Comment by the_af
3 days ago
To be clear, I've seen it in the wild, but not here where it's discouraged to pick on words instead of focusing on the substance of what's being said.
3 days ago
To be clear, I've seen it in the wild, but not here where it's discouraged to pick on words instead of focusing on the substance of what's being said.
Here's a better example. Use "a few bad apples" wrong, and you'll likely get a response. A few bad apples will cause the entire barrel to spoil rapidly, so a few bad apples is a big deal. But it's often used to say the opposite, that a few bad apples isn't a big deal.
Wow, I guess I never thought about the "few bad apples" figure of speech! Interesting. But regardless, everyone understands what it means in common use, even if it's logically wrong, and I swear I've never seen anybody be a pedant about it here.
And really, it goes against the spirit of HN to hyperfocus on idioms instead of addressing the meat of the argument...
As a personal observation, if an LLM was figuratively looking over my shoulder and pointed out something like "well, ackshually, 'a few bad apples' means..." I would delete the fucker.
A few bad apples is a great idiom though that applies to so many places. For examples, teachers often report that more than 2 troublemakers in a classroom ruins the entire class. A few bad cops destroy trust in all policemen, ruining the the entire force, et cetera.
And more relevant to us, a couple bad lines of code sprinkled in the millions in your code base can ruin the entire thing....
I wish I had posted a better example, but I couldn't recall anything at the moment and still can't. It's usually a more interesting complaint than the old man shaking fist at clouds of the usage of the word literally.
OK, but let's dig deeper.
Would you prefer to be corrected on some logical fallacy/mistake you made in your argument, by another human being (and yes, maybe get slightly upset about it, we're human beings after all), or have both sides present bot-mediated iron-clad comments, like operators sparring with robots?
I prefer the raw, flawed human version. Even if, yes, I make a silly, avoidable mistake, or get upset, or make you upset in the heat of the argument. Maybe when I cool down I will have learned something.
I don't want flawless robotic arguments. I want human beings. (Fuck, that last bit sounded like an AI-ism, but I promise it's me, a human!).