News for hackers. Stuff that matters.
Re:Venture capital firms to include “diversity riders” in te... by eaurouge on Friday August 28, 2020 @08:02PM (#24311865)
We all carry biases, including towards/against groups we wouldn't expect to be biased towards/against. We probably all start with blank slates, but, unless you've had zero interactions with the outside world, you'd have to work actively to retain that blank-slate outlook.

In aggregate, we run the risk of collectively creating various forms of systemic bias. We can test for these systemic biases (we have) and prove the existence/non-existence of said biases (we have).

Re:Venture capital firms to include “diversity riders” in te... by raxxorrax on Friday August 28, 2020 @11:37PM (#24313226)
> prove the existence/non-existence of said biases

No, you have not.

Re:Venture capital firms to include “diversity riders” in te... by throwawaysea on Friday August 28, 2020 @08:13PM (#24311952)
> We can test for these systemic biases (we have)

How do you propose testing this? Typically, a difference in outcomes is taken as evidence of systemic bias, but it isn't the same thing. For example, if one cohort commits more crimes, they are going to have more police interactions, more arrests, and so forth. But that's not a systemic bias. Unfortunately the recent "anti-racist" authors like Ibram Kendi have been pushing a false and expansive definition, where they term any difference in outcomes (or as they put it, any racial inequity), as "racist". That's not what most people would call "racist".

> prove the existence/non-existence of said biases (we have)

As for the existence of biases, I assume you are referring to the 'implicit association test', which is commonly used as a tool to identify unconscious biases. The IAP methodology has been challenged and debunked many times. Studies have also found that measures of implicit bias do not predict how people act in real life. See https://newdiscourses.com/2020/06/flaws-white-fragility-theo... for more.

Re:Venture capital firms to include “diversity riders” in te... by jtsiskin on Friday August 28, 2020 @11:33PM (#24313186)
There are many ways to conduct studies on bias. As a simple example, they can do things like send out identical resumes with different genders/races and see who gets called back. You can find a ton of these; here is a meta-analysis which looks at many such studies: https://www.pnas.org/content/early/2017/09/11/1706255114
Re:Venture capital firms to include “diversity riders” in te... by skitout on Saturday August 29, 2020 @04:51AM (#24314731)
Example of bias from some VC against woman (was done is Sweden, know for its gender equality): https://hbr.org/2017/05/we-recorded-vcs-conversations-and-an...