What Atlanta Can Teach Tech About Cultivating Black Talent | Hacker News Reader

What Atlanta Can Teach Tech About Cultivating Black Talent

wired.com

17 pointsrak6 comments

Nice! It is also worth mentioning that in addition to the work at Clark Atlanta University, Spelman, and Morehouse, there is Black leadership in computing at Georgia Tech. Charles Isbell is the Dean of the College of Computing and Ayanna Howard does robotics research there. It helps that many of the "100 black " people in machine learning referred to in the article reside in Atlanta.

What can Atlanta teach tech?

Not much, reading the article.

The fundamental reason that their are fewer Black people working in computing as that they score at the very bottom for Math in the SAT:

https://i1.wp.com/www.brookings.edu/wp-content/uploads/2017/...

And the explanation for that is probably deep in history and DNA: Africa as a continent has forever been far less developed than any other region, yielding no overall selective pressure for any of the traits which today produce high Math scores.

Or, perhaps, much less magical non-genetic factors like more likely being brought up with less access to higher-quality education, nutrition, stability, and safety.

Here’s an article from the institution whose graph you reference about why this gap exists - https://www.brookings.edu/articles/unequal-opportunity-race-... - which explains that many of the differences in education available. Black children are more likely to go to schools that have less funding, offer fewer advanced classes, have more students in them, and have lower-qualified teachers. And experiments that involved children going to better schools showed they had better outcomes.

So, yeah, let’s control for things we actually know affects these outcomes first before jumping to genetic factors.

To downvote him or to say genetics may not be part of it is just being ignorant. Men an woman have completely different genetics, while Black vs Asian do as while. We are not all equal in terms of genetics, so don’t discount that argument. Having lived in Atlanta and having hired devs in Atlanta, I can easily comment on this since I have first hand knowledge. Unlike other parts of the country, Atlanta has a large black population (18% of the US is black, and mainly concentrated in south eastern US). In my last hiring, 80% of interviews were with black developers. Due to the demographic, the talent in that region will likely be black. If you go to Texas, the talent may likely be white. I’m really tired of this PC stuff with black vs white. The black population as a whole is very low outside the southern US. Companies will cultivate the best they can afford. While there are racists out there, the very large majority of hiring managers could care less of your skin color and care more about your abilities. As far as schools and funding, this has to do with regions and parents need to step up. We live in a capitalist society. There are no hand outs. Be the best or get run over by someone better than you. I grew up extremely poor in a broken home, with no college education. I worked hard and built everything I have on my own through sweat and tears.

It’s easy to take a statistical summary of two groups, and discover that attribute X has some distribution for one group, and another for the other. But figuring out why requires much more effort.

In order to say something is genetic, you need to rule out confounding factors that apply. For example, twin studies attempt to reduce the effects of upbringing (cultural influences, home stability, access to education and other opportunity). This is effective in things like trying to determine whether there may be a genetic component to homosexuality - for example, there’s a statistically significant likelihood (that doesn’t exist for non-twin siblings) that if one twin is gay, the other one is.

However, you can’t aggregate twin studies of one group (“white people”) and another group (“black people”), since you’re not getting around the underlying confounding factors. So, while such twin studies do find that genetics play a role in academic performance (since if one twin does well, the other is statistically significantly more likely to do well too), it isn’t helpful as a mechanism here.

Anyway, jumping to genetics for things that have clearly more local causes is at best lazy. I don’t doubt there’s some genetic factors out there, but we don’t have an easy way to untangle them from other factors.

But we do have several different studies that show a connection between family income and academic performance (even at the same school), parental education levels and academic performance (ie, generational effects), class sizes (in early education) and academic performance, teacher quality, school funding, and so forth.

Reply on news.ycombinator.com