4 ms·

So I guess the hacker news downvote police got after this one because it was actually informative? Am I reading this right?

alex43578 7mo ago
More because it's AI-slop that doesn't say much more than "government and big business bad".
- GorbachevyChase 7mo ago
  I don’t see a trace of mark down or em dashes. I think maybe you’re perhaps a little too emotionally invested in politics
  - sneak 7mo ago
    https://git.eeqj.de/sneak/prompts/src/branch/main/prompts/LLM_PROSE_TELLS.md https://git.eeqj.de/sneak/prompts/src/branch/main/prompts/LL...
  - alex43578 7mo ago
    Firstly, I could care less about the regional politics of Hawaii's infrastructure. I was just answering why this was likely getting downvoted. As for it being AI, GPTZero puts it at 99% AI. "Insolvency insurance" is used out of context, incorrectly mixing the financial metaphors he told the AI to use with the more-relevant idea of flood insurance (was insolvency supposed to be the AI's attempt at a pun around liquids?). There's the classic AI "it isn't X, it's Y" structure structure at the end. The whole thing reads as a prompt of "Recontextualize the potential flood caused by the failure of Wahiawa Dam in Hawaii through a lens of politics, business, and finance". Markdown, em-dashes, and emojis were AI-slop 101 a year ago. You gotta keep up.
    - ses1984 7mo ago
      There is actually an em dash in there, it isn't a disaster—it's the violent…
      - alex43578 7mo ago
        I didn't say there wasn't an em dash, but thought it was worth pointing out the totality of things that stuck out to me as AI evidence.
  - pfannkuchen 7mo ago
    > disaster—it's
tokioyoyo 7mo ago
One cool thing about reading different forums for decades is you get this instinct on spotting AI-generated content. Obviously not 100%, and if you tune your prompts properly it’s impossible to catch. However this just doesn’t sound human-y. No clue how to explain it. But it also sucks because I’m sure I incorrectly tag some real comments as AI slop.
- user2722 7mo ago
  You are right. However I've found most market absolutists write similar to this. The fault is always someone else's. I had thought I had gotten the contents humanly sociopathic enough to go below the LLM radar. I actually intended to leave a comment explaining but I started to lose points and deleted the explanation; I was however unable to delete the main comment.
  - AndrewKemendo 7mo ago
    I wouldn’t trip It’s pervasive at this point that “its AI” is a lazy retort when people have to use a portion of their brains
    - user2722 7mo ago
      Ya but in this case, it _is_ AI. It was just an experiment though -- most market absolutists talk like this.

alex43578 7mo ago

More because it's AI-slop that doesn't say much more than "government and big business bad".

GorbachevyChase 7mo ago
I don’t see a trace of mark down or em dashes. I think maybe you’re perhaps a little too emotionally invested in politics
- sneak 7mo ago
  https://git.eeqj.de/sneak/prompts/src/branch/main/prompts/LLM_PROSE_TELLS.md https://git.eeqj.de/sneak/prompts/src/branch/main/prompts/LL...
- alex43578 7mo ago
  Firstly, I could care less about the regional politics of Hawaii's infrastructure. I was just answering why this was likely getting downvoted. As for it being AI, GPTZero puts it at 99% AI. "Insolvency insurance" is used out of context, incorrectly mixing the financial metaphors he told the AI to use with the more-relevant idea of flood insurance (was insolvency supposed to be the AI's attempt at a pun around liquids?). There's the classic AI "it isn't X, it's Y" structure structure at the end. The whole thing reads as a prompt of "Recontextualize the potential flood caused by the failure of Wahiawa Dam in Hawaii through a lens of politics, business, and finance". Markdown, em-dashes, and emojis were AI-slop 101 a year ago. You gotta keep up.
  - ses1984 7mo ago
    There is actually an em dash in there, it isn't a disaster—it's the violent…
    - alex43578 7mo ago
      I didn't say there wasn't an em dash, but thought it was worth pointing out the totality of things that stuck out to me as AI evidence.
- pfannkuchen 7mo ago
  > disaster—it's

tokioyoyo 7mo ago

One cool thing about reading different forums for decades is you get this instinct on spotting AI-generated content. Obviously not 100%, and if you tune your prompts properly it’s impossible to catch. However this just doesn’t sound human-y. No clue how to explain it. But it also sucks because I’m sure I incorrectly tag some real comments as AI slop.

user2722 7mo ago
You are right. However I've found most market absolutists write similar to this. The fault is always someone else's. I had thought I had gotten the contents humanly sociopathic enough to go below the LLM radar. I actually intended to leave a comment explaining but I started to lose points and deleted the explanation; I was however unable to delete the main comment.
- AndrewKemendo 7mo ago
  I wouldn’t trip It’s pervasive at this point that “its AI” is a lazy retort when people have to use a portion of their brains
  - user2722 7mo ago
    Ya but in this case, it _is_ AI. It was just an experiment though -- most market absolutists talk like this.