I find the IAT somewhat ridiculous. For example, it asked whether I associate black with "sports". Well, I do, but there is nothing "implicit" or "biased" about it. It is my lived experience as a high-school sprinter.
In addition, stereotype accuracy is one of the largest and most replicable effects in all of social psychology. It's not "wrong". What would be wrong is not adjusting for the individual once you get to know them, but AFAIK that same research also shows that most people drop the stereotypes quickly once they actually get to know an individual.
And yes, there are people who don't. In other news, stupid people exist and earth still round.
But now that they are generally "equal", it's detrimental
to them. I.e. It's entirely plausible that society is self-
correcting against the unfairly-claimed bias by...being
I think "overcorrection" is a valid concern (if people begin devaluing women's opinions thinking they're diversity hires or somehow hired at a lower bar), but I for one haven't observed us being there yet. There's still this yawning divide between being a man vs a woman in both society and in the smaller sphere of tech. There are many things we take for granted as men: e.g people don't assume I, a man, work in marketing despite sitting with other engineers -- these sorts of things negatively affect women who are otherwise fully / more than qualified to do their jobs.
Anecdotally, a number of women in my life who worked in tech have since left the industry citing aggressions of varying levels. This is concerning :(
While I absolutely wish that any person felt welcome in tech regardless of race, gender or other choices like orientation, I'm not convinced this is a good approach. It's an explicit statement that they are desperate for diversity, and as a result it calls in to question the justification for those hires. How much did the candidates diversity factor in vs. their competence?
They claim that non-minorities won't be disadvantaged by them and I believe they are being honest here, but it still sends a very mixed message.