Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

davidrozado.substack.com

Post reply on HN
Print view View original

451–460 of 697 posts

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#451

Post by unshavedyak »

luckylion wrote: post #321
unshavedyak wrote: post #237

Earlier quoted context omitted.

That's interesting. I can understand being reasonably anti-Israel, but i'm suspicious that many people take the leap from being anti-Israel to being anti-Jew. Especially when you consider how many people of both the Jewish faith and/or ethnicity live within the United States. Tbh it feels like a straw man, but we're speculating anyway, so i can't fault you for it. I appreciate the discussion nonetheless. I'll definit…

> I can understand being reasonably anti-Israel Pretty much all "I'm only against Israel's policies" is shy Antisemitism. You can easily test this by asking them about other countries that have similar (or much harsher) policies as Israel and see whether the "no no, it's not about being Jewish, not at all" person is similarly upset about those and demands Boycott, Divestment, Sanctions. 99% aren't and it really is on…

Not sure i follow, are you assuming i'm in favor of other countries policies that match Israel?

To be fair i know little on the subject, and don't claim to. The only reason i said "i can understand .." is that i know it's a hot-button topic. Israel is often portrayed in the west as being overly militaristic towards neighboring areas. The United States itself does similar, depending on who you ask.

I judge both quite harshly on this front. I also know very little about world politics. Both in ignorance and on purpose, heh.

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#452

Post by erichocean »

sterlind wrote: post #354
jMyles wrote: post #349

Earlier quoted context omitted.

> Why is this considered good / normal / expected, but s/men/blacks/g and s/women/whites/g ...because the former holds true across correlations (and is difficult to confound anyway, since the group 'men' and the group 'women' are distributed in like proportion through correlated groups - rich and poor, white and black, around the world, etc), while the latter is very difficult to convincingly show because controlling…

It also makes a lot more sense biologically. Testosterone and estrogen have strong mental effects, and alter brain development. The brain is sexually dimorphic in complex ways, borne out by many studies. The brain is not racially dimorphic at all.

> The brain is not racially dimorphic at all.

Well, since dimorphic means "occurring in or representing two distinct forms", I think you are likely 100% correct.

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#453

Post by nhchris »

oneeyedpigeon wrote: post #365
tomp wrote: post #311

> men have a bigger tendency for violent behavior than women Why is this considered good / normal / expected, but s/men/blacks/g and s/women/whites/g (or asians, or muslims/christians) and it's discriminatory? (Statistically, both statements are justified. Morally, neither is, as we should treat people as individuals, not as members of X group.)

These statistics you're referring to, how comparable are the relative differences? Do they show that men are 4 times more violent than women and black people are 1.1 more violent than white people, for example? What are these figures?

That's a fair question. According to FBI homicide statistics for 2018 [1] and US demographics [2], the Black homicide rate per 100k is 7.9, while the White+Hispanic rate [3] is 1.2. The female (of all races) homicide rate is 0.42.

So the Black homicide rate is 6.7x higher than the White+Hispanic rate, while the White+Hispanic homicide rate is 2.8x higher than the female (all races) rate.

Edit: A different source [4] gives different figures, but these are "age adjusted": 20.9 for non-Hispanic Blacks, and 2.6 for non-Hispanic Whites.

[1] https://ucr.fbi.gov/crime-in-the-u.s/2018/crime-in-the-u.s.-... - and using homicide instead of something as nebulous as "violence" because it is the statistic most difficult to distort, and is immune to over-policing.

[2] https://en.wikipedia.org/wiki/Demographics_of_the_United_Sta...

[3] The FBI counts Hispanics as White, and has a separate 'Ethnicity' category where they mark Hispanic or Latino. Assuming all Hispanics in that table were counted as White yields an even lower White homicide rate.

[4] https://www.cdc.gov/mmwr/volumes/66/wr/mm6631a9.htm

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#454

Post by lmm »

chitowneats wrote: post #322
debacle wrote: post #319

Earlier quoted context omitted.

I think it's fair to point out that men commit most violent crimes. That's not good/normal/expected it's just a fact. AI could be used to better understand these associations (e.g. why is there a correlation between Asians and academic performance) and maybe help social leverage advantages more equitably.

Is there anything we could point about women? Nothing as terrible as being violent as far as I know, but what about other things? I agree with your premise but I believe it follows we must speak clearly in all cases.

Women who never marry get paid more than men who never marry once you adjust for hours worked.

(Once you adjust for education level this disappears)

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#455

Post by slg »

tomp wrote: post #433
lcnPylGDnU4H9OF wrote: post #413

Earlier quoted context omitted.

> men have a bigger tendency for violent behavior than women > black people have a bigger tendency for violent behavior than white people It seems way easier to justify that first statement. Are there physiological differences between a black man and a white man that are as significant as the differences between testosterone and estrogen?

The causal relationship between testosterone and violence in humans is weak and disputed. Should be as much of a reason to doubt it and brand such statements as bigotry as “blacks are more violent”, no?

That may be true, but we should still avoid false equivalencies. There is scientific evidence for a casual relationship between testosterone and violence. Even if that is debated and there is no full consensus, there is still more evidence for it than there is for the idea that a person's race has a casual relationship on their propensity for violence all else being equal.

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#456

Post by slg »

philwelch wrote: post #427
slg wrote: post #395

Earlier quoted context omitted.

>Statistically, both statements are justified. What can you point to that statistically justifies that "Black people have a bigger tendency for violent behavior than white people"? Showing crime statistics isn't enough. You need to show that given all the details about a person being the same, a Black person is more likely to act violently than an identical white person. You basically need to correct for all the soci…

> Showing crime statistics isn't enough. You need to show that given all the details about a person being the same, a Black person is more likely to act violently than an identical white person. You basically need to correct for all the societal reasons that result in people committing violent crime. I think you're conflating the questions of "which demographic group engages in higher rates of violence" and towards t…

They are the same question when someone implies that the membership of the demographic group is the "why".

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#457

Post by gwright »

jfoief wrote: post #336
debacle wrote: post #319

Earlier quoted context omitted.

I think it's fair to point out that men commit most violent crimes. That's not good/normal/expected it's just a fact. AI could be used to better understand these associations (e.g. why is there a correlation between Asians and academic performance) and maybe help social leverage advantages more equitably.

Parent's point was that it wouldn't be acceptable when done regarding some protected group (ex. blacks vs whites), so it shouldn't be acceptable when it comes to men vs women

There is no reason to think that interpretation of metric that compared "men" and "women" must be aligned with that same metric when men/women is substituted with some other populations. When you make the substitution you completly change the nature of the comparison. Whether either comparison is "acceptable" depends on the metric and the populations. There is no ethical rule that says all comparisons for group X must be acceptable or all comparisons for group Y must be unacceptable.

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#458

Post by chitowneats »

mxkopy wrote: post #326
chitowneats wrote: post #322

Earlier quoted context omitted.

Is there anything we could point about women? Nothing as terrible as being violent as far as I know, but what about other things? I agree with your premise but I believe it follows we must speak clearly in all cases.

This 'men are violent, therefore women must also be bad in some way' rhetoric makes no sense. There's no purpose for it other than assuaging sore feelings.

I didn't say they must be bad. I asked if we could make any group-based observation at all.

Also, your implied statement of "women, on average, have no traits that are unhelpful to the flourishing of human civilization", strikes me as terribly naive.

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#459

Post by throwaway894345 »

SpeedilyDamage wrote: post #268
throwaway894345 wrote: post #223

Earlier quoted context omitted.

> I don't think any honest defense of CRT said any of the things you claimed here, where have you heard those arguments? I've heard them all over for many years (including on this forum for a good while a few years ago), but I agree they aren't "honest" (and the dishonesty is precisely the problem). > Things that are "Nazi" or "white supremacist" ideas and ideologies get called out, and those ideas and ideology exist…

The Google memo was hateful, anti-diversity, and he got exactly what he deserved (he was fired). According to his LinkedIn[0], he's been working at an unnamed startup for the past 4 years, so it's not like he was "cancelled" in any real sense. And Lee Fang and David Shor are fine? I literally cannot find either controversy you're referencing on the front page of Google or Google News. [0] Not sure I'm allowed to post…

> The Google memo was hateful, anti-diversity, and he got exactly what he deserved (he was fired)

Yeah, this is flat out wrong. The "memo" (it wasn't a memo, but the media's gonna media) is publicly available and pretty short so we can trivially verify claims that it was "hateful" and anti-diversity. Notably, Damore's position on diversity via the document:

> I strongly believe in gender and racial diversity, and I think we should strive for more.

And his position on hate:

> Many of these differences are small and there’s significant overlap between men and women, so you can’t say anything about an individual given these population level distributions. ... Reducing people to their group identity is bad and assuming the average is representative ignores this overlap. (this is bad and I don't endorse that)

> According to his LinkedIn[0], he's been working at an unnamed startup for the past 4 years, so it's not like he was "cancelled" in any real sense.

I don't know anyone who defines "cancelled" as "permanently unemployable", but in any case I hope we can agree that getting someone fired for minor ideological differences is reprehensible. Maybe you would argue that these differences are very significant, but I don't see anything in the document that would fall outside of the American Overton Window (it's probably more moderate than many of his critics professed viewpoints on diversity, which is pretty obviously the actual reason he provoked such a reaction).

> And Lee Fang and David Shor are fine? I literally cannot find either controversy you're referencing on the front page of Google or Google News.

Lee Fang was harassed by colleagues and pressured to resign because he quoted MLK in support for non-violent protest and for Tweeting an interview with a black man who expressed concern about crime in his community.

David Shor is a data analyst who was fired for citing research on the efficacy of non-violent protest. He was fired because people on Twitter were contacting his company's management on Twitter and demanding his termination.

Of course, if your perspective is that firing someone for these kinds of minor ideological offenses is totally fine (so long as they're able to get another job?), then we're probably going to have to agree to disagree. That said, I doubt very much that the people celebrating or defending these terminations would be so cheerful when people on their side of the spectrum are terminated (and for whatever it's worth, I'm a left of center independent).

Re: The unequal treatment of demographic groups by ChatGPT/OpenAI content moderation

#460

Post by DaveExeter »

xp84 wrote: post #447
Llamamoe wrote: post #439

Earlier quoted context omitted.

One claim I've seen is that men commit more violent crime because it's more common for a man to be physically stronger than another person, while the rates of non-physial violence are much closer across gender boundaries.

> non-physial violence Come again? What's non-physical violence? Mental violence? Writing someone a mean letter?

Proxy violence.

https://youtu.be/W0FByIEijXI?t=22

Post reply on HN