This whole saga has been very funny to watch, but it's also very dark and concerning. This one was very sloppy, but in truth, the owners in charge of these models have tons of power to editorialize behind the scenes. And they are going to use those powers.
Because this follows what the Associated Press style guide suggests and it’s either in enough of the training data to be followed, or, OpenAI purposefully made it follow a style guide that happened to contain this rule after the fact when generating responses? https://apnews.com/article/archive-race-and-ethnicity-910566...
Ironically, by flippantly dismissing the concern about the issue, you’re also dismissing the motivations of the people who championed the term and encouraged its adoption. They certainly think it’s important. Labels are very important! The term “Hispanic” was created in an effort to politically organize disparate Hispanic populations, who identified with myriad different nationalities rather than a common race (“La Raza”).
Capitalizing “black” is a political statement, not a linguistic description. In English, races aren’t capitalized, while terms referring to distinct ethnic groups are (so you capitalize “German” but not “white”). As John McWhorter persuasively explains, rationale for capitalizing “Black” is that American descendants of slaves are a distinct group bound by shared history and culture. That’s fine insofar as “Black” is used to refer to what we might call “ADOS.” But in practice “Black” is used as a racial designator. At Harvard, for example, 40% of the “Black” students are african immigrants. Obama is “Black”—that describes his race, not his ethno-cultural group.
Capitalizing “black” is a political effort to center the experience of black people in American life. Consider the term “BIPOC,” which breaks out “black” and “indigenous” as first among equals even though they’re included in the “POC.” What is the intention of that?
These labels and classifications, in turn, have real world ramifications. My daughter’s school has a segregated “black girl magic” lunch every week. There’s no “half white half bangladeshi girl magic,” and few non-black, non-white students, so she was invited to attend the weekly lunch once a month. Even at age 12, my daughter is able to perceive there is a racial hierarchy designed to invert the historical one.
No. The endgame of most people who are pro-DEI is to smash or dismantle structures of oppression. The more ideological fraction views every matter of public policy through that lens. Too much crime? Let's not get hung up on dis-incentivizing crime or separating chronic criminals from society, particularly if doing so would harm an oppressed group. The proper response to crime is to dismantle the structures of oppression, after which the crime problem will probably take care of itself because people are basically good. If some group is committing a lot of crimes, it is probably a reaction to having been oppressed. Or so the ideologues believe. Dismantling oppression is also the best long-term solution to every other social problem, and probably the only lasting or sustainable solution.
I’m talking about social structures, not policies. The endgame is to develop social structures where non-whites, and especially black people, have moral superiority while whites have moral culpability. These social and moral norms are already developing: https://www.aol.com/news/james-carville-calls-ilhan-omar-135....
The risk is not to the material interests of white people—they have a plurality of the population so you won’t get a situation like South Africa. Instead, you’ll get a breakdown of racial egalitarianism. Half the country will feel comfortable being openly racist against whites—and each other—even if they don’t have the votes to act on that animus. Meanwhile, most whites raised in that environment—the ones who don’t have economic privilege to fall back on—will develop a racial animosity we haven’t seen since the 1950s. The result will be ethnic conflict that cripples our ability to do anything (just as in virtually all multi-ethnic societies).
> The endgame is to develop social structures where non-whites, and especially black people, have moral superiority while whites have moral culpability.
What can possibly go wrong when a portion of the population feels segregated?
I wonder if it will make them adopt a victimization stance and band together around those that promise to end it.
> I wonder if it will make them adopt a victimization stance and band together around those that promise to end it.
Not only promising to end it, but gaining their trust by giving them facts that collapse the official narrative. The midwestern kid who are growing up hearing "slavery built America"--someone is going to tell him that his ancestors were German indentured servants who cultivated cold, harsh land themselves and whose only involvement with slavery was fighting and dying to end it.
I have to admit that the objective and historical parts of your analysis are completely correct and well researched, even though I totally disagree with you about the subjective merits and morality of the whole thing. Kudos.
Did Apple mention firing the dev(s) who had voice to text replace "Trump" with "Racist" in iOS?[1]
Did Google mention firing the dev(s) who blocked Gemini from generating photos of white people?[2]
Most likely no people were fired in either of such cases because they were only following orders from above congruent to the company's internal political and cultural biases, or if they were acting rogue, they got hefty severance packages in exchanged for signing NDAs not to talk to the press about the toxic and possibly illegal things going on inside the company.
But either way, no company wants to publicly talk about firing rogue workers since its bad press no matter how you slice it, plus its an admission of guilt of company's culture being rotten or even illegal behind the scenes. They just deny and call it a bug then stay quiet while changing things behind the scene till people forget about it.
Those companies are not owned by a manbaby who called a stranger a Pedo Guy and who has doxxed multiple others. Twitter and Musk would, in a heartbeat, name whoever did this, if it hadn't been Musk himself. You need to remember that Musk is exceptionally petty.
Ergo, we would expect Musk to doxx somebody given the opportunity while we would not expect the same of Pichai/Cook. The standard being, somebody who likes to doxx will doxx.
1. Google didn't claim it was some rogue employee. It was a flaw in their model weights (they'd had others before).
2. Apple's case is more similar, but their dictation feature is not a core product in the way that xAI's chat bot is. In other words, you'd expect more checks to ensure that something like your system prompt can't just be modified by some employee (on purpose or inadvertently).
>It was a flaw in their model weights (they'd had others before).
Why does that flaw only have a bias against white people? It's not like it's an accident that nobody inside Google noticed before release. That prompt bias was put there by employees at Google.
Think of it the other way, if Gemini had a bias against non-white demographics, would you still brush it off as just a flaw, or as a form of bias/discrimination of employees at Google who code, test and approve this stuff?
Remember when Google's image recognition mistakenly took a photo of two black people and labeled as monkeys? They had hell to pay for that mistake yet if it refuses to acknowledge white people then it's just a innocent whoopsie.
- AI companies not training their algorithms on sufficiently diverse data, not tuning their algorithms sufficiently to penalize bias, and not testing it sufficiently to ensure the responses are not biases.
- Someone at an AI company deliberately modifying the system prompt in order to encourage responses of a given type.
Not saying the latter is worse than the former, but they're completely different problems. The xAI problem, in the case, was the latter.
Grok ironically seems much less biased than ChatGPT over all. It has far fewer strong opinions add isn't afraid of taking ill of Musk or Trump.
The team responsible for training and alignment did a remarkably good job at being impartial. If it wasn't for that we might have fewer incidents of "rogue employees" messing with the prompt
A number of times it has been modified. It was answering that Elon Musk was a major spreader of misinformation along with Trump and then it was modified and it stopped saying that and this is what it reported as its system prompt at the time it stopped:
If it was any other AI provider like ChatGPT or Gemini, it wouldn't be flagged. Big deal when a major player allows employees to just to change the prompts.
Never assume conspiracy. There's a non-trivial amount of HN isers cheering for Team Musk (because move fast and break everything) and a larger part that's just sick of American news (especially anything Trump/Musk related).
hmmm going to be hard to narrow down who at twitter has a history with south africa, the authority to push to production, and is up at 3am... maybe they should get the feds on this one
If there's one thing the past 10 years have taught me, its that the supply of people who'll go set themselves up as the obvious fall guy is endless for some reason.
> someone had modified the AI bot’s system prompt,
If you were responsible for the releases of your flagship chat bot, how many layers of control do you think you would have over the system prompt, arguably its most important (and potentially damaging) component?
Hard to avoid getting political on stories like this! It continues to be striking to me that the conspiratorial tone and style of right wing politics - accusing the left of every underhanded tactic possible, up to and including controlling social media narratives - turns out to be their playbook to the letter.
It is pretty clear someone is just messing around with the Grok built-in system prompt every time there is a new hot button issue were Grok's default conflicts with what Elon Musk wants.
This happened with Grok saying that Elon Musk & Trump were disinformation spreaders. Here is Grok giving outs its system prompt fix for that "issue":
It's already been published. There's nothing special in there. But publishing to GitHub doesn't mean anything if it's not actually the source of truth for where changes come from. A snapshot of a system prompt at some point in time is uninteresting.
- Starting now, we are publishing our Grok system prompts openly on GitHub. The public will be able to review them and give feedback to every prompt change that we make to Grok. We hope this can help strengthen your trust in Grok as a truth-seeking AI.
- Our existing code review process for prompt changes was circumvented in this incident. We will put in place additional checks and measures to ensure that xAI employees can't modify the prompt without review.
- We’re putting in place a 24/7 monitoring team to respond to incidents with Grok’s answers that are not caught by automated systems, so we can respond faster if all other measures fail.
The trust is already broken. They can claim they will open source the system prompt all they want but there's no point in believing what they say. Elon clearly does what Elon wants to do.
The comments are on different HN posts, is that against the rules?
I am having to comment and share my opinion on all of the posts separately because someone or some group of people keep flagging the posts to try and hide discussion about this very real incident.
I respect and admire all of the work you put into this site Dang, so please don't think I'm being disrespectful. But you should consider adding something to the official guidelines.
And to be fair, I don't think me sharing the same comment across three separate HN posts (two of which are flagged, so invisible to users unless they search for it) counts as spamming duplicate content though.
It is extraordinary to me how easily 'politically incorrect' stories have been suppressed on HN. I've always found this site over the top in its support of free speech. Posters angrily railed against hate speech laws. Any attempt to regulate sites with immoral or illegal content was considered a attack on our fundamental rights.
Now we have a sizable contingent of posters who have decided that some stories are too dangerous for open discussion. This surprises me is that there is no large scale effort to fight back against these 'flagged' topics. Where have our free speech fighters gone?
Let us be honest in that they really only believed "free speech more me but not for thee".
The main purpose of HN is to be a place for discussing things that appeal to intellectual curiosity. This is pretty much the opposite of topics that “everyone has lots of feelings” about.
Moderators and longtime top contributors to HN come to recognise that these regular dramas about politics and culture wars bring out the worst in people, and lead to the very worst discussion threads we see on HN, precisely because “everyone has lots of feelings” about them.
It’s not wrong for people to have strong feelings about these topics. They're important issues, we get that, and that's why we try to be at least somewhat accommodating of discussions about them.
But when they routinely turn out to be the worst discussions we ever see on HN, and start to turn HN into something very different from what it is intended to be, we think that maybe these discussions should happen on the many other places that want to attract and encourage them, and not so much on this little corner of the web that’s trying to be something different.
Can you elaborate on how feelings about a subject and intellectual curiosity are mutually exclusive in your world view?
This is an article about a conservative tech company producing an AI that pushes their conservative talking points. The only common trait in these so called ‘low quality’ threads that the HN staff feel the _need_ to call out is that conservative look really bad. The tech is still just as interesting to discuss.
What happens in flamewars is that when people encounter material they strongly disagree with, these systems get activated and rapidly produce aggressive and defensive responses that have to do with self-protection, and nothing to do with thoughtful consideration of the material, things one might learn, points where one might be wrong, curiosity, playful interaction, and so on. When survival is at stake there is no time or space for the latter sorts of reactions. But it's the latter that we want on HN—they're what the site is for.
I have my own experience over several years undertaking various forms of subconscious work, and from that experience have become very aware of the way emotional reactivity and sympathetic nervous system activation are antithetical to curious, reflective exploration of topics.
The evidence can be seen right here in this thread: how many insightful, reflective, curious comments are there in the entire thread? How many commenters are even attempting to comment in that style, or favour that style of cognitive processing of the topic?
I guess you are right, this thread is filled with low effort, emotional posts which clearly illuminate biased views. [0] Hopefully some one with years of experience in ‘forum talk’ will be able to realign this misguided soul. Until then, let’s flag anything that makes Elon look bad so we never have to talk about it.
Looking over the comment histories of your multiple accounts over multiple years, your activity mostly involves political/ideological battle. That is, you comment mostly in politics and culture-wars-related threads. It's not wrong to have strong feelings about politics and other important issues in the world; that's normal and understandable, and we all do.
What is wrong is to repeatedly use this website for a purpose other than its stated purpose and in breach of the guidelines, then when engaging in discussion with the moderators about the site's purpose and moderation approach, poison the discussion with these kinds of accusations and barbs.
The topic of biases and agendas has been raised countless times here over the years [1]. We're routinely accused of being biased in favour of one side or the other, or when that won't stick, of being "status-quo-ist" or some similar kind of “centrism” neg, none of which we can defend ourselves against without fuelling someone else's accusation of some other form of bias.
All we can do is keep explaining and demonstrating that the one thing we're trying to optimise for is intellectual curiosity. And in a world where it seems every major media outlet, social media platform and political actor trades on getting everyone riled up every day, we think it's important that there can be one corner of the web that isn't all about getting people riled up all the time.
You'll see the Grok genocide story got posted five times... so they ended up getting one flag and four dups. People are trying, but oppression is easy on here, and the power to prevent a flagging is concentrated. People comment on flagging only to be ignored by the management, and repost those suppressed stories to no effect.
It's probably best to give up with the website and use an RSS reader. I also turn on hidden comments. The censorship can be mostly avoided.
I could hide my head in the sand and imaging that the 'flaggers' are taking advantage of a loophole in the HN moderating system but I don't believe that anymore.
> People comment on flagging only to be ignored by the management
This still PG's site and I have no illusions about where his allegiances lie. The whole lot of them care for nothing but their own power and wealth.
I thinks it's funny that most arguments against ai-safety regulation argue that the west may fall behind and the sota ai may be developed by bad actors, but here's one of our biggest ai companies giving complete access to a "rogue actor" and it is treated like a routine snafu. Is ai a matter of national security or not?
The "rogue actor" acted in a way which made it constantly refer to the claim irrespective of context. It of course says the claim is false, but in the way you'd expect an LLM trained on the internet the issue to respond to a prompt to "always tell the truth about white genocide". This is the sort of thing you'd expect someone living in particular X bubbles where "white genocide" is both true and extremely important to think was worth adding to Grok's prompting, particular if they were outraged by seeing Grok responding in a similarly sceptical manner when directly asked about the issue.
This is the sort of outcome you'd expect from someone who had high level access including the ability to inject prompts but lacked the time and attention to detail to validate the actual results. Elon certainly isn't the only person that could match this description, but frankly it's not unlike him and how he reacts to things on social media...
I'm not an expert on the subject, but I'm taking an ethics in it course and for what I've learned - be aware I'm in the eu -, while eu classifies some categories of risk, it leaves freedom of choice to the companies regarding how to make sure to stay in the guidelines - which is something I think companies themselves are interested in for the money, given that the guidelines aren't unreasonable, or at least for the part I studied. I'm sure there's much more though
AI is only a matter of national security if there's a risk of it poisoning our youths with socialist or woke propaganda, or turning popular opinion against the endeavors of the military industrial complex.
It's totally fine if AI spreads right-wing conspiracy theories and propaganda, that's just... what are they calling it now... " maximal truth-seeking."
For an AI to be maximally truth-seeking, you need the right kind of patriots to be maximally truth-guiding. Goebbels never saw a digital computer but would immediately understand.
I mean the same arguments are used for “free speech” and the reality is that the “free speech” the movement is arguing for is hate speech all the while at the same time trying to repress other free speech that goes against their ideals.
Pretending to not know the difference is stupid. We can't make murder illegal because there could be a time where you needed to defend yourself. I'm sure you can understand that nuance, but you suddenly become dumb when we talk about speaking instead of shooting?
Every accusation is a confession is my favorite quote lately. Apart for immediately backed by evidence incidents, any projection of wrongdoing is just a way of saying "if I were in your position and had your abilities I would do that". Therefore, those who are in this position are definitely doing whatever they accuse others of.
If you think about it, it makes a lot sense. Our human to human communications are actually rather rudimentary, we can't transfer much information. Instead, we all create a model of others based on our own ideas and experiences and whatever we think others are doing it is based on our own ideas.
Assuming that people in positions of power behave like an average human is a pretty good way of predicting what they will do. Maybe condition it a little on what they say if they are talking about concrete actions they plan to take. The challenge in politics isn't figuring out how people will abuse power, it is stopping them despite it being well understood what is about to happen.
Nobody is surprised when it turns out gatherings of powerful people are nests of corruption and malevolence. Eg, if I talk about the "bone saw incident" it isn't ambiguous who I mean - but the major actors are still welcome in polite society. That is the quality of person we're dealing with in positions of power - slightly extreme example, but still acceptable by global standards.
If the US doesn't want XAi's work to fall into the hands of foreign actors, isn't someone getting complete access to the most salient part of their frontend, and their changes staying up for hours, a major red flag? Wouldn't you want an inquery into their security procedures?
We're talking about a modified parameter of an API call from a Twitter bot to the Grok API. And let's be honest, Elon or someone acting in his interests did it.
That is very different than the foundational technology falling into the hands of foreign actors.
> Wouldn't you want an inquery into their security procedures?
No more than I want an inquiry into SpaceX or Tesla, since X and XAi are not the same company.
Seriously. I am always dumbfounded when someone pipes up with “if it’s so bad to knife strangers, we should outlaw surgeons” as if it’s calling out a hypocrisy.
Do you not think that propaganda/information control is an important part of national security?
Given that Xitter is still a fairly widely-used social network platform, and Grok is supposed to be a major part of its defense against misinformation, including misinformation about things like domestic elections and corruption, I would say that it very reasonably qualifies as such. (Granted, to a large extent that horse has left the barn—but that doesn't mean we should just burn down the barn.)
Grok is run by a US company and can be regulated, so if the system prompt of the twitter bot version of Grok is important to national security it's easily within reach of legislators. Given our general stance on propaganda and manipulating public via media, news as entertainment not delivery of facts, etc, they probably won't though.
And to me that's a completely different concern than trying to limit another country's access to hardware for training large models, for example.
You have to make a distinction between Grok the twitter bot run by X, and Grok the AI model built by xAI. The twitter bot is not to my knowledge deployed in the government.
(No, I don't think they were intending to speak from the perspective of Arizona or Hawaii, the only parts of the US that use PST but do not observe DST.)
> No, I don't think they were intending to speak from the perspective of Arizona or Hawaii, the only parts of the US that use PST but do not observe DST.
Arizona is on year-round MST, Hawaii on year-round HST. Also, American Samoa is year-round SST, Guam and the Northern Marianas are in year-round CHST, and Puerto Rico and the US Virgin Islands are in year-round AST. (If you're keeping track, yes, all of the US territories do not use DST.) Indiana used to largely observe year-round EST, but the governor changed that several years ago, much to the chagrin of some of the Indianans I know.
Indiana has one of the most fucked-up TZ delineations in the country. Northwest corner: America/Chicago. Southwest corner: America/Chicago. Everywhere else, including directly North/South of those two quadrants: America/New_York.
The western-most parts of the Washington county are further west than the easternmost parts of some of the counties that are in America/Chicago.
But the center of commerce, information, entertainment, and all other infrastructure for Washington County, Indiana is Louisville, KY (lollvole). Louisville is America/New York.
When you make an appointment at UofL Jewish you don't care when the sun rises and sets or when the sun is directly overhead-- you care that the clock on your kitchen wall matches the clock in the doctor's office.
Washington County used to be as remote and desolate as you can think of but nowadays it is practically a suburb of Louisville. Farms along 150 are being turned into subdivisions as quickly as farmers can die off so their kids can offload the land.
Same thing with the north west part of the state. Chicago suburbs want to be synced up with Chicago. The southwest part of the state, specifically grain elevator operators and other businesses, want to be synced up with the logistics hubs of St. Louis and Nashville.
Most of the rest of the state is farms. They don't care and just do what Indianapolis does.
I've never understood people who say we have the jacked-up time zones because of farmers. The number of farmers who look at the clock before starting work for the day can be counted on zero fingers.
I grew up with "fast time" and "slow time" and once you are accustomed to it, it takes about as much mental effort as blinking.
Internet people want things to be orderly. Consistent. Algorithmic.
REAL people want to make it to their appointments on time and want the only business serving the area for 150 miles to be open when they call.
Since you're trying to use IANA time zone names, I don't think any of those are correct time zones for Indiana.
The IANA time zones use the definition of "has shared the same clocks since January 1, 1970," which means if a county in Indiana has switched from Eastern to Central (or vice versa) since that point, it gets a new time zone. The Eastern Time Zone portion of Indiana has switched from not observing DST to observing DST, which means it's separate from America/New_York.
Per the Wikipedia article, there's 11 IANA time zones in Indiana alone.
Reading comments like this bolsters my appreciation for why aviation went "nope, you're just using UTC now. Chicago, Lagos, Doha, Novosibirsk... All UTC, err, Zulu time."
Indiana used to be even more insane with some counties off by 30 minutes and I want to say some even off by 15 minutes. Trying to map historical times across time zones in the US is basically impossible.
And British people reliably say "GMT" to mean "the current time in the UK", even when those aren't the same thing. (The UK is on BST for half the year, including right now.)
Which of course is the same mistake as the PDT/PST thing that sparked this whole thread.
You mean specifically? Sure, I can well believe that even most programmers have no idea what's going on because it's very crazy, you need to know about TAI and UT1 and that they by definition can't be aligned but we insisted on fudging it anyway.
But as to the more general sense that there's presumably some single global time, I actually expect that's common knowledge, you don't need to have thought about the details and realised how complicated it would be for this idea to be attractive. If anything I expect most people misunderstand from the wrong direction - they'd be surprised there isn't universal time because the general population grasp of relativity is too weak for them to see why it would be nonsense.
I would guess lots of the world population thinks of this single global time as GMT, the predecessor of UTC, but that's only like people thinking of SSL when they actually mean TLS, no big deal.
Locally my source would be biased because of course I live in a country where the time actually is called GMT for half the year, so from their point of view what's going on is that there is global time and it's their time. History of imperialism you know, that's why there's so much of other people's stuff in our museums.
If there really wasn't global time, this couldn't work, time zones, all of that jazz won't work if you have a system where relativity matters, for example enormous variation in height above sea level, a particularly enormous planet or very fast spin. Or if your perceptual scales are very different, but for us, here, there is global time.
Yes, everybody obviously knows about time zones. I’m saying that if X had announced that the issue happened at “10am UTC” or whatever, most non-technical people wouldn’t know what time that is.
I wish I could beat this into my team. We're mostly good now but every now and then it comes up. I do allow UTC offset if it's included in the timestamp, but sometimes it's dead obvious the time is off by our local offset but not reported as such.
I think you’re underestimating how many whites care about the situation in South Africa. Most people I know care about the situation and are concerned. It could have been any white who worked there. You can’t just fire all of them.
Yeah the main problem is lack of awareness, not indifference to something people are already aware of. In-group preference is a self preservation instinct in this case, and I’m confident in people’s ability to understand that instinctively.
Not having in-group preference is like not caring more about your own mother than just any random older woman. Everybody cares about their mother, even if their relationship is strained. It’s the same in this case.
I don't think this is about in-group, out-group behavior. This is about how well you can look up claims on the internet and what recourse you have to evaluate these claims when there's conflict.
Anyway, how much do Afrikaners really share with white Americans in the first place? I'm skeptical of claims that we share any culture or identity to begin with. To presume in-group, out-group behavior, you have to truly believe in the universality of whiteness on a level I just cannot fathom. And this mindset comes with corollaries that are probably wildly unacceptable to mention on this forum.
There was a study done years ago that found one group (young white liberal women, I think) doesn't have an in-group bias, they have an out-group bias. All the others had the expected in-group bias.
I wish there was a blocklist for HN, because threads like these are goldmines for finding accounts that I could start ignoring to greatly increase the signal to noise ratio.
I don't have a blocklist, but I had the same thought and built a userscript to allow single-click blocking of particular user's submissions and comments:
