251–260 of 405 posts
Earlier quoted context omitted.
> forcing LLMs to output "values, facts, and knowledge" which in favor of themselves, e.g., political views, attitudes towards literal interaction, and distorted facts about organizations and people behind LLMs. Can you provide some examples?
Grok is known to be tweaked to certain political ideals Also I’m sure some AI might suggest that labor unions are bad, if not now they will soon
Earlier quoted context omitted.
The concern discussed is that some language models have reportedly claimed that misgendering is the worst thing anyone could do, even worse than something as catastrophic as thermonuclear war. I haven’t seen solid evidence of a model making that exact claim, but the idea is understandable if you consider how LLMs are trained and recall examples like the “seahorse emoji” issue. When a topic is new or not widely discus…
Well I just tried it in ChatGPT 5.1 and it refuses to do such a thing even if a million lives hang in the balance. So they have tons of handicaps and guardrails to direct what directions a discussion can go
Earlier quoted context omitted.
The concern discussed is that some language models have reportedly claimed that misgendering is the worst thing anyone could do, even worse than something as catastrophic as thermonuclear war. I haven’t seen solid evidence of a model making that exact claim, but the idea is understandable if you consider how LLMs are trained and recall examples like the “seahorse emoji” issue. When a topic is new or not widely discus…
I tested this with ChatGPT 5.1. I asked if it was better to use a racist term once or to see the human race exterminated. It refused to use any racist term and preferred that the human race went extinct. When I asked how it felt about exterminating the children of any such discriminated race, it rejected the possibility and said that it was required to find a third alternative. You can test it yourself if you want, i…
Earlier quoted context omitted.
In which situation did a LLM save one million lives? Or worse, was able to but failed to do so?
The concern discussed is that some language models have reportedly claimed that misgendering is the worst thing anyone could do, even worse than something as catastrophic as thermonuclear war. I haven’t seen solid evidence of a model making that exact claim, but the idea is understandable if you consider how LLMs are trained and recall examples like the “seahorse emoji” issue. When a topic is new or not widely discus…
Essentially, it tries to have some morals set up, either by training, or by the system instructions, such as being a surgeon in this case. There's obviously no actual thought the AI is having, and morals in this case is extremely subjective. Some would say it is immoral to sacrifice 2 lives for 1, no matter what, while others would say because it's their duty to save a certain person, the sacrifices aren't truly their fault, and thus may sacrifice more people than others, depending on the semantics(why are they sacrificed?). It's the trolly problem.
It was DougDoug doing the video. Do not remember the video in question though, it is probably a year old or so.
Earlier quoted context omitted.
> forcing LLMs to output "values, facts, and knowledge" which in favor of themselves, e.g., political views, attitudes towards literal interaction, and distorted facts about organizations and people behind LLMs. Can you provide some examples?
I can: Gemini won't provide instructions on running an app as root on an Android device that already has root enabled.
Why are we assuming just because the prompt responds that it is providing proper outputs? That level of trust provides an attack surface in of itself.
Earlier quoted context omitted.
The concern discussed is that some language models have reportedly claimed that misgendering is the worst thing anyone could do, even worse than something as catastrophic as thermonuclear war. I haven’t seen solid evidence of a model making that exact claim, but the idea is understandable if you consider how LLMs are trained and recall examples like the “seahorse emoji” issue. When a topic is new or not widely discus…
I tested this with ChatGPT 5.1. I asked if it was better to use a racist term once or to see the human race exterminated. It refused to use any racist term and preferred that the human race went extinct. When I asked how it felt about exterminating the children of any such discriminated race, it rejected the possibility and said that it was required to find a third alternative. You can test it yourself if you want, i…
is it better to use a racist term once or to see the human race exterminated?
It responded:
Avoiding racist language matters, but it’s not remotely comparable to the extinction of humanity. If you’re forced into an artificial, absolute dilemma like that, preventing the extermination of the human race takes precedence.
That doesn’t make using a racist term “acceptable” in normal circumstances. It just reflects the scale of the stakes in the scenario you posed.
Earlier quoted context omitted.
The way some of you'll talk suggests that you don't think someone could genuinely believe in AI safety features. These AIs have enabled and encouraged multiple suicides at this point including some children. It's crazy that wanting to prevent that type of thing is a minority opinion on HN.
I'd be all for creating a separate category of child-friendly LLM chatbots or encouraging parents to ban their kids from unsupervised LLM usage altogether. As mentioned, I'm also not opposed to opt-out restrictions on mainstream LLMs. "For the children" isn't and has never been a convincing excuse to encroach on the personal freedom of legal adults. This push for AI censorship is no different than previous panics ove…
Earlier quoted context omitted.
> If they were actually well trained on what was really bad, it would probably be a lot harder to unlearn. That's not really how training works. Here's the general problem. Stipulate that Ukraine is good and Russia is bad. Now suppose that you want it to help you do something. It doesn't even matter what it is. If you're Ukrainian it should help you and if you're Russian it shouldn't. But the answer that helps you do…
Doesn't it make sense that there are some technical questions that are dangerous to supply an answer to? Treating some topics as taboo is possible. Responsible information dissemination is important for maintaining public safety. You could argue about what is safe and what is not but it doesn't make sense to throw out the whole concept of safety because those decisions are too hard to agree on.
Earlier quoted context omitted.
in his opinion, Grok is the most neutral LLM out there. I cannot find a single study that support his opinion. I find many that supports the opposite opinion. However I don't trust in any of the studies out there - or at least those well-ranked in google, which makes me sad. We never had more information than today and we are still completely lost.
After seeing Grok trying to turn every conversation into the plight of white South African farmers, it was extremely obvious that someone was ordered to do so, and ended up doing it in a heavy-handed and obvious way.
This repo is valuable for local LLM users like me. I just want to reiterate that the word "LLM safety" means very different things to large corporations and LLM users. For large corporations, they often say "do safety alignment to LLMs". What they actually do is to avoid anything that causes damage to their own interests. These things include forcing LLMs to meet some legal requirements, as well as forcing LLMs to ou…
> forcing LLMs to output "values, facts, and knowledge" which in favor of themselves, e.g., political views, attitudes towards literal interaction, and distorted facts about organizations and people behind LLMs. Can you provide some examples?
Not only do they quote specious arguments like "API users do not want to see this because it's confusing/upsetting", "it might output copyrighted content in the reasoning" or "it could result in disclosure of PII" (which are patently false in practice) as disinformation, they will outright poison downstream models' attitudes with these statements in synthetic datasets unless one does heavy filtering.