
Why take something we know to be false as your starting point though?
For example, 117 people have run the 100m in less than 10 seconds. Do you know how many of them are women? Of course you do. Almost everyone is willing to admit this is a genetic difference. Do you know how many of those men are black? 113. Most people are also willing to admit this is genetic, especially when you point out that we've identified specific genes giving west Africans an advantage in sprinting. But suddenly when the same observations are made about IQ, then it becomes "racist pseudo-science", despite a complete inability to offer any valid criticism of the science involved.
For example, people with "non-white sounding" names don't get hired at the same rate as those with "white sounding" names when they have the same skillset. http://www.nber.org/digest/sep03/w9873.html
I think these biases operate at a low level, below most of our understanding. I like that GH is having this conversation in the open and working to rid themselves of these biases as much as possible.
To me, this is all relevant socioeconomic data. Why not look at the distribution of gender, race, sexuality, religion, etc, in your workplace? If there's a discrepancy, why not figure out what the root cause is? Given the starting point of "all these groups of people have the same capacity for talent and skill", why is the distribution skewed? What does it hurt to ask?