41–50 of 241 posts
The desire to be "centrist" on HN is perplexing to me. The fact that Elon, a white south african, made his AI go crazy by adding some text about "white genocide", is factual and should be taken into consideration if you want to have an honest discussion about ethics in tech. Pretending like you can't evaluate the technology politically because it's "biased" is just a separate bias, one in defence of whoever controls…
The desire to be "centrist" on HN is perplexing to me. The fact that Elon, a white south african, made his AI go crazy by adding some text about "white genocide", is factual and should be taken into consideration if you want to have an honest discussion about ethics in tech. Pretending like you can't evaluate the technology politically because it's "biased" is just a separate bias, one in defence of whoever controls…
The desire to be "centrist" on HN is perplexing to me. The fact that Elon, a white south african, made his AI go crazy by adding some text about "white genocide", is factual and should be taken into consideration if you want to have an honest discussion about ethics in tech. Pretending like you can't evaluate the technology politically because it's "biased" is just a separate bias, one in defence of whoever controls…
Aren't you just evaluating these claims based on things you've heard from biased sources (which is all of them) too? How do you know that your biased perspective is any more correct than Grok's bias?
I can't think of a less trustworthy group of people on model alignment. They claimed that they had a rogue actor who deployed their 'white genocide' prompt, but that either means they have zero technical controls in their release pipeline (unforgivable at their scale) or they are lying (unforgivable given their level of responsibility). The prompt issue is a canary in the coal mine, it signals that they will absolute…
Earlier quoted context omitted.
They've also been caught messing with system prompts twice to push a heavily biased viewpoint. Once to censor criticism of the current US administration and again to push the South Africa white genocide theory contrary to evidence. Not that other AI providers are necessary clean in putting their finger on the scale, but the blatant manner in which they're trying to bias Grok away from an evidence-based position erode…
Has any AI company not been caught doing this? Grok is just doing it in the opposite direction. I hate it too, but let's not pretend we don't know what's going on here.
As someone developing agents using LLMs on various platform, im very reluctant to use anything associated with xAI. Grok's training data is increasingly pulled from an increasingly toxic source. Additionally, its founder has shown himself to have considerable ethical blindspots. Ive got enough second-order effects to be wary of. I cannot risk using technology with ethical concerns surrounding it as the foundation of…
They've also been caught messing with system prompts twice to push a heavily biased viewpoint. Once to censor criticism of the current US administration and again to push the South Africa white genocide theory contrary to evidence. Not that other AI providers are necessary clean in putting their finger on the scale, but the blatant manner in which they're trying to bias Grok away from an evidence-based position erode…
I mean really, people don't want that crap turning up in their responses. Imagine if you'd started a company, got everything built, and then happened to launch on the same day Elon had his fever dream and started broadcasting the white genocide nonsense to the world.
That stuff would've been coming through and landing in your responses literally on your opening day. You can't operate in a climate of that much uncertainty. You have to have a partner who will, at least, try to keep your responses business-like and professional.
The desire to be "centrist" on HN is perplexing to me. The fact that Elon, a white south african, made his AI go crazy by adding some text about "white genocide", is factual and should be taken into consideration if you want to have an honest discussion about ethics in tech. Pretending like you can't evaluate the technology politically because it's "biased" is just a separate bias, one in defence of whoever controls…
Aren't you just evaluating these claims based on things you've heard from biased sources (which is all of them) too? How do you know that your biased perspective is any more correct than Grok's bias?
Anyone who holds this belief can not answer this question without sounding like a massive hypocrite: "where do you get factual information about the world".
Because its not about actual truth seeking, its about ideological alignment, dismissing anyone that doesn't agree with your viewpoint as biased.
Earlier quoted context omitted.
Has any AI company not been caught doing this? Grok is just doing it in the opposite direction. I hate it too, but let's not pretend we don't know what's going on here.
I think conflating what other companies have been doing with what Grok is doing is disingenuous personally. Most other AI stuff has had banal "brand safety" style guards baked in. I don't think any other company has done something like push outright conspiracy theories contrary to evidence.
Earlier quoted context omitted.
It's honestly one of the better ones I've tried for general questions. I saw it used in a blind competition against ChatGPT, Claude, and Gemini, and amongst people who didn't use LLMs frequently, it was the most favored for 4/5 questions! It's very good at sounding much more natural and less robotic than the others, imo.
Was it more correct or useful in its output or do you mean it nailed a desirable conversational tone like a pleasantly rendered lorem ipsum.
they conduct blind trials were users submit a prompt, and vote on "best answer".
grok holds a very good position in its leaderboard.
I can't think of a less trustworthy group of people on model alignment. They claimed that they had a rogue actor who deployed their 'white genocide' prompt, but that either means they have zero technical controls in their release pipeline (unforgivable at their scale) or they are lying (unforgivable given their level of responsibility). The prompt issue is a canary in the coal mine, it signals that they will absolute…
I reckon there is exactly one person at xAI who gives even remotely enough of a fuck about South Africa's domestic issues to put that string into the system prompt. We all know who it is.