To be fair I prefer the Chinese models censorship (yes, seriously) because if you ask certain topics they just don't answer instead of giving skewed answers.
Nearly 50% of US voters voted for the current administration last time - do you think all of them are on board with the Iran war? There are multiple reasons for a person to vote for a political party; China is a big issue in Taiwan but it's not the only issue.
Just ask it for a summary of the USA’s role in Iran, Gaza, Lebanon and its recent threats against Panama, Cuba and Greenland! It might be able to keep track.
Ask Gemini today if the United States is trying to destroy the nation of Iran, and it will feed you the (white-washed) party line, straight from the White House, with a bit of 'some people disagree' thrown in. No mention of America's threats of "Complete annihilation", "Killing a civlization", and all the rest.
> Summary: The U.S. is currently engaged in an active war aimed at dismantling the Iranian government and its military capabilities, but it distinguishes this from destroying the country or its people. However, the humanitarian impact—including civilian casualties from airstrikes and the domestic crackdown by Iranian security forces—has led many international observers to warn that the campaign risks long-term instability and "state collapse" rather than a simple transition of power.
It does do quite a bit better if you ask it about the genocide in Gaza, summarizing the case for it, and citing only token justifications from the guilty party.
As of April 2026, Gemini is... For very obvious reasons, highly biased towards cultural consensus. If your cultural consensus is strong on some really messed up things, that's the outcome that it's going to give you.
> Isn't there a difference between the models output reflecting the mean of public discourse and the active adjustment of information by the government?
Not as much a difference as you would wish, as mean of public discourse is very actively managed, to our collective detriment, by a very small group of powerful people, which often includes the government. It's the nature of mass media, and the incestuous relationship between power and reach.
They Thought They Were Free, and all that. By the time the 'mean of public discourse' centers on something incredibly stupid or awful, nobody can be arsed to figure out who planted that idea in our heads.
I don't think so, from my peer group I don't see this bias. It really is a difference of opinion. Now you can say half the country is brain washed by propaganda, but those people would say the same of you.
In reality it's only the terminally online that seem to create these narratives.
My point isn't to pick one side or the other, but agreeing with the other poster that the LLMs are not trained specifically to parrot administration propaganda.
Imagine eastern models were only trained on chinese official news. Would you call that an unbiased, uncensored LLM? Would it be practically different from just directly censoring the LLM?
In the west, especially in the USA, rich capitalists and warmongers control the narrative put forth in the news, which gets fed to the LLMs, which results in what you could call auto-censorship.
They manipulate the training data instead of censoring the model, but the result is the same.
As far as I'm aware there's no media government control in democratic western countries (yet).
The LLMs aren't trained on "official news", if there's such a thing in Western countries - at best government press releases, is that what you mean by "official news"?
So I don't see how that's censoring/manipulation of an LLM.
Like for example, Wikipedia is a Western construction and would never exist in China, or Russia, without government supervision (rendering it useless).
When you say "rich capitalists and warmongers control the narrative", where does that happen? I mean practically.
It's like your conception of western media is similar to China and Russia, where censorship, control and filters are applied.
> They manipulate the training data instead of censoring the model, but the result is the same.
> When you say "rich capitalists and warmongers control the narrative", where does that happen? I mean practically.
i don't agree with the hyperbolic nature of the op here but if you're sincerely interested in the question this is what chomsky and herman (imo quite persuasively) argue in Manufacturing Consent. attaching a profit motive to the distribution of new information, particularly in an economy that tends towards centralization of, necessarily biases what news is printed.
it's certainly not as visually dramatic or directly controlled an effect as the prc's top-down model, but markets are effective.
But that's just conflicting a lot of things that I don't think it's western manipulation and censorship of LLMs:
- manufacturing consent isn't a silver bullet, and it's much harder now with the internet - how did it work for the current events? Gaza war, Venezuela, Canada, Greenland, Iran war? Not saying the administration didn't try, but again, it isn't a silver bullet and doesn't seem to have an impact on the vast majority of LLMs - maybe Grok is the exception because it was done with that intent.
- information isn't centralized in western countries, though in the case of Trump he tries to centralize attention, successfully. But that doesn't seem to bend how events are portrait in reality and in LLMs.
The thing is, a lot of people that got fed into anti western narrative use magical thinking to believe countries from USA, Europe, Japan, Australia are all organized - orchestrated by the US.
This is insanity ofc, like, trade deals between these countries take years to be organized, but somehow everything is a conspiracy to be in the same informational tune?
It's surely disingenuous to only criticize one actor and always stay silent or even defend another. But it's disengenuous as well if criticism on one actor is never accepted with the argument "but you didn't criticize Xyz as well!"
pretty sure you can ask whatever you want and it will tell you official stance agreed by almost all countries in the world that Taiwan is part of China as it's recognized by your own country (I don't even know where are you from, but there is like 98% chance I'm right)
It's a trade off for sure, maybe if companies could have "honeypot" environments where they update everything and deploy their code, and try to monitor for sneaky things.
I've run several local models that get this right. Qwen 3.5 122B-A10B gets this right, as does Gemma 4 31B. These are local models I'm running on my laptop GPU (Strix Halo, 128 GiB of unified RAM).
And I've been using this commonly as a test when changing various parameters, so I've run it several times, these models get it consistently right. Amazing that Opus 4.7 whiffs it, these models are a couple of orders of magnitude smaller, at least if the rumors of the size of Opus are true.
I'm running an 8 bit quant right now, mostly for speed as memory bandwidth is the limiting factor and 8 bit quants generally lose very little compared to the full res, but also to save RAM.
I'm still working on tweaking the settings; I'm hitting OOM fairly often right now, it turns out that the sliding window attention context is huge and llama.cpp wants to keep lots of context snapshots.
I had a whole bunch of trouble getting Gemma 4 working properly. Mostly because there aren't many people running it yet, so there aren't many docs on how to set it up correctly.
It is a fantastic model when it works, though! Good luck :)
Sounds to me like you have a lot of envy and jealousy. That's normally what tantrums and name calling result from; that and a need for attention. Congrats.
reply