I am guessing a major factor is that a lot of industries, women often have to prove themselves to be as skilled as or more skilled than their male counterparts due to entrenched patriarchal structures. If you regularly have to prove you know how to code, you're probably a lot less likely to rely on ChatGPT to do it for you.
I am having difficulty following this line of reasoning, can you please help clarify? Why would being forced to prove your worth dissuade you from using a productivity tool? Are you implying women likely don't have access to use it at all, or they don't trust the output because the stakes are too high?
In some ways I bet it's also just the whole "made for and by men" thing. That might be because of the AI companies being male-dominated (the tech industry is like that right now at least) but it's also maybe just the training data. Probably most of the training material used for these AIs come from men, simply because there is more material online made by men (again, probably because the tech industry is male-dominated and lots of other cultural reasons).
It's the same with video games. Most video games today are made by men and for men and it's kind of a positive feedback loop since the people playing those games will be men, and those people will be more likely to get into game development. And when making a game, you go for the largest market so you go for the existing male market instead of trying to cater to women.
Shout-out to software engineers having the lowest use outside of work. I have a coding chatbot that work wants me to use. Even when I have it set up right with only the correct tabs open in my IDE it just hallucinates stuff that looks ok but doesn't actually work.
I mostly just ask it if it poops. One time I got it to admit that its garbage collection routine could be roughly compared to the act of pooping and it was the best day ever.
Statistics is such a weird area of math that idk if anything social can be deduced based on them.
What does it tell you exactly about the world that in a sample 550 men used ChatGPT vs 450 women?
There’s a reason for replication crisis in psychology.
Big ifs and unbiased in what way? There are detectable and undetectable biases. Physical and mental. It’s god damn pseudoscience outright because there’s no way think of every possible way it can be fricked due to human complexity undetectable by simple questionnaires
The only proper way to approach psychology and psychiatry is to analyse the brain neurons one by one and map it all for each one person. It’s just not possible yet. But it is deterministic, sane than blindly shooting substances and seeing what sticks. Those are primitive early methods akin to bloodletting due to absolute lack of deeper understanding of the brain