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 technology.
"Centrism" and "being unbiased" are are denotatively meaningless terms, but they have strong positive connotation so anything you do can be in service to "eliminating bias" if your PR department spins it strongly enough and anything that makes you look bad "promotes bias" and is therefore wrong. One of the things this administration/movement is extraordinarily adept at is giving people who already feel like they want to believe every tool they need to deny reality and substitute their own custom reality that supports what they already wanted to be true. Being able to say "That's just fake news. Everyone is biased." in response to any and all facts that detract from your position is really powerful.
It's far more likely that an employee injected malicious code, exactly as said. Elon's become a divisive figure in a country filled with lots of crazy people, to the point of there been relatively widescale acts of criminality, just to try to spite him. Somebody trying to screw over the company seems far more believable than Elon deciding to effectively break Grok to rant about things in wholly inappropriate contexts.
Didn't this guy hit the salute in front of the entire world? To me it seems very likely that he would inject a racist prompt. Far more likely than a random hacker doing so to discredit him.
If that were the case, Musk absolutely would have shared the details of who this person was, why they hate freedom so much, how they got radicalized by the woke mind virus, etc.
First, I think the fact that grok basically refused to comply with those hamfisted instructions is a positive signal in the whole mess. How do you know other models are just as heavily skewed but just less open about them? The real alignment issue today is not about AGI, but about hidden biases.
Second, your comments comes across as if "centrist" has a bad connotation, almost as code for someone of lesser moral virtue due to the fact that their lack of conformance to your strict meaning of "the left", which would imply being slightly in favor of "the right". A "desire", as you called it, perhaps arising from uncivilized impulse rather than purposeful choice.
In reality, politics is more of a field than a single dimension, and people may very well have their reasons to reject both "the left" and "the right" without being morally bankrupt.
Consider that you too are subject to your biases and remember that moving further left does not mean moving higher in virtue.
It's difficult to make the claim that the AI not complying with a racist prompt is a positive signal for the organisation that wrote the racist prompt.
> Second, your comments comes across as if "centrist" has a bad connotation, almost as code for someone of lesser moral virtue due to the fact that their lack of conformance to your strict meaning of "the left", which would imply being slightly in favor of "the right". A "desire", as you called it, perhaps arising from uncivilized impulse rather than purposeful choice.
Centrism and compromise are the enemies of extremists.
Centrism is also the ultimate defense of the status-quo, meaning you have a bias towards the status-quo.
The fallacy here is that the status-quo is reasonable therefore being a centrist is reasonable and being a not-centrist is unreasonable.
Just because the status-quo is the status-quo and is in the "middle" does not make it reasonable. For example, the status-quo in Israel right now is performing a genocide. The centrists in Israeli politics are pro-genocide. The "extremists", as you say, are anti-genocide.
The current political landscape of the US is far-right. Where does that leave centrists? This is up to you to dissect.
The current political landscape of the US is not far-right. The current government may be, but everything in life is cyclical.
Democrats in 2024 lost more votes relative to 2020 than Republicans gained between the two elections. Which is why some people say Kamala "lost to the couch"--which is a comforting but myopic take because losing to the couch means your arguments are less convincing than those of the other party
> First, I think the fact that grok basically refused to comply with those hamfisted instructions is a positive signal in the whole mess.
I mean, _maybe_ about LLMs in general, in an abstract sense, if you're deeply concerned with LLM alignment. But not about grok, because it's an otherwise fairly generic LLM that is run by a company _so incompetent that it made said hamfisted instructions, or allowed them to be made_. Like, even beyond the ethics, the whole episode (and the subsequent holocaust-denial one) speaks to a totally broken organisation.
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.
LLMs can't truth seek. They simply do not have that capability as they have no ability to directly observe the real world. They must rely on what they are told, and to them the "truth" is the thing they are told most often. I think you would agree this is a very bad truth algorithm. This is much the same as I have no ability (without great inconvenience) to directly observe the situation in SA. This means I am stuck in the same position as an LLM. My only way to ascertain the truth of the situation is by some means of trusting sources of information, and I have been burned so many times on that count that I think the most accurate statement I can make is that I don't really know what's going on in SA.
Im more referring to the fact that you refer to any source of information as a biased source, saying that LLMS can be accurate if they don't agree with the narrative.
One good reason is because you have no logical reason to think it did. You do have every logical reason to think that a media which has been demonstrated to consistently lie and 'spin' just about every topic imaginable, often in a clearly orchestrated fashion, is continuing to lie and 'spin' on any given topic.
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 technology.