Emil O W Kirkegaard

But who is going to call out OUR OWN FIELD for trumping up this sort of nonsense? And by "nonsense" I mean "systematically ignoring a great deal of evidence that contests your preferred narrative."

Stereo acc work is no more immune to criticism than any other area. But there is a world of difference between critically grappling w/work contesting your preferred narrative, & simply ignoring. Ignoring large bodies of lit that contest your conclusions should not be an option.

Nothing in this thread addresses policy or ethics. Empirical evid. of accuracy in any stereotype says EXACTLY ZERO about whether or how to go about diversity and inclusion, affirmative action, or ANYTHING ELSE. It is just the damn data.

An accurate stereotype does not justify discrimination, harassment, or mistreatment of any type, just as knowing that NJ is warmer than Alaska (in general) does not mean it would be a good idea for me to walk outside shirtless in shorts today.

However, the rather high accuracy of them does force a purely rational agent to take those stereotypes into account, even when dealing with individuals https://emilkirkegaard.dk/en/?p=5788 . So it is a little bit thornier than being "just data".

Agreed. It's not that easy. Ignoring diagnostic information?

Esp. since it has actual consequences. E.g. doing the "right" thing and ignoring the information will make you worse at predicting people's behaviors and at correctly understanding the world. It doesn't feel right to me to demand people act in ways that are harmful to them.

This is also more complex than it seems, prob too much for twitter. Short version: 1. For important things, it behooves decision makers that to max out diagnostic individuating info; 2. Evid of strong stereo acc ≠you can/should assume your stereos are always right.