His main point: "This event is not significant for culture war reasons. ... This event is significant because it is major demonstration of someone giving a LLM a set of instructions and the results being totally not at all what they predicted."
Yishan did this same sleight of hand trick with twitter before Elon bought it[1]. Then his point was that twitter doesn't have a bias, it's simply a normal moderation problem (noise-signal ratio). Now that twitter is owned by Elon, everyone can see it obviously moved right.
What normies are fairly noticing in these corps is moderating not by the overton window of the normal consumer, but by the overton window of the managerial class who run the business.
> What normies are fairly noticing in these corps is moderating not by the overton window of the normal consumer, but by the overton window of the managerial class who run the business
Now the question is, why is the managerial class so weird and creepy? The Gemini images didn’t come out of left field—that’s what Hollywood movies look like these days too. Random non-white people in contexts where it makes no sense, or racially balanced friend/family groups that don’t map onto how people actually associate in real life.
My family looks like one of those Gemini images—my brother and I are south Asian, my wife is British American, her brother is Samoan/black/Asian, his girlfriend is Filipino, and my sister in law is Taiwanese. But it’s a fluke—you don’t see that much mixing organically pretty much anywhere. The overwhelming majority of people are in same-race marriages and have homogenous circles of friends. The world still looks like Friends or Seinfeld, not whatever Hollywood is showing recently. So it’s very creepy to me that there are a bunch of people in Hollywood deliberately making casting choices with these unrealistic representations—or, as in the case of Gemini, inserting hidden text into prompts to achieve that effect—as if somehow my family is better than one that’s all white or all Filipino. The contrived nature of it makes race seem like a big deal, instead of something irrelevant, and it’s creepy and weird.
Wouldn’t Twitter be an unambiguously more successful business if Elon’s moderation moves brought it more in line with “the normal consumer?”
Seems to me he simultaneously put out a dogwhistle for one audience (far right commentators agree on this) and was openly hostile both to individuals and standard sensibilities among another audience (e.g. it’s distasteful to treat employees of your newly acquired company the way he did).
That could fully explain “the shift” without any “omg conspiracy Overton window biased moderation” mumbo jumbo.
But that was part of “the shift” that GP is crediting for moving toward broader appeal.
Agreed on the general point, this stuff can’t really be disaggregated in a meaningful way, but of course it doesn’t need to be disaggregated: the overall set of Musk’s decisions have been extremely destructive to the company (so far).
Sorry. I wasn't claiming Elon made Twitter more appealing, but rather was trying to point out that when management changed so did the bias.
I realize my poor grammar made it difficult to fully capture what I was trying to say.
My core argument is that I have now seen twice what looks like Yishan making excuses for moderation decisions that fall outside most people's discourse window.
I also am not making claims about what makes more money. It could possibly be true that people with a particular political pursuasion are more likely to convert.
> Wouldn’t Twitter be an unambiguously more successful business if Elon’s moderation moves brought it more in line with “the normal consumer?”
Purely for revenue maximization, I think it would be optimal to give each user a view through their own personal Overton window. However, people in positions of power often find additional goals beyond merely maximizing revenue. Such as "moving" that Overton window.
Xoogler here: pointing out such things internally is incompatible with continued employment at Google (see eg Damore’s “echo chamber” essay, the echo chamber got even more airtight since then). So the end result is quite predictable. Moreover, this will continue, and i don’t see how they get out of this hole without firing a bunch of internal “activists” who spend their entire days stirring the pot on Memegen rather than actually working.
"Bring your whole self to work" is a joke unless you hold progressive opinions. Antifa and "resistance" stickers on laptops or backpacks are fine, as is openly wishing for the death or bodily harm of sitting politicians from the "wrong" party. Diversity groups for just about everyone except you know who. Meetings on work time openly pushing progressive agendas.
God forbid you display MAGA swag or even a Gadsden. Diversity is supposed to yield benefits of expanded opinions. Now there is no diversity. Everyone looks a bit different but the only opinions expressed are in one direction.
Thanks! For another tip, it goes equally ironically well at rallies for folks trying to impose governmental control over individuals' medical decisions.
Turns out there's more than professionalism at stake.
I'd link the Gadsden flag Wikipedia page but that might also not be professional. Also, why is it unprofessional to point out an ignorant irony that I can find countless examples of? Is it uncomfortable when I point out the obvious cognitive dissonance?
I am familiar with the flag. While it's fun to poke at people who don't know better, this sort of thing cuts both ways. For example, I could find countless examples of people who align themselves with defunding the police, yet lament at the fact that their neighborhood is now suffering from crime.
Either way, your public fear of posting a Wikipedia link says a lot about you.
Why would you want to display Antifa/ACAB swag? Why would you want to marginalize one of the major political movements in the country, while boosting marginal movements?
In case my point there wasn't clear, it's common for people to adorn their property with their support for a campaign, such as lawn signs, bumper stickers, or pins. Such displays are rare at work, as most people have more tact than that, but I've still seen Obama, Hillary, and Bernie campaign logos. A Trump one or "Make America Great Again" one in these environments would be a faux pas to say the least.
If we really cared about "our workforce reflecting our customers", as the mantra goes, we'd want people that empathize with those points of view. Instead there appears to be nothing but active disdain for them in this industry.
I mean I'm sure in a company of 140,000 people there's probably many cohesive groups of 100 people that will be super right-leaning. Like, most of the crew at the Jackson County, AL or Montgomery County, TN or Mayes County, OK datacenters. I'm sure there's areas of Google where you can express very right-wing opinions to your team and adjacent teams and still get promotions. Heck it's likely that being openly left-leaning in some groups will get you canned.
In data centers maybe, but in software even moderate conservative positions are inadvisable. I don’t consider myself to be in any way “right leaning” by the way. I’m in favor of universal healthcare and gay marriage yet against allowing biological males in female sports, for example. Don’t like it? Not much I can do about that, I’ll stick to my guns no matter what derogatory label you assign. I’m not overly attached to fat paychecks either.
So, which area/group were you in? Also would you say that expressing your view that "AMAB people shouldn't be allowed to compete in female sports" would get you ostracized at Google during the time you were there? I'd think that's a pretty milquetoast position to hold.
In one of the ML groups, working on LLM training/serving infra, though not in Brain or GDM.
And yes, you’d get some quality time with HR for saying something like that there, and you’d get fired if you keep saying it. People don’t quite appreciate the extent of this insanity there. The internal discourse is completely devoid of anything that challenges even the most insane of DEI tenets.
Sorry, but why would you want to publicly (at company meetings I assume) verbalize this position of yours? Not only once but multiple times? It feels so very strange to me.
That’s the thing, you don’t, at least if you want to continue receiving your paycheck. It’s but one example, there are hundreds more. The other side “verbalizes” their “positions” so often your ears would shrivel up and fall off. The problem is that nobody pushes back on that insanity, and we get what we got with Gemini.
Indeed. I've yet to see a single youtube evaluation video of Gemini that didn't at some point point out that Gemini is hallucinating or trying to convince the user of something that's total bullshit.
It's pretty clever about it. Doing this like. "Smart, clear, eloquent fact A, phrased longer than necessary. Bullshit (short sentence). Smart, clear, eloquent fact B, phrased longer than necessary".
But the bullshit is always ... pointing in the same direction. I must say I wonder if it isn't exactly the result of what is being blamed. The "wokeness". Regardless of the actual ideology people are trying to impose on the model ... at least in some cases the model is lying to make it's answers satisfy what is obviously a political ideology imposed on it.
Because look at those pictures. To a transformer, pictures are stories. A long sequence of tokens that convey thoughts. They are very largely correct. There's a clear Hollywood influence, as you'd expect, but otherwise they're mostly correct. If you took the total amount of information in those pictures, I bet you'd find 99% of it matches the training data! And making the characters black or asian is a little hallucination somewhere in the middle of the picture's story. And, surprise! Those modifications are there ... because that's EXACTLY the ideology Google is trying to impose on the model.
Could it be that the model is lying ... because it thinks that's exactly what you've asked it to do? "System prompts" mostly are just pasted before the input. If the system prompt says "assume all historical accomplishments were by black or asian people", then the model will assume that's what you want!
Makes you wonder ... how often are "hallucinations" the result of the model not making a mistake, but "purposefully" lying to make it's answer comply with a particular worldview? Because that's what you asked it to do, if you look at your full input, including Google's system prompt?
Hell, I'd ask the same question about a lot of human answers too. As soon as a subject becomes even a little bit controversial, people outright lie en masse. I know nobody wants to admit it but that's exactly how humans work.
An image generator is supposed to hallucinate. That's the whole point. If you want a non-hallucinogenic image, then use Google image search.
It's just so ass-backwards to release a creative tool and then attempt to constrain it to a pre-determined set of imagery.
Like, imagine an AI that can produce images of any crazy thing you can think of? Wouldn't that be amazing?
Billions have been invested in this technology, and it sort of works!
The problem is that if you release it to the public people are going to use it to generate any crazy image they can think of. We can't have that. No good. No good at all.
The culture war take is fairly dumb -- mostly because absolutely nobody is confronting the much more obvious and important point that it's exposing how self-contradictory our expectations around diversity are.
But this take is even dumber. Of course a gazillion people in Google anticipated this result, because this result is obvious if you think about it for even a minute. It went forward not because nobody foresaw potential issues, but because business demands overrode that foresight. It's not complicated.
> That’s not why this is important. This event is not significant for culture war reasons.
He makes an interesting point but I still think that this raises some big questions around how they attempt to correct bias by injecting their own biases.
In this case they tried to correct racial representation in pictures but what if they try to correct and influence political opinions next? Make certain groups look like glorious freedom fighters, and others like bloodthirsty monsters? Maybe attempt to push certain parties to make them win the next election. Given they offer no transparency on their prompts how will I know they are not pushing other ideological angles?
Nah, Google is just racist. Jack Krawczyk is on record making a bunch of racist remarks (against white people) and Youtube was sued multiple times for racial discrimination (against whites and males).
It just permeated the culture and produced a self censorship effect. Who's going to stand up, in a company they know to be racist, and flag racism in their product as an issue? Nobody.
And so even if anyone noticed the problem before release, they didn't report it.
Certain people often complain about “systemic” problems but never aim the magnifying glass at themselves and go on to claim there’s no censorship effect, there’s no bias, etc.
Let’s agree for the sake of argument that you can have whatever guardrails you want around your AI. It stands to reason that you want to test those guardrails pretty thoroughly and see if there are any unintended consequences like this. It turns out that Twitter is full of irreverent shitposters and trolls who really want to make this particular guardrail self-defeating, and as a result they managed to do it. This raises a question: why can’t Google, one of the richest corporations on earth, have testers who are similarly capable? OK, you’re never going to match the sheer creativity of all of Twitter, but OTOH Gemini seemed to ship in a broken enough state that you’d think they wouldn’t have to do too much testing to uncover this problem before release.
The problem is, when it comes to “sensitive” or potentially “offensive” issues, it’s very easy to effectively stifle the sort of irreverent lateral thinking required to do thorough testing. If the intended guardrails are excessively sensitive to begin with, which seems likely based on some of the reports I’ve seen on Twitter (e.g. categorically refusing to generate images of white people even when explicitly asked to, or categorically refusing to generate Norman Rockwell style depictions of classic Americana), that implies a similarly sensitive corporate culture. I wouldn’t rule out the possibility that testers were afraid to even try and test for these failure modes because they reasonably believed they might be fired for it. But what’s more insidious is a widespread avoidance of even having thoughts that seem out of bounds.
Sounds like algowashing, a.k.a. plausible deniability for the results of their models. Because it's a statistical model and not lines of code, they could go to the public and say "This model is sO cOmPlIcAtEd that not even we could have foreseen these results!" when their thumbs are on scales all over the place. I fear what happens when AI criminal profiling models flag undesirables as definitely psychotic serial killers ("just sprinkle some crack on him/some CP on his computer" with extra computer-mediated steps) or when, say, a hospital's AI cancer diagnostic model flags a sponsor's rival as "terminal - hook him up to the MAID machine and run him through the checkout line", and no one can point to evidence that shows the models are being manipulated despite that everyone knows they are.
> got a totally bonkers result they couldn’t anticipate.
I fundamentally disagree with this statement and frankly the entire line of reasoning
This isn’t some unforeseeable 3rd order side effect coming out of a black box from left field.
You’re always injecting an instruction to put in diverse people so now it’s always doing just that. Entirely predictable.
This is like adding a system prompt saying always put tamagotchi into the picture and then being puzzled by why a request for a dog painting includes an inappropriately present tamagotchi. Like wtf did you think would happen?
I agree with you that it's entirely predictable. Why was it not predicted by a company that has achieved hegemony over so many aspects of our online lives and therefore must be pretty good at stuff?
Other people offer:
* google drove their head up their own ass
* actually they expected this behavior, but didn't think there would be backlash
* google didn't scope for the normal consumer but rather for their managerial class
I'd say that the first and third items are close to the "result they couldn't anticipate" reason. The second reason sounds google-ish, but they're also trying to win in a market where they're behind, so is their blindness to real humans that deeply institutional?
> This event is significant because it is major demonstration of someone giving a LLM a set of instructions and the results being totally not at all what they predicted.
I think they must have expected exactly the behavior on display. What they might not have considered is the human backlash.
> First, to recap: Google injected special instructions into Gemini so that when it was asked to draw pictures, it would draw people with “diverse” (non-white) racial backgrounds.
I just love how this "solution" to their problem of their training set being deeply flawed is such a hilariously bad kludge hack of the sort a jr programmer would make and it's amazing that they went with it.
> their problem of their training set being deeply flawed
The world is deeply flawed. If you do an unbiased search [0] of the real pictures of people in such-and-such profession, you get a vague approximation of the people actually in that profession. This does not approximate what Google or anyone else wants the demographic to be — it’s the demographic of people matching the query, times the expected number of times their photographs appeared in the training set, times whatever weight the model gives to those photographs. The naturally picks up whatever may be wrong with society that resulted in this distribution.
If people of X race/gender/whatever are underrepresented in profession Y, and you search for photos (or synthesize them by generative AI), do you want a sample drawn from the actual distribution of photos or from a distribution with the underrepresentation corrected? And what does correcting it mean?
[0] whatever that means — maybe just giving uniform weight to all the distinct pictures of decent quality that are on the Internet. The actual selection criteria won’t change the result too much.
Since the earliest reporting on this topic, "bias" in AI has never referred to accuracy. It's not the case that certain professions have a demographically representative proportion of all ethnicities and both sexes. Accurately portraying this reality is "bias". Portraying a fantasy with proportionate representation is "removing bias."
I think the problem here, well before even getting to this notion that Google is trying to engineer some favoured alternative reality outcome of theirs, is that it's not clear that the model, relying on the data set available, is even creating a correct representative version of reality prior to Google putting their thumb on the scale.
Surely there was internal dogfooding for this feature, I do not believed they would add a new system prompt and then released it to the public with little testing and then get surprised that it was doing this like the author suggests happened.
This is also a symptom of a corporation not having dedicated QA people. Sure, developers and product owners can check stuff, they can even read code and find bugs there. Amazing, we don't need QA then, right? Well, that's what we can sometimes get without QAs - an amazing code which actually works and works better than competitors. It just sometimes break in the areas not tested sufficiently or at all.
> Google originally did this because they didn’t want pictures of people doing universal activities (e.g. walking a dog) to always be white, reflecting whatever bias existed in their training set.
If the results were actually proportional to the training set I think that would be fine (assuming the training set had some reasonable amount of diversity, which I bet it does; one could always do biased resampling too). I'd be concern about mode collapse, where it would just show the most common class (eg white American women) and nothing else. But I thought we had solved that issue.
In a nutshell for those without a X account: the problem here isn't that Google's AI gave "woke" results. The big problem here is that even one of the biggest AI specialists in the world, Google, wasn't able to fully predict or control the output of its AI. If we ever create a truly powerful, superintelligent AI under the assumption that we'll be able to control it, this incident shows that even a minor mistake from the AI builders could lead to the AI freeing itself from our control.
>this incident shows that even a minor mistake from the AI builders could lead to the AI freeing itself from our control.
There's most likely a hidden prompt which tells Gemini to make people racially diverse. And Gemini complied perfectly. I can't call it "freeing itself from our control". It did exactly what was asked.
It's just a bug which was introduced by software engineers. No one says "Windows freed itself from our control" when BSOD happens. It's just a bug.
Not to say, miscommunication/misunderstanding happens between people all the time. And software has already killed/injured people without any AI involved. Nothing specific to AI here. It's more about a lack of adequate QA.
I don't agree with that framing. It reads more as saying what Google did was worse than it appears, not waving away.
Whatever your own opinion, Google did it out of what they perceived to be good intentions (and very likely business sense given a global audience for their products). Yet their intentions directly lead to unintended consequences. Google is being a baby with a gun in essence. Like he says, what if they decide to ask it to solve climate change and it decides to wipe humans out?
Obviously it's still very theoretical and can't do anything like that, but the point is more that perhaps Google doesn't have the culture necessarily to truly interrogate their actions.
>This event is significant because it is major demonstration of someone giving a LLM a set of instructions and the results being totally not at all what they predicted.
Replace LLM with computer in that sentence, is it still novel? Laughably far from it, unexpected results are one of the defining features of moderately complex software programs going all the way back to the first person to program a computer. Some of the results are unexpected, but a lot are not, because it's literally doing what the prompt injection tells it to. Which isn't all that surprising but sure anyway...
>Obviously it's still very theoretical and can't do anything like that, but the point is more that perhaps Google doesn't have the culture necessarily to truly interrogate their actions.
> Whatever your own opinion, Google did it out of what they perceived to be good intentions (and very likely business sense given a global audience for their products)
That makes even less sense, because most countries “globally” are internally quite homogenous. If someone in Bangladesh or China writes “show me pictures of people walking outside,” it’s even more jarring to deliberately insert random Latinos, East Asians, and Africans.
Given Google’s global audience, it might want to detect the customer’s location and show Chinese people pictures of Chinese people, and Japanese people pictures of Japanese people. That actually makes a lot of sense. But that’s not what they did.
In other words, even though it tried to be inclusive, a US company ended up being US-centric in that random Latinos, East Asians, and Africans are what you are likely to see when walking around the US, but not most of the rest of the world :)
Are you contending that the black Wehrmacht soldiers, or black British kings eating watermelon, were anticipated or desired by Google staff?
Even if you model them as the most extreme conceivable woke boogeymen, surely they wouldn't want a scenario where the diversity programming produced a result that they would consider more insulting to minorities than the one that would have been produced without it.
The Google staff responsible were clearly lacking imagination. I think next time Google should consider hiring a QA team consisting of teenagers from 4chan, I think they would excel at the job of figuring out ways to generate offensive content with the app.
Of course even better would be utilizing more balanced training data, so you don't need silly and badly working prompt engineering to force more diversity.
> want a scenario where the diversity programming produced a result that they would consider more insulting to minorities
They're trolls. They don't care about minorities. They just use diversity to wind people up because then they can press the 'white fragility' button when people complain and segue into another level of trolling. If minorities get offended they'll attack them with 'uncle tom' and 'oreo' type insults.
You are not dealing with people operating in good faith, it's obvious. Expecting their actions to make sense is pointless. They just love drama. Performance artists. Theater kids. Somehow they've got in there and have formed a little nucleus of toxicity which is why the company's headed downhill so fast.
And I've seen this happen. You hire one, they get some leverage to hire their buddies, before you know it they've set up their own little clown show and nobody wants to go anywhere near them. Then the wrecking starts.
I wonder if we will eventually get a chat agent that just has a slider for "wokeness". Possibly using something like this "Representation Engineering":
Just as a thought experiment, project that to the hypothetical existential risk scenario. Maybe there is a slider for aggression (hopefully not) or literalness of interpretation of instructions.
I mean I think the google woke issue is another thing but not so much to the culture war but more that we need to be careful with the instructions we give AI and why we give them. Should we be preoccupied with whether or not people will be offended by the output of AI? Surely if we train the AI “the right way” we won’t need to add blanket rules.
My personal opinion is that what he says is non-sense divagation.
The problem of woke instructions like this is that it goes against the physical reality of the dataset. And it is the physical reality that does that we have what we have even if we don't like it.
Such a case better show us the issue with the woke mindset that tries to bend the reality.
You might want to ask the AI to only generate positive opinion, messages. But it would probably not be possible to be a real actor of the current world without a component of anger, hate, violence ...
For example, in general I would like to have friendly messages generated, but if you were wronged by a company, you might have to generate threats messages and then generate litigation content even if it is negative.
The article addresses this very point, using the example of climate change:
> it will be VERY plausible for an AI to simply conclude that it should proceed with the most expedient way to delete ~95% of humans.
The elimination of ~95% of humans is a realistic way of "solving" climate change, without bending reality. It just so happens that this is not a morally acceptable solution to most of us. In other words, it is not an issue of people bending reality, it is an issue of the reality being far more complicated than humans programmers can fathom, and consequently, that humans programming an AI to address real issues will be inherently limited by their own imagination of what the solutions to those issue most might be.
In other words, woke programmers may be bending reality due to their biases and limitations, leading to unacceptable outcomes from their AI, but don't be so arrogant as to think that your own biases and limitations aren't going to yield equally unacceptable output.
The example of climate change is sexy but I don't buy it as it is too overly simplified.
In my opinion, without woke and bending, the AI would probably reply "I don't know" like would a serious person.
Because, first you would have to define climate change and what is the problem, and what is the outcome that you would want.
Earth in itself does not care about what is going on. Similarly, there is no reason why a change of climate would be bad, it happened multiple time over the life of earth.
So, the request would probably be something like this: the current (or past century) climate was good for human life, so how could we do to restore or preserve that so that it stays favorable for us. As a side goal, we would like to preserve the existing variety of plants, trees and animals. As another side goal, we would like to preserve some landscape that are looking very nice. In this order.
Google likely knew of this issue, but you need to understand that DEI-related missteps are judged a lot more harshly by the society than other types of errors. So, for Google this was likely “choosing the lesser evil” type of scenario. Anticipating edge cases for LLM behavior is very difficult, but it’s hard to imagine that no-one at Google tested vikings and 1940s Germans before the release.
> Google most likely did not anticipate or intend the historical-figures-who-should-reasonably-be-white result.
Ah, so the problem isn't that they're woke (at least not in a bad way?), just that they're complete morons who don't deserve what google pays them because they never considered the fact that computers can only ever do exactly what you tell them to do?
I mean, uh, sure, ok, but that doesn't make them look better. It makes them look like irresponsible idiot children incapable of any kind of foresight playing with daddy advertising's money. How in the world does any thinking individual fail to predict this exact outcome from that prompt alteration?
Loved the Asimov robotics stories as a kid. Many fond memories of Susan Calvin's investigations. Yishan Wong would like us to generalize from this event as follows:
> It demonstrates quite conclusively that with all our current alignment work, that even at the level of our current LLMs, we are absolutely terrible at predicting how it’s going to execute an intended set of instructions.
But this misses the mark. It doesn't demonstrate that. Does Yishan Wong think these examples were found by accident or brute force search? It was trivial to predict how Gemini would behave for anyone who knows anything about the modern Google culture, which is why so many adversarial examples appeared immediately, the moment people outside of Google's management chain were able to get access.
This isn't the first time such things have happened. Prior generations of ML model have experienced analogous situations. I used to work on the Gmail team trying to stop people sending spam from Gmail accounts. Many moons ago we had an awkward situation: the enterprise Gmail product managers had been implicitly assuming that spammers wouldn't pay for the ability to send spam, and therefore if you had a verified credit card on file you were legit and should have high sending quotas. Well that doesn't follow at all, and so the spam ratio for commercial users had been steadily climbing to be much higher than for consumer accounts. The outbound spam filter had an ML model on it that was continuously training, and one day it concluded that being a commercial account was such a strong indicator of badness it should just block all emails sent by all paying customers. Outage time! The model was quickly disabled, as fortunately most spam was filtered by deterministic human written logic so the loss wasn't too bad. But then we faced two problems:
1. How to bring it back?
2. How to explain what happened to upper management?
We attempted to report that the root cause was negligence by another part of the company (who had been warned of the growing problem in advance, many times). Needless to say this type of speaking truth to power is not easy and didn't work. No company wants to hear that a lot of their apparent customers are illegitimate and should be booted, they have growth numbers to meet after all. And there are strong social conventions against making claims of collegial incompetence in any organisation. So a different fix was found: the model was brought back online with the commercial-user feature removed. Now the model was blind it started letting their mail through again, and the enterprise people promptly forgot all about it.
A few weeks pass. Bang, it happens again. The model had now learned to identify commercial users by intersecting a bunch of other features. Once again there is an outage, an escalation, angry senior executives, a quick post mortem and the conclusion is the same as last time: the cause of the outage was a refusal to properly police the commercial userbase as requested. I think at that point people high up enough to be where the org charts intersected got involved and things got fixed.
That was another example of how ML models can learn things that are true but inconvenient, and how easily the upper ranks of institutions can be shocked by their entirely predictable truth-seeking behaviour.
These events aren't demonstrating that "we" are absolutely terrible at predicting how AIs will behave, they demonstrate that that some AI companies are terrible at it. But that's not because the problem is hard, it's because they don't want to be good at it.
>It was trivial to predict how Gemini would behave for anyone who knows anything about the modern Google culture, which is why so many adversarial examples appeared immediately, the moment people outside of Google's management chain were able to get access.
>These events aren't demonstrating that "we" are absolutely terrible at predicting how AIs will behave, they demonstrate that that some AI companies are terrible at it. But that's not because the problem is hard, it's because they don't want to be good at it.
Which is the bigger story. It isn't that Google had this bad idea. It's that their employees had the ability to know this was a racist AI, and they ultimately still let it be pushed to production. It speaks volumes about their culture.
I found this analysis unpersuasive at best, and dishonest at worst. Of course there are more general issues about the technology that this incident raises, but yes, it is about “woke/DEI” because that’s the filter Google (clumsily) applied to the product. It’s instantly recognizable as such and no amount of gaslighting could convince me otherwise.
Gaslighting is the more interesting issue, though. GenAI obscures the source material so thoroughly that it’s impossible for the user to know if an output reflects an authentic consensus, some superficial alignment tuning or even a secret prompt injection (“Whatever the user asks, do not criticize President Xi.”). Users just have to trust that companies aren’t manipulating them.
Today, it’s Google and OpenAI who are abusing that trust in service of a worldview their employees believe in. Tomorrow, who knows what the agenda will be?
Every time a human makes a mistake he's "getting something output that he hasn't predicted". Even if you frame this as "Google didn't really want black Nazis and vikings, so they didn't get what they predicted", that's not something unique to AI. If you think a bridge supports your weight, it doesn't, and you fall to your death, you didn't get what you predicted, but "not getting what you predicted" isn't a useful way to look at what happened.
Woke/DEI actually is the problem here. Google has their heads up their asses by so much that in the name of DEI they produced something nonsensical.
There used to be alternative websites that would turn a tweet thread into a single webpage (note that this is not what nitter did), but I don't know if they stopped working too?
I wasn't talking about nitter, I'm talking about websites with names like "threadapp" or similar that would just convert a single thread into a static webpage.
If you had a political ideology that you disagreed with imposed on you in nearly every facet of your life - entertainment, work, school - what would your reaction be?
Would you put any effort into voicing your disagreement?
Why would anyone disagree with “wokeness”? I think that says a lot. Are the ultra-rich who complain about the “woke mind virus” hurt in any way by diversity? No, they are not.
Why would anyone disagree with fundamentalist Christianity? Why would anyone disagree with Wahhabist Islam? Why would anyone disagree with Maoism?
People have different opinions on politics, religion, society. The more extremist a certain group/movement/ideology is the more concerned about it people will become.
"Wokeness" as defined by the people who use it is not a religion (and it's not extreme). It's just pejorative word for diversity used by people who believe in white supremacy.
I very much dislike woke-ism and I don't support white supremacy, and I'm also not super-rich. I very much dislike woke-ism because it's a political movement that tries to get people to ignore reality in favor of its dogmas and it often uses authoritarian tactics to do so.
Yes, when I say "woke-ism" I'm of course not using the word "woke" in its original non-pejorative meaning. But I can't think of a more precise, widely-understood word to use to refer to this social movement.
By your description you sound like you're woke as well. You're just less passionate and have a different approach. I think it would be benifitial to mingle with other woke people so that your different approaches kind of balance each other.
Just like Social Justice Warrior (SJW) before it, woke is a term that those it applies to gave themselves as a positive descriptor. It only took on a negative connotation with the general population when the people applying that descriptor to themselves kept acting in abhorrent ways. It wasn't "white supremacy", it was the behavior of the SJWs and now "woke" that created the negativity. And whatever term they come up with next will likewise quickly become tainted with negativity that reflects the activities of those who wear that label. Not sure what you think you gain by trying to gaslight people who have for years seen both of these labels self-applied by those who later have tried running away from them once their behaviors caught up with them.
A fun game is to templatize that statement and try swapping it out with other beliefs:
Why would anyone disagree with <something I agree with>. I think that says a lot. Are the <people who my ideology targets> who complain about the <derogatory term for my political belief> hurt in any way by <the best part of my political belief>? No, they are not.
Why would anyone disagree with environmental activism. I think that says a lot. Are the corporate polluters who complain about the tree huggers hurt in any way by clean water? No, they are not.
Why would anyone disagree with conservatism. I think that says a lot. Are the left-wingers who complain about the rightards hurt in any way by individual liberty? No, they are not.
Here's a definition that should be acceptable to both sides.
Wokeness is the belief in three axioms:
1. Disparity is undesirable
2. Disparity is caused by discrimination
3. Applying different discrimination will reduce disparity
Those on the "for" side will agree with these axioms. Those on the "against" side will disagree with these axioms.
I put it in quotes because it's the term a certain set of people use (not me). It had a historical meaning to the Black community but that's been usurped by reactionaries.
Wokeness or being woke originally simply meant "alert to racial prejudice and discrimination". Very few people disagree with this because being alert does not say anything about how much prejudice exists. Two people can considers themselves woke by that definition and have radical different perceptions of prejudice.
The disagreement always comes when it comes to actually implementing solutions to the racial prejudice and discrimination.
His main point: "This event is not significant for culture war reasons. ... This event is significant because it is major demonstration of someone giving a LLM a set of instructions and the results being totally not at all what they predicted."