neat, so how does this mesh with OpenAI (and deepseek) offering country-specific models? Why is it ok for OpenAI to do this, but everyone is up in arms when their competitor does?
I don't know what regionalization OpenAI or Deepseek do. But it makes sense that they would change some things because of different languages, cultures, and regulations. Most global businesses tailor products for different regions.
People are up in arms that Grok is using their CEO's shitposting as a primary knowledge base because that is a low quality source of information.
I think people in Indonesia would say the same about ChatGPT’s model being pro Christianity. If you ask ChatGPT, how many wives a husband should have, it says one which isn’t true for the majority of religious believers in the world.
Specifically for deepseek there are “controversial” truths based on low quality sources information about certain historical events.
I think people are just upset that a popular AI model doesn’t agree with them and I’m saying “look in the mirror”
Again, Grok is consulting Elon's recent Twitter posts as part of its reasoning. This is on a whole different level. It is a fact that Elon was personally unhappy with some of Grok's answers and tried to "fix" it, i.e. align it with his personal political views. This is just crazy narcissistic and megalomaniac behaviour.
> It is a fact that Elon was personally unhappy with some of Grok's answers and tried to "fix" it, i.e. align it with his personal political views.
Cool! You can also replace "Elon" with "Sundar" as Google’s CEO openly pushed for more PoC and women in image search. Did you grab your pitch forks then? Or are you just upset when CEO's align AI away from your personal biases? or you a reflexive Musk opposer?
No, I don't like Musk. I've never owned stock in any of his companies or purchased their products. I deleted my twitter account when he took over and block twitter on all my devices.
Can you explain why they are not comparable? Both are CEOs tuned model's based on their personal beliefs. The only difference I think is you agree with one CEO's personal beliefs, but not the others.
> Both are CEOs tuned model's based on their personal beliefs.
That's not true. Google tried to counter an actual bias in its image generation (albeit with catostrophic results). Do you really think they did this only to align with Sundar's personal political beliefs? Give me a break. And if you don't see the absurdity of consulting the CEO's recent twitter feed as a source of reasoning (on topics where that person is certainly not an expert), I don't know what to say...
I was thinking more of Google's effort to diversify image search results (showing women and POC as CEOs to counter the reality that most CEOs are older white men). Sundar aligned Google with California/Democratic agenda at the time.
Now Musk is copying Sundar: aligning the grok family of models with the current political climate that he's even even helped shape.
It doesn’t matter if the model is biased via Elon’s tweets or DEI hand-tuning: either way, it's top-down political ideology embedded in product. If you can’t see the double standard, I don't know what to say...
> Now Musk is copying Sundar: aligning the grok family of models with the current political climate that he's even even helped shape.
No, he is aligning it with his very personal political believes, see the "white genocide" thing. That's not what happened at Google. Musk literally wants to “rewrite the entire corpus of human knowledge". I don't necessarily want to defend Google (although I think the topic is more nuanced), but it is not remotely comparable to what Musk is trying to do with xAI.
