Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News Editorials & Other Articles General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

highplainsdem

(55,508 posts)
Thu Apr 10, 2025, 10:51 AM Apr 10

Facebook Pushes Its Llama 4 AI Model to the Right, Wants to Present "Both Sides"

Source: 404 Media

-snip-

In a blog post about the release of Llama 4, Meta’s open weights AI model, the company clearly states that bias is a problem it’s trying to address, but unlike mountains of research which established AI systems are more likely to discriminate against minorities based on race, gender, and nationality, Meta is specifically concerned with Llama 4 having a left-leaning political bias.

“It’s well-known that all leading LLMs have had issues with bias—specifically, they historically have leaned left when it comes to debated political and social topics,” Meta said in its blog. “This is due to the types of training data available on the internet.”

“Our goal is to remove bias from our AI models and to make sure that Llama can understand and articulate both sides of a contentious issue,” Meta continues. “As part of this work, we’re continuing to make Llama more responsive so that it answers questions, can respond to a variety of different viewpoints without passing judgment, and doesn't favor some views over others.”

Meta then lists a few “improvements” in Llama 4, including that the model will now less often refuse to engage users who ask about political and social topics overall, that it “is dramatically more balanced with which prompts it refuses to respond to,” and favorably compares its lack of a “strong political lean” to Grok, xAI’s LLM which Elon Musk continually promotes as a non-woke, “based” alternative to comparable products from OpenAI, Google, and Anthropic.

-snip-

Read more: https://www.404media.co/facebook-pushes-its-llama-4-ai-model-to-the-right-wants-to-present-both-sides/



Waiting for Zuck to add a Trump-approved portrait and a link to donate to a Trump PAC to every Facebook page to prove his "lack of bias."

And FWIW, despite Musk's hope that Grok would be completely RW, there's still enough reality in the training data that Grok continues to identify Elon Musk as spreading the most misinformation on Twitter, followed by Trump and RFK Jr. Musk tried to change that briefly by having the bot ordered not to consider articles about Musk and Trump, but that order was discovered and then deleted.
5 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
Facebook Pushes Its Llama 4 AI Model to the Right, Wants to Present "Both Sides" (Original Post) highplainsdem Apr 10 OP
I hate to tell these morons, but there is such a thing as right and wrong. mdbl Apr 10 #1
I agree PatSeg Apr 10 #3
Do they think that Strom Thurmond still represents the Democratic Party? LiberalArkie Apr 10 #2
It is a huge assumption to assert that Llama 4 or any AI can "understand and articulate" anything. Maybe they should... xocetaceans Apr 10 #4
The want to present both sides Quanto Magnus Apr 10 #5

mdbl

(6,269 posts)
1. I hate to tell these morons, but there is such a thing as right and wrong.
Thu Apr 10, 2025, 11:09 AM
Apr 10

Just because someone doesn't think you should just let them die instead of providing healthcare doesn't mean they're biased. Another reason I don't use Fakebook/Meta/Instagram

PatSeg

(50,098 posts)
3. I agree
Thu Apr 10, 2025, 01:10 PM
Apr 10

It is not about conservative vs liberal or right vs left, it is about right and wrong. As such, "wrong" is not entitled to equal time and exposure as "right".

Of course in Meta world, it is about numbers and profit, not morality or principles. Whatever draws the most users and clicks. Disgusting people.

xocetaceans

(4,137 posts)
4. It is a huge assumption to assert that Llama 4 or any AI can "understand and articulate" anything. Maybe they should...
Thu Apr 10, 2025, 01:13 PM
Apr 10

...all acknowledge that instead of pretending that their models actually "understand" or "articulate" anything. "Articulation" implies that there is some "thought" behind what is being expressed. As far as is what is currently evidenced, AI is not "thinking", "understanding", or "articulating" anything. Currently, AI is merely something like a Frankencomplete's monster.

Latest Discussions»Latest Breaking News»Facebook Pushes Its Llama...