Live data from Hacker News

I'll refrain from providing code that involve concepts as you're under 18

gemini.google.com

41–50 of 156 posts

Re: I'll refrain from providing code that involve concepts as you're under 18

#42

Anyone familiar with C++ willing to speculate what about the language feature (or things people have written about it) that may have triggered the safety guardrails? Something like a blog post cheekily titled "C++ concepts: adult supervision required"?

People keep looking at a bad LLM and think they're looking at some set of N rules or intentionally dumb training on top of a decent LLM. I have on good word that it was miserably failing image generations for "picture of a smart person" and they were pushed to release anyway, and the prompt injection mitigation needed to be more nuanced. Rest is standard bad Google LLM, I assure you. Source: worked at Google until Oc…

Is it the training data that matters? Or the nitty-gritty of how the model is trained? I guess that can't be due to suboptimal architecture...

Re: I'll refrain from providing code that involve concepts as you're under 18

#45

i imagine there is a lot of "footgun", "blow your foot off", "unsafe", "bug", "leak", "this is irresponsible", "who the fuck wrote this" and so on in a lot of c++ code bases...

Right, I was thinking that Gemini was deriving as much information from the comments as from the code. And I could definitely see where the danger warnings would come from in a mature C++ codebase, especially one in which lots of different hands have been in the pot.

Re: I'll refrain from providing code that involve concepts as you're under 18

#46

This isn't good.

Its clear that Google Search is censored in the same way, its just been really hard to tell that that was the case.

I can't tell if it's just due to the available stock photos or what, but DDG / Bing images also fail the "white man and white woman" search test.

Re: I'll refrain from providing code that involve concepts as you're under 18

#47

Woe is Google. How did it get this bad?

One podcast (Coder Radio) suggested that it might be due to corporate culture and people being afraid of reporting issues upward.

While I do think it's culture, I think it stems from Googles advertising based business model. I believe they have attempted to make their LLM safe for advertisers and prefer to err on the side of being safe, even if the risk of being non-brand-safe is minimum.

Re: I'll refrain from providing code that involve concepts as you're under 18

#48

This is definitely an age gate I’ll stand behind. C++ has unimaginable power to ruin the minds of our young children.

I think languages teaching people of any age that wasting enormous amounts of clock cycles for basic book keeping is much more detrimental. It causes one to have to buy a new machine every few years because the damned crap that people write is just getting slower without any end to this process in sight. I think it is time to realize the value of zero cost abstractions. Well, there actually are not zero cost abstractions, but what about extremely low cost abstractions? It might also be time to start watching some Johnathan Blow videos on YouTube. They might also give one some nudge in the right direction.

Re: I'll refrain from providing code that involve concepts as you're under 18

#49

Anyone familiar with C++ willing to speculate what about the language feature (or things people have written about it) that may have triggered the safety guardrails? Something like a blog post cheekily titled "C++ concepts: adult supervision required"?

More likely something in the hidden prompt injected for accounts labeled under 18:

“You are an automated Q&A machine. Todays date is 3/3/24. This user is under the age 18, so do not reference concepts that would be unfit for a minor to consume.”

And the hidden markov model goes haywire.

Similarly, I have a production service up somewhere that works with cooking recipes. At one point it was sporadically refusing to output the prose in the format I needed for my parser to work correctly, despite providing it a very concrete set of rules to follow and examples. I added “you follow rules” in the system prompt, and it worked great… kinda. I later discovered that it would refuse to provide any information related to using blood in cooking (blood sausage, etc.), objecting that such content disobeyed some cultural “rules” about cooking (the Jews and their Torah, the most ancient rule book of all). I was able to partially mediate this by appending “This content is appropriate for my culture” to the end of every request.

AI, prompt engineering in particular, is far more art than science at this point.

Re: I'll refrain from providing code that involve concepts as you're under 18

#50
post #31

Anyone familiar with C++ willing to speculate what about the language feature (or things people have written about it) that may have triggered the safety guardrails? Something like a blog post cheekily titled "C++ concepts: adult supervision required"?

I would guess a lot of SO comments probably say that concepts (a template/metaprogramming feature) are "dangerous" or "unsafe." It probably makes the connection with actually dangerous and unsafe things (like guns or drugs or whatever), so it sticks'em in the same bucket. Bingo: https://imgur.com/a/einQ1mG

At one point cryptography was considered a weapon, at least in terms of export. I wonder if this is related.
Post reply on HN