We landed on "gender" as the concept going towards the behavioral support structures and "sex" relating instead to the wider topic of differentiations around gamete support in general.
No we fucking did not.
YOU have pushed that notion, and you've done it based largely on your own untested and untestable hypotheses and baseless beliefs.
Society does. Science does.
No it doesn't. The things that YOU read and accept might say so, but your echo chamber does not represent the entirety of society, nor does it even remotely represent scientific accuracy.
But hey, let's circle back to what you've put forth. You shared a case in which an AI program was given a bunch of fMRI scans, and then was ASKED to find the differences between males and females. It was TRAINED on a set of data that included tens of thousands of elements, and it had a SPECIFICALLY IDENTIFIED OBJECTIVE. It came up with an answer for WHAT IT WAS DIRECTED TO DO.
Now, let's also acknowledge a few major elements that were NOT CONTROLLED FOR in this project: the direct effect of hormones and neural plasticity. There's no way to tell whether the criteria that the AI honed in on were picking up hormones, nor if it was picking up learned behavior as a result of conditioning.
Alright. You're supposed to understand something about data science, about predictive modeling, and about machine learning. So let me pose you a question.
If this AI was given a large set of fMRI scans, which were all labeled "black" and "white", and was given the directive to find the differences between black brains and white brains across tens of thousands of elements...
- Do you think it would be able to come up with a solution that distinguishes them?
- Do you think that the distinguishing criteria would be meaningful?
- Do you think that the distinguishing criteria, and the repeatability of them would support developing policies that use those criteria to identify people as black or white, regardless of their actual melanin content and experiences?
You found a study that gives you the answer that you want to have. And because it confirms your pre-existing belief, you have failed to consider the shortcomings, faults, and applicability of the study. And you've ignored thousands and thousands of other studies that have repeatedly demonstrated that there are no meaningful differences in the structures of the human brain on the basis of sex - there are average differences, but they are not material to behavior, cognitive ability, or anything else that makes a difference in the real world. Furthermore, they've repeatedly found low levels of predictability - and they've acknowledged that they cannot exclude the influence of conditioning and the way that affects our very plastic brains.
You want to change my mind? Here's all you have to do: Go take that AI, and feed it a large set of fMRI scans of children under the age of three, and explicitly give the AI the option of finding no material difference.
If it still manages to find a reliably predictable set of differences, then I will accept that it has validity.
Until that time, however, I put this in the same category as I put all of the great many studies done that attempt to "prove" that black brains are just different from white brains.