3 ms·
"And when the floor goes, the thing that decides what is worth keeping is no longer the cost of making it." I'm 100% confident that a human did not write this
by no_multitudes 2mo ago
"And when the floor goes, the thing that decides what is worth keeping is no longer the cost of making it."
I'm 100% confident that a human did not write this sentence unless they were deliberately trying to sound like an LLM. Just look at it! You don't need Pangram. There's a classifier model between your ears that knows how humans write.
- tfrancisl 2mo agoI'd retrain that model if I were you.
- no_multitudes 2mo agoI'm generally quite open to the possibility that I'm wrong. I said 100% for this one because it is tooth-grindingly obvious. Could you give me some examples of long-form writing like this from before 2024? Generally to retrain my model I need data.
- Tiberium 2mo agoTo be honest with you, I think this is pointless. However hostile this might sound, the active userbase of HN seems to be extremely dismissive to the thought of LLM detectors and they think that all those LLM written articles are actually human written. I don't know if this will change, perhaps it'll only get worse from now. (You can see my recent experience in another thread https://news.ycombinator.com/item?id=49212027 https://news.ycombinator.com/item?id=49212027 )
- slopinthebag 2mo agoIf you don’t know how to write well, the AI slop is less obvious. That’s my theory at least, I don’t understand how people can read these nonsense sentences and believe a human actually wrote it.
- runtime_terror 2mo agoRunning Llama 3 between your ears? Don't you and all the people in this thread claiming AI slop know that AI was trained on human writing?
- no_multitudes 2mo ago>Don't you and all the people in this thread claiming AI slop know that AI was trained on human writing? Do you think this implies that it is impossible to distinguish LLM-generated text from human text? I don't think that follows at all. >Running Llama 3 between your ears? I'm always happy for a chance to improve my classifier! Can you provide some long-form writing from before 2024 that uses this style?
- runtime_terror 2mo agoNo, I dont think there is any reliable data to be able reliably classify AI text. So far what I've seen is humans actually fail this task reliably
- no_multitudes 2mo agoInteresting hypothesis! I find that my internal sense of whether the article is AI almost always agrees with Pangram when I check it, and almost never triggers on pre-2024 text, even if I didn't know beforehand when it was written. I think this cuts against your hypothesis -- if I was just recognizing a generic style of writing, I would expect it to fire on pre-2024 text, and if I was just recognizing text written in 2026, I would expect it to fire much more often when reading blog posts of already-established writers who I trust.
- jmilloy 2mo agoThat sentence doesn't read as llm to me.
- no_multitudes 2mo agoCan you give an example of long-form writing that sounds like this from before 2024?