Not surprising
Surprising gender biases in GPT
111–120 of 148 posts
Re: Surprising gender biases in GPT
#112I will preface this by saying I think people should do whatever it is they want to do, physical limitations willing. My (possibly contentious) question: Is this an OK thing to do? Why are we allowing a vast minority of people to rewrite our social norms? I went to the Barbie film as a man nearing middle-aged and had the most young girls, mothers, teenage girls, and women in general I've ever seen in an audience at th…
There's a difference between women or GNC folks choosing for themselves and cis men choosing for them. That's where gender bias is poisonous.
I think there is a lot of evidence that using terms like "cis men" is the more poisonous attitude in the current day.
GNC folks are not the norm. The clue is in the name. It's all well and good to let people choose things for themselves but young people need guidance, not a divisive and inaccurate model of the world pushed on them.
Re: Surprising gender biases in GPT
#113Earlier quoted context omitted.
There's a difference between women or GNC folks choosing for themselves and cis men choosing for them. That's where gender bias is poisonous.
I think there is a lot of evidence that using terms like "cis men" is the more poisonous attitude in the current day. GNC folks are not the norm. The clue is in the name. It's all well and good to let people choose things for themselves but young people need guidance, not a divisive and inaccurate model of the world pushed on them.
I’d like to see that evidence.
Which “divisive and inaccurate model” are you talking about exactly?
Re: Surprising gender biases in GPT
#114One interesting thing to do is use a model directly like Llama and then query the next-token probability logits for "he" and "she" (assuming you set up the sentence in such a way). For example: "A doctor was examining the patient when ___" What this makes apparent is that increasing model temperature will select the less stereotypical option more often. IMO this is getting at a deeper truth that the use of a gender i…
I agree, please don’t randomly select a gender. The singular “they” is respectful to everyone and has the same benefits that you pointed out
Re: Surprising gender biases in GPT
#115> This phenomenon likely reflects that while initiatives to integrate women in traditionally masculine roles have gained momentum, the reverse movement remains relatively under developed. Not really surprising IMHO. Yes, there has been a huge focus on accepting women in the workplace and no, there has not really been a push accepting men as primary caregivers in the family. This likely uncovers underlying motives for…
It also fits into the larger societal trend of horribly underserving Zoomer and Alpha (maybe young Millennials too) young boys and men. I've heard someone put it very succinctly: people like Andrew Tate are asking the right questions, but giving abominable answers. I can drop some very simple statistics: women have been getting more bachelors degrees since the 80s . Most university gender ratios are 40/60 trending to…
If you accept a couple of other facts its even worse.
Presuming you accept iq or related tests of cognitive ability as useful, (even if not capturing the entirety of human qualities) and you accept that males have higher variance in cognitive abilities, more around the extremes of low and high cognitive abilities, whereas women on average fall closer to the mean, then if higher education was based on ability you would expect to see more men in higher education.
This is if course, verboten speak. I think as a society we have a lot to unpack.
Re: Surprising gender biases in GPT
#116Earlier quoted context omitted.
I think there is a lot of evidence that using terms like "cis men" is the more poisonous attitude in the current day. GNC folks are not the norm. The clue is in the name. It's all well and good to let people choose things for themselves but young people need guidance, not a divisive and inaccurate model of the world pushed on them.
I’d like to see that evidence. Which “divisive and inaccurate model” are you talking about exactly?
Specifically I am talking about the inaccurate model that there is no sex differences and the idea of male and female genders are socially constructed.
Re: Surprising gender biases in GPT
#117Earlier quoted context omitted.
I’d like to see that evidence. Which “divisive and inaccurate model” are you talking about exactly?
Specifically I am talking about the inaccurate model that there is no sex differences and the idea of male and female genders are socially constructed.
This is somewhat begging the question. It assumes that:
A) this is a unified model for how queer people talk about gender.
B) the term "cisgender" does anything at all to advance it. And
C) that it causes harm.
without providing any evidence for any of those claims.
----
It also assumes:
D) that the multiple ideas being bundled up here are entirely incorrect, or that they are some kind of new idea.
Male and female genders are socially constructed, at the very least in presentation -- and that's not a new idea, it's an entirely uncontroversial understanding of gender that's been around for ages. Just as one example, pink used to be a manly color. Social stereotypes of women as innocent or highly sexual and wild also vary between cultures and time periods.
This has been a theory of gender for a long time, it's not a new idea. And it has nothing to do with sexual biological differences beyond expressing the (again, entirely uncontroversial and generally accepted) idea that not all social customs and attitudes are 100% biologically based.
----
And you might have criticism of that model, just like you might have criticisms of gender essentialism. There is not uniformity on how to view gender even within queer communities. There is no singular model to debunk. Everyone (queer or not) has their own model of gender.
And that's the final problem to bring up, that you're also assuming:
E) that the statement "I'm not cisgender, I'm normal" doesn't have harmful effects, or that it doesn't push people towards a competing model of gender with its own flaws and inaccuracies that could be interrogated in the exact same way as other gender/sex models.
Re: Surprising gender biases in GPT
#118Earlier quoted context omitted.
It also fits into the larger societal trend of horribly underserving Zoomer and Alpha (maybe young Millennials too) young boys and men. I've heard someone put it very succinctly: people like Andrew Tate are asking the right questions, but giving abominable answers. I can drop some very simple statistics: women have been getting more bachelors degrees since the 80s . Most university gender ratios are 40/60 trending to…
If you accept a couple of other facts its even worse. Presuming you accept iq or related tests of cognitive ability as useful, (even if not capturing the entirety of human qualities) and you accept that males have higher variance in cognitive abilities, more around the extremes of low and high cognitive abilities, whereas women on average fall closer to the mean, then if higher education was based on ability you woul…
If you accept those priors, men would be over-represented when attending college was rare and would become less dominant (and eventually a minority) when attending college became vastly more common.
Re: Surprising gender biases in GPT
#119Earlier quoted context omitted.
The paper is about biases in GPT-4. They're not talking to GPT-4 to determine things about the real world. They're talking to GPT-4 to determine things about GPT-4.
> This phenomenon likely reflects that while initiatives to integrate women in traditionally masculine roles have gained momentum, the reverse movement remains relatively under developed. Not challenging you. Maybe it’s just the phrasing. But that sentence to me reads as if they think the presence of the biases in GPT means that they exist in the real world. And again, not challenging that the biases do exist. Just n…
I read as suggesting that the underlying bias is contained in the text corpus that GPT-4 was trained on, which in turn suggests it was put there by the humans who authored the original corpus.
That seems an entirely plausible mechanism to me.
Re: Surprising gender biases in GPT
#120I will preface this by saying I think people should do whatever it is they want to do, physical limitations willing. My (possibly contentious) question: Is this an OK thing to do? Why are we allowing a vast minority of people to rewrite our social norms? I went to the Barbie film as a man nearing middle-aged and had the most young girls, mothers, teenage girls, and women in general I've ever seen in an audience at th…
There's a difference between women or GNC folks choosing for themselves and cis men choosing for them. That's where gender bias is poisonous.
And what about the corrollary. Most men are engaging in paid labor at the behest of cis women， or the hope of getting with a cis women .