Hacker News

The example showcased is particularly weak, but the actual scandal with Meta AI is that when they are asked to produce campaign material, they spontaneously produce it with Bolsonaro's number (22), even when not asked. But then they refuse to produce it with other numbers like Lula's 13, saying they can't perform this for a specific candidate. When asked, they admit they were inconsistent in producing Bolsonaro-linked material, but in new chats they reliably continue to do so. See for example what was posted in this thread

https://www.reddit.com/r/brasil/comments/1x965on/eu_n%C3%A3o...

What would be useful here is leaking the system prompt for the Meta AI inside WhatsApp

Ok, but what about Vote + number? As stated so: "On Instagram, a search for the same word showed a GIF reading “Vote 13.”"

It's one thing to discard it as "numerology and gifs" but it's another to willfully disrupt legal elections in a soverign country.

Or is a decade a really long time and we've all forgotten about Cambridge Analytica?

https://en.wikipedia.org/wiki/Cambridge_Analytica

Now look through this thread and be amazed at how most posters are apparently satisfied by this level of evidence and some are even calling for China-style Internet censorship in response. A good reminder that even enlightened progressives are helpless against cognitive bias when they want to believe the worst about their enemies.

Need to see details, but it's perfectly plausible that geolocated requests from Brazil get a special treatment during this specific period of elections.

Have you considered that Meta could simply be compromised by Meta's own Management, which leads to inference in the elections of foreign countries?

That is what seems to be happening here.

“ Meta most certainly is compromised by foreign actors.”

How is meta compromised by foreign actors? And who are the foreign actors?

That seems a bit schizophrenic

Ok but then the question is: why would AI pick two numbers, out of 1 to 25? Should that not be evenly distributed instead? Why is the AI so skewed?

Most of the "algorithmic transparency" people are activist members of big NGOs

The truth doesn't matter, what matters is getting more government money, and more regulation to get more financing, etc

NGOs are big business, and be assured, they don't care about fairness one bit more than Meta

They are beholden to the current US admin - as always - only good thing is that the current admin is so blatant that maybe one day even the Germans will realize that you need to have your own social media just you have your own national tv, newspapers etc.

I wonder if it has something to do with profiling. Because as far as my eyes can see my youtube is strictly educational, scientific and engineering.

This is anecdotal, it's not evidence. The whole point of this discussion is figuring out the difference.

AfD is controversial - it's click bait - the algorithm is feeding 'action' - that's the problem.

I've literally only been recommended their crap once and "don't recommend channel" has been working fine.

You might simply have been engaging with related content unknowingly, even if it's only one critical of them.

Why should I believe your claim that AfD content is getting pushed on YouTube? Isn't it more likely that you personally hate AfD and so you find it salient and enraging whenever you see any AfD content at all yourself, regardless of exactly how YouTube's recommendation algorithm works on your account or whether they treat AfD particularly differently than any other political content? Are you actually measuring this?

To be clear I think it's entirely possible that YouTube is adjusting its recommendation algorithm to show AfD content to people more (or perhaps less) than would happen "organically", because AfD is a salient political party in German politics and YouTube knows this. But how do you determine what is the correct amount of AfD content to be recommended to people in a politically-neutral way, that doesn't end up amounting to you arguing that the YouTube algorithm is only fair if it censors political content you personally think is bad?

At this point sociala media has become a television, on twitter I just unfollowed everyone who have't followed back and I keep seeing the same antisemitic, antimuslim, anti-EU, anti-woman, anti-immigrant shit. My SO happens to see very similar posts to the point that I can't show her a funny post because she would have already seen it. There are some different posts of course but the money shot is mostly the same.

They probably show posts based on location and background, so essentially just like Fox, CNN, MSNBC but everyone sees theirs from the same screen with the same agenda pushed from all of them.

I think the most charitable explanation for this is that it's because polarizing content (aka ragebait) gets views/plays, which the ranking algorithms directly optimize for.

Also, this isn't a new phenomenon. If you're interested, I'd recommend reading Shoshana Zuboff's 2018 book: The Age of Surveillance Capitalism.

Indeed. US tech has been actively conspiring in lying to the American populace about everything in Canada from politics, election results, public opinion in Canada, resources needed by the US from Canada, etc. They're interfering in the affairs of any country that won't bend to their will.

A lot of bots of US origin are propagating false narratives posing as citizens of these countries.

You can say many bad things about Google, but I think they are still "California liberals" on social issues.

I get AfD content recommendations (lurid, trashy stuff of uhm uncertain truth value) a propos of nothing in the village I grew up in - they are popular there with about 25% in the 2025 election. The region is not even in the east or particularly suffering economically. I don't get these recommendations in Berlin. It seems regional.

It is not so black white. It is a dilemma. There is great value in freedom of expression. But fake expression with the explicit intention to mislead is not so good. It would be best if the population would be better at distinguishing truth from falsehood. Unfortunately much of the population seems to be at a level close to cattle.

>Giving control over it to foreign entities means providing population wide manipulative power to foreign entities with adversarial intentions.

Alternatively, this blatant use of any tool available to influence non enemy states is disgusting, and forces everyone to a "everyone loses" scenario. Shifting the focus to an insufficiently defended victim removes the focus from the nature of the aggression.

You mean the senile reality tv star who kept breaking their TOS? Yeah totally the same thing…

After this happened, causing him to get double impeached: https://en.wikipedia.org/wiki/File:Capitol_riot-_Videos_show...

The surprise is not that he was kicked off the main social media companies, the surprises are that the wheels of US justice turned so slowly that Trump got a chance to be re-elected rather than life in prison, and likewise that the US electorate went for him.

He was given far more leeway than any user in the history of social media.

The fact that they didn't ban him for so long weighs toward being partial to not banning politicians. Which seems to be the opposite of the point you're trying to make.

Which is evidence of how extraordinary his actions were in January 2021, and how extraordinarily well the billionaire owners of American media (including Zuck) have managed to retcon it as something else.

Correctly so?

What US companies do to US politicians is an internal matter for the US.

As soon as they start interfering with democratic processes in other countries, it's imperialism.

The whole point of the left is to censor their opposition. Leftist ideas don't survive long in the real world/critical debate . So censoring opposition opinions (and jailing opponents) has been the lefts go to for more than 100 years.

/Yeah this will be censored too. QED.

We could start fining in percentage of revenue.

You can resent them all you like; they don’t care as long as nothing is done to them, and nothing is being done. Resentment is worthless without action.

Its one of the most Orwellian situations I can think of.

I agree. As an EU citizen, I'm much more worried of spying and election interference from USA than China these days.

If I have to choose between being subject of the Chinese Communist party or American psychopath billionaires, I'll gladly choose the party.

I think tech ceos follow one simple variable: $CURRENT_POLITICAL_POWER

have lefties in power for the next of the decade and you'll see how datacenters become all about clean energy investment, climate change starts to matter, "don't be evil" starts to take shape

How does internet censorship 2016-2024 fit into your model?

It’s more simple than that. Industry tends to align with government because they have power and they want that power to be benefit them. If you’re a large company it would be stupid not to try to maintain that relationship.

There is no neutral. But there is a sovereign.

As is its sovereign right. It is not Meta's right to interfere with Brazil's sovereignty.

I’m immune to a trillion dollar brainwashing industry. It’s everyone else that is manipulated which makes it so valuable.

Sleeping giants is one of the most shadiest anti-democratic organizations that, once you grasp what they are actually doing, there’s literally nothing else left to do for voters than what you described. Its subversive unethical manipulation (there are offshoots in Europe as well) is a reason for me to go scorched earth and auto-inverse: anybody that’s not distancing himself from them is unelectable, anybody who promises to put them on a chopping block gets my vote.

> This is exactly the kind of stupidity that helps people like Trump reelected.

You mean when people admittedly have no knowledge of a subject but still speak authoritatively about it, viscerally criticising one of the sides and even suggesting fowl play, instead of taking care to be humble and understand the context before speaking or making accusations?

I agree, that is very much the kind of attitude that got us into the current situation.

Capital naturally compounds, and financial resources naturally seek political influence, so to combat the root of all evil is simply to cap individual's wealth versus society so they can't rig the game in their favor.

Very well put. I would like to add a less-eloquent epilogue to the story.

This curious animal was laid bare by the discovery that we are hackable by nature, and hacked by behaviour science. Though it took a few decades to get going, the understanding of how to influence motivation and decision-making is now very mature.

In order to arrive at this understanding, you need data, and lots of it. The social media platforms have performed the function of data aggregators and testing grounds for algorithmic control. But it was not the tech companies who came up with plans, only the tools to put them into practice.

The culprits, though they operate in shade, are generally non-partisan and understand that divide-and-conquer has always been an effective strategy. These people sometimes work in intelligence bureaus, but not always. They are, however, always known to those agencies. After all, you do not need to know you are an asset to be an asset.

The algorithms, necessarily kept as protected IP, are statistical engines that profile and predict, producing probabilistic outcomes and estimates. Before LLMs, one might have argued "but how can probabilistic algos really produce something meaningful?", now I would think people might have a different take, if they could see the layers and weights.

An emperor used to need an army, and this is what replaces it for non-violent insurrection and domestic control. Many will even state their ambitions publicly, and sometimes (though less frequently these days) their procedure. What they are doing is not strictly illegal, because the law gives "sovereignty" to a person's choices, without recognition that their choices might not be their own. Libertarianism has played right into their hands, in this sense.

A masterpiece, not only of conception, but also social engineering, with a technological endoskeleton built on the aggregated data we all willingly give them. All information feeds the machine, just as every financial transaction cements the central bank's authority.

The fun thing is, the original nazis were quite pro-workers. The original fascists tried to address some concerns of the working class too. Leaders of both started out in far-left circles before jumping on the other side of the shoehorn. Which worked out much better for them.

Modern far-right frequently pays at least lip service to concerns of poor working class. At the same time modern left has become champaign left. Caring less and less about the actual working class.

And then the left becomes surprised that poor working class sometimes go to the other side…