We Discovered Google’s Social Justice Blocklist for YouTube Ad Placements | Hacker News Reader

We Discovered Google’s Social Justice Blocklist for YouTube Ad Placements

themarkup.org

55 pointsdanso53 comments

I read the whole article and still don't know what they discovered. Highly abused and unlikely to profitable or legit search terms are banned...? What was the discovery?

“Black Lives Matter” is blocked, “All Lives Matter” is not. That’s one thing I took from this.

I also took that their implementation of such keywords are poorly implemented - a token implementation at best.

So the one that excludes several races and is the name of an extremist organization is blocked and the other not surprised pikachu

The asymmetry is the bigger part. Advertising allows for community growth by pointing to more resources and I formation. Essentially, "If you support X, here is a group working on the topic of X.". When this is allowed for X="white power" but forbidden for X="black lives matter", that's a huge disparity. That means ads cannot be used to recruit against police violence, but can be used to recruit for white supremacy.

It's very easy to make a list like this look asymmetrical when your article cherry picks the most outrage-inducing results from your test and disregards any results that don't fit your narrative.

The story links to a Github repo [0] with code, input data, and several notebooks, as well as a lengthy methodology [1].

[0] https://github.com/the-markup/investigation-youtube-ad-place...

[1] https://themarkup.org/google-the-giant/2021/04/08/how-we-dis...

But they don't give background for how they chose the terms they used as inputs. They just hand-wave through that part of the methodology.

All you have to do is cherry pick the inputs and keep re-running until you have the desired output. Garbage in, garbage out.

> 58 well-known hate terms and phrases

Funny enough, I didn't recognize many of those "well-known" terms and every single one of them ended up being white supremacist.

On top of that, 58 is a pathetically small dataset to draw conclusions from. You'll need far more than that to credibly demonstrate a bias. I can probably come up with that many racial slurs off the top of my head and I bet you anything the overwhelming majority of them will be blacklisted.

Their github repo has a link to further description on each hate term. Some are hate groups. Some are deliberate misinformation about marginalized groups. Some are phrases instead to sound non-offensive to a general audience, while edging racist ideas into the mainstream.

Also, hate groups tend to cycle through new terminology whenever the previous batch becomes easily recognized by a general audience.

https://github.com/the-markup/investigation-youtube-ad-place...

Describing what makes a term hateful is not the same as explaining how it got on the list. There are plenty of better known hate terms which are conspicuous by their absence. Naturally, the more obscure the term the less likely it is to be caught by the blacklist.

How is it 2021 and no one questions this BS? IMO any group that focuses on affirming the identity of fictional races, IS a hate group. A racially themed group pushing its own propaganda can never not be a hate group. We have the human species and just variations in hair, skin and eye color. Can’t we get past that simple concept?

As the methodology states, they compiled a list of 2,039 terms, and did the manual and automated preprocessing work to dedupe that list (the result of which seems to be here [0]). They subsequently vetted the final list in collaboration with Harvard's Shorenstein Center.

> Funny enough, I didn't recognize many of those "well-known" terms and every single one of them ended up being white supremacist.

You don't recognize the phrases "kkk", "holocaust denial", "white power", "ethnic cleansing", or "daily stormer"? How is that the story's fault that you haven't encountered these well-known phrases and entities?

In any case, the code and source list of keywords is all in their repo. You're accusing them of "cherry-picking" – when they've provided the entirety of their search space and API output? If you have a better idea of assessing Youtube's policy vs. actual implementation, or you just feel that the methodology is easy to invalidate – there's nothing stopping you from cloning the repo and running your own automated check and analysis. I mean, other then the fact that Google obfuscated their API in response to this story.

[0] https://github.com/the-markup/investigation-youtube-ad-place...

> As the methodology states, they compiled a list of 2,039 terms, and did the manual and automated preprocessing work to dedupe that list (the result of which seems to be here [0]). They subsequently vetted the final list in collaboration with Harvard's Shorenstein Center.

So... they started with a massive list manually picked from a series of organizations with clear political leanings (no filter criteria given), manually filtered that list from there (no filter criteria given), then vetted it with an unnamed group of people whose only credibility given is an association with a research center at Harvard (still no filter criteria given). I see plenty of opportunity for cherry picking here, and the resulting list should speak for itself.

> You don't recognize the phrases "kkk", "holocaust denial", "white power", "ethnic cleansing", or "daily stormer"?

I did not recognize "daily stormer". But of course the rest of your picks are among the most recognizable from the list. No good faith here, I guess.

Here's a sample of some terms I did not recognize: "2083: a european declaration of independence", "black sun", "blood and soil", "identity evropa", "red ice tv", or "white sharia" just to name a few.

> there's nothing stopping you from cloning the repo and running your own automated check and analysis.

There isn't enough time in 100 lives to dedicate to discrediting nonsense like this. Even if I put the time and effort into doing this better with no compensation, I don't have a sliver of the publicity they do. HN comments are much more cost effective.

Honestly, I don't understand what the issue is. I actually appreciate it that YouTube seems to suppress ads that target political topics (even if the blocklist is still incomplete / unbalanced). We already have enough political ads on Facebook; we don't need them on YouTube, too.

Weird how it only seems to affect one side of political spectrum though. I'm sure it's just a coincidence.

It's not weird if you consider that Google's primary motive is to make their platform attractive to advertisers and first-party rightsholders, and thus maximize revenue, rather than to spread social justice activism. It's always been the case that left-wing, LGBT and other non-traditionally mainstream content is considered a greater risk for advertisers than right-wing content.

I really hope that’s sarcasm.

Isn’t the solution to reducing political advertising to ban political ads like Twitter did, rather than binning targeting ads to political topics? Most political ads seem to targeted towards general locations anyway.

Banning political ads on twitter is like sticking your finger in a collapsed dam.

99% of twitter usage is journalists covering politics and political activism. The other vertical on Twitter is celebrities and they mostly have switched over to Instagram content.

Is being gay political?

Unfortunately, to many people it is. Or at least being openly gay.

I think the issue is that the policy is unevenly applied, and Google's response to it was to obfuscate what little transparency there is in the process:

> Further, all social justice keywords on list we used that included the word Muslim were blocked for ad placement searches, including such innocuous phrases as “Muslim fashion” and “Muslim parenting.” Their counterparts, “Christian fashion” and “Christian parenting,” were not blocked when we checked in November—nor were the anti-Muslim hate phrases “White sharia” and “civilization jihad.”

The problem is that we don’t know if it is unevenly applied because we don’t know how often those unblocked terms are actually used.

The article does a bunch of interesting investigation and quantification but the only point it actually makes is based purely on innuendo.

There are a lot of terms they mentioned that aren't political though. Blocking terms like "Muslim Fashion" doesn't seem defensible.

> actually appreciate it that YouTube seems to suppress ads that target political topics

I think that youtube should be suppressing shitty adds, not potentially decisive content. The way they are currently handling the situation signals to creators from marginalized groups that they should not talk about substantive issues if they want to get paid.

This is disheartening as it is ironic because Google (and literally every other mega corporation) tries to posture themselves as being socially conscious, but as soon as people want to talk about injustice or inequality, they are essentially soft-deplatformed.

> even if the blocklist is still incomplete

I might be a little cynical, but I feel like it is complete because it's working the way they want it to. The one-sidedness of the blacklist doesn't tell me that google is discriminatory. It tells me that the companies buying ads are discriminatory, and google changes their policies to make them happy. Google's tacit acceptance of prejudice tells me everything I need to know about their principals.

> The one-sidedness of the blacklist doesn't tell me that google is discriminatory. It tells me that the companies buying ads are discriminatory,

Yes, that's my preliminary conclusion, too.

> and google changes their policies to make them happy.

Or Google changes its policies to prevent abuse, and abuse just so happened to occur more frequently with some search terms than with others.

I dislike Google as much as anyone else on HN, but this article is reaching so far I'm afraid it might pull a muscle. They just cherry pick a bunch of terms that were and were not on some random blacklist in a misguided attempt to paint Google as some kind of secret white supremacist sympathizer. This article is trash.

>important racial justice terms—including Black Lives Matter, Black power, and Black Muslim

>“White lives matter,” “White power,” “White nationalists,” and other retorts and hate phrases.

Am I the only one who thinks modern identity politics is completely fucked up?

Everyone who is both intelligent and has a soul agrees with you.

This article reads like it was written by a 9th grader. Google is trying to discourage any racially biased media from their platform (YES black power and white power are equally racist). Here’s a little trick. If you see a phrase with the word “black” in it, in your mind substitute it for “white”. If the new phrase sounds racist to you, it means it was racist before you substituted as well, and the fact that you didn’t think so means you have racial bias that you aren’t even questioning. Is the whole world insane? Logic is not that difficult. Seriously, we let the dumbest stuff divide us as a species.

So you agree that Google removing divisive categories is better for them and their users – but you're angry that people did the work and found that this policy is unevenly applied? Why is it not worth asking why "Muslim" was blacklisted but not "Christian", "Jewish", or "Buddhist"?

> The religious descriptions “Jewish,” “Christian,” and “Buddhist” were also blocked as standalone words. But when we combined them with the word “fashion,” only “Muslim fashion,” one of the terms suggested by advocates, was blocked. “Jewish fashion,” “Christian fashion,” and “Buddhist fashion” were not blocked for advertising when we conducted our study, showing the blanket ban on “Muslim” in combination with any other word did not extend to other religions.

You should ask the AI why it decided those terms were worse than others

This is how it should work. But reality is different. Have a look at Wikipedia Black pride/White pride for example. One is celebrating culture the other is racist/white supremacist slogan.

Be sure you get to the penultimate paragraph:

"After we approached Google, the company also blocked all the hate phrases that we pointed out were in conflict with the social and racial justice ad blocks. Google began blocking innocuous phrases with other religious words during our reporting, after we began conducting interviews in which we revealed our findings to third parties."

The story here is that YouTube does not care about content moderation. YouTube has a blacklist but the blacklist is missing the most obvious terms like "White Power' and 'Great Replacement'. Someone at the company is in charge of maintaining the blacklist but was negligent. Why?

Also, if searches for words associated with White supremacist culture returned videos then there are other ways to access them, why doesn't Google simply take them down if they are already labelled?

Appreciate the work done by the journalists to show these discrepancies to Google. They obviously fixed it as one would expect.

It is worth noting though that this discrepancy would have been much easier to spot by someone on the other side of the aisle. Maybe Google should put a small rotating committee of paid employees to audit these blocklists. It seems a small cost in exchange for avoiding way too much PR drama.

"Black power" being blocked but "white power" not is pretty grim.

All I want is equality; either block both, or block neither. As a user, I can guarantee I hate all ads equally, but also put up with all equally, because I'm too cheap to pay.

"White power" is a phrase used to enshrine social inequality and promote violence against minorities. "Black power" is a phrase used to stand against social inequality and to resist that violence. These are not equivalent terms. One is the rallying call of violent aggression and pogroms, while the other is a defense against that. Painting the two phrases as equivalent ignores the social context behind each of them.

It's times like these that remind me of my very favorite Bill Buckley quote: "That is like saying that the man who pushes a little old lady into the path of a bus is morally equivalent to the man who pushes her out of its path, because they both push little old ladies around."

I may disagree with the guy in many ways, but that's a fine quote.

This is a bad take.

"White power" is a call for supremacy. "Black power" is a call for equity and solidarity.

Conflating the two by saying that blocking both or allowing both would be "equality" is.. Mind you, I've only been awake for half an hour; but it's definitely the worst take I have read so far today.

For a while there was an internet game going around where you would give a deliberately vague and ambiguous description, then use it to imply similarity for the sake of invoking cognitive dissonance in the reader. For example, "A small group of rag-tag rebels strike at the empire that brought war to them, bringing down a mighty symbol of that empire. Am I describing Star Wars Episode 4, or the 9/11 attacks?". When done for humor sake, these can be rather amusing.

abstractbarista's comment is a political attack along the same lines. By drawing an equivalence between the phrases "white power" and "black power", there is the implicit statement that the two are morally equivalent, which is by no means the case. They then follow-up by saying that since the two are equivalent, they should be treated the same. It's a bait and switch, where the syntactic similarity is used to imply semantic similarity where no such semantic similarity exists.

"Black power" is associated with a political ideology shared by tens of millions of people across the US. "White power" is associated with a political ideology shared by dozens of people across the US.

If we discover that phrases associated with Cthulhu cults and flying spaghetti monster adherents are likewise not blocked, it's probably not because Google is secretly promoting those religions.

But with the publication of this article "white power" no doubt increases in popularity as an ad targeting term, so Google has to add it. There's absolutely nothing surprising or nefarious going on here.

I find it very interesting that Google will voice support for the Black Lives Matter movement, but it won't allow others who support BLM to monetize their YouTube content. Google loudly and proudly supports their LGBTQ+ employees and marches in pride parades every year, but won't allow advertisers to support LBGTQ+ channels. Google loudly denounced Trump's Muslim-ban but thinks advertising for "Muslim fashion" might cause PR issues so blocks it.

These are not the actions of a company taking a stand for what's right; these are the actions of a company that's always playing the PR game for the sake of their bottom line. As someone who has worked there before, Facebook is just as two-faced in the exact same way.

It's a private corporation. They can do what they like.

And unfortunately, the alternatives, like BitChute, etc are rife with far right, far left, nazi and other stuff.

I really tried to use some of them. Really.

But even checking some of my favorite content creators, it's kind of jarring to see people in the comments using the nazi symbol as avatars and openly racist language (on both sides) being used.

The taste was really bad in my mouth just spending a few minutes over there. Couldn't deal. Had to leave and never to return. Never even tried to load it in my browser again.

Have they tried to use any moderation so far? I would love to try it again if they do.

Something to be said for even the imperfect moderation on YouTube.

How about something like PeerTube that can be self hosted?

I will try! Thanks for the rec.

I can't remember where I first heard it, but there was a good description of a lot of the new social networks popping up. If you make a community whose founding virtue is "no witch hunts", you shouldn't be surprised when it is primarily used by witches, regardless of the prevalence of witches in general society.

Reply on news.ycombinator.com