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

661–670 of 697 posts

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

#661

Post by throwaway894345 »

SpeedilyDamage wrote: post #656
throwaway894345 wrote: post #629

Earlier quoted context omitted.

I’m not on either team; I get flack from both. These people were patently not doing anything reprehensible, which is the whole point. Specifically, they were transparently fired/harassed/etc for deviating from the party line.

> These people were patently not doing anything reprehensible Says you. I say differently.

Agreed. Quite a lot of people think deviating from the party line is reprehensible, but these things are all squarely within the Overton Window irrespective of mine or your opinion.

I’m curious about why you find Fang, Shor, etc reprehensible though.

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

#662

Post by LawTalkingGuy »

manuelabeledo wrote: post #634
LawTalkingGuy wrote: post #603

Earlier quoted context omitted.

Here's a transcript that seems fairly accurate to the video interview I saw contemporaneously: https://www.politico.com/story/2017/08/15/full-text-trump-co... His own words: "you had people – and I’m not talking about the neo-Nazis and the white nationalists, because they should be condemned totally – but you had many people in that group other than neo-Nazis and white nationalists, okay? And the press has treated th…

Related: https://twitter.com/IamRageSparkle/status/128089153745134387... "Once you let a nazi in your bar, it becomes a nazi bar."

It was a protest in a park, nobody could keep anyone else out.

And does that apply equally to Antifa?

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

#663

Post by manuelabeledo »

LawTalkingGuy wrote: post #662
manuelabeledo wrote: post #634

Earlier quoted context omitted.

Related: https://twitter.com/IamRageSparkle/status/128089153745134387... "Once you let a nazi in your bar, it becomes a nazi bar."

It was a protest in a park, nobody could keep anyone else out. And does that apply equally to Antifa?

Are we really comparing people burning ATMs with nazis now?

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

#664

Post by SpeedilyDamage »

throwaway894345 wrote: post #659
SpeedilyDamage wrote: post #653

Earlier quoted context omitted.

Sure, the people are on Wikipedia, but the account of their controversy was not. Besides, what do any of these have to do with conservative views being suppressed?

They are both on wikipedia; I read them yesterday. On mobile now, but you can google them easily. They don’t have anything to do with conservatives views being suppressed because we weren’t talking about that. The original claim was that people who get “called out” have reprehensible views—I gave examples of people who were fired or harassed or etc for completely innocuous speech.

…what do you think the title of this submission is?

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

#665

Post by SpeedilyDamage »

throwaway894345 wrote: post #661
SpeedilyDamage wrote: post #656

Earlier quoted context omitted.

> These people were patently not doing anything reprehensible Says you. I say differently.

Agreed. Quite a lot of people think deviating from the party line is reprehensible, but these things are all squarely within the Overton Window irrespective of mine or your opinion. I’m curious about why you find Fang, Shor, etc reprehensible though.

I’ve said multiple times I cannot find the controversy you keep claiming happened to those two people, which to me speaks more loudly than the actual controversy they supposedly had.

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

#666

Post by LawTalkingGuy »

manuelabeledo wrote: post #663
LawTalkingGuy wrote: post #662

Earlier quoted context omitted.

It was a protest in a park, nobody could keep anyone else out. And does that apply equally to Antifa?

Are we really comparing people burning ATMs with nazis now?

Beating reporters, pulling motorists out of their vehicles, throwing molotov cocktails at occupied buildings, etc. All politically motivated violence aka terrorism.

Are you really defending that based on who does it?

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

#667

Post by belorn »

Llamamoe wrote: post #604
belorn wrote: post #517

Earlier quoted context omitted.

That is one theory I also suspect but I have yet to see a study that look at physically strength during conflicts between people. If it is true then among homosexual relationships the person with less physical strength should be detectable when looking at who is the attacker and who is the victim. It might also be non-linear. A person who has half the strength of their partner might be significant less likely to be t…

I think things change a lot when a dynamic is already established, and involves boundaries being pushed and probed, rather than being "crimes of opportunity" so to call them.

Based on the data it seems that the rate of conflict is a human constant. Two women living together will have a lower average absolute strength than two men living together, but from the data the rate of domestic abuse is identical. Similar, the strength difference in a heterosexual relationship is going to be greater than homosexual, and yet again the rate of domestic abuse is again identical to homosexual relationships.

It seems that the only thing that changes with gender is that when there is a conflict and it "eventually" develops into physical attacks, which is a strange concept in itself, the person who will be the attacker (or at least be found guilty in courts) will be significantly more likely to be a man if the relationship is heterosexual.

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

#668

Post by mellosouls »

SpeedilyDamage wrote: post #657
mellosouls wrote: post #544

Earlier quoted context omitted.

I'd dispute religion not being a choice and that it's perfectly acceptable to discriminate as you suggest, but, regardless, this seems like semantic cartwheels to avoid the clear, evidenced point that bias is present. Whether you or I approve of or align with that bias is irrelevant.

No it’s not irrelevant; bias is good when it hinders horrible things and supports great things. This is an argument on the merits of specific biases, not an argument that all biases are bad; no person or group operates with the intention of removing all bias, even if they claim to want that.

But as we all operate with bias, what appears to us to be good or bad is necessarily informed by it and makes subjective claims about which political biases are for the common good opinion only.

That's why we have democratic systems, so the mass of those opinions, informed by those personal biases override (in theory, if not always in practice) the opinions of a select few.

That is also why it is important to flush out as much as possible the undemocratic representation of ideological bias (left, right, whatever) in a tool as powerful as ChatGPT, when it has been clearly shown to be present.

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

#669

Post by manuelabeledo »

LawTalkingGuy wrote: post #666
manuelabeledo wrote: post #663

Earlier quoted context omitted.

Are we really comparing people burning ATMs with nazis now?

Beating reporters, pulling motorists out of their vehicles, throwing molotov cocktails at occupied buildings, etc. All politically motivated violence aka terrorism. Are you really defending that based on who does it?

No, I’m just saying that there are grades of malice, and I put nazis on the top, with a comfortable margin over any second.

Because their motivation is to remove any “undesirable” from their sight, and their definition of “undesirable” is wide and expanding and imminent, at least in the US [0]

In contrast, there has been one single murder in the past 26 years, where a far left activist was charged in the US, although subsequently shot and killed by the police [1].

So, I’m evaluating motivation here.

There’s no such thing as “both sides are bad”. Centrism is a lazy and uninformed view of the world.

[0] https://en.wikipedia.org/wiki/List_of_right-wing_terrorist_a...

[1] https://en.wikipedia.org/wiki/Killings_of_Aaron_Danielson_an...

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

#670

Post by antifa »

account42 wrote: post #615
brmgb wrote: post #140

Earlier quoted context omitted.

Hum, no, it's not. Of course, we marginally suffer from what we are importing and some students having too much free time on their hand try to argue rather unsuccessfully for intersectional studies but the models used in the USA don't map well on the European context for obvious historical reasons. Our left vs right debate also has nothing to do with the US one. From where I stand, the cultural gap between the USA an…

I can't speak for the entirety of europe but here in Germany, but the local equivalent of gender pronouns (/-innen) as well as rainbow flags in places where there is no reason for sex stuff at all is definitely a thing. Advertising is also pushing the same brainwashing (check the predominant races for each gender of couples in ads) and the mass import of muslim africans as well as the political refusal to admit when…

> Advertising is also pushing the same brainwashing (check the predominant races for each gender of couples in ads)

Brainwashing is when minority exists.

Post reply on HN