That isn't about the social hierarchy, it's about the switch. If the couple would always use 존댓말 - which is not the norm nowadays but still a large enough minority that it's not weird - then obviously it'd be a switch to 반말 that would be shocking; in fact it would sound even worse to suddenly do so during a fight.
> As an obvious foreigner, you are 100% exempt from all this. In fact, if you get it right, it'll make them feel weird and they'll try to avoid that.
Are you sure? In Korea, everyone who seriously cares about this stuff absolutely prefer it if you get all of it right. Of course, plenty of people (probably most under 30) secretly hate it and are therefore thankful if you don't get it right as it gives them a semi-out to be slightly more loose as well.
> I read a story once about someone who went to McDonalds in Japanese, and the clerk flipped the menu to the English side. The person flipped it back over, and the clerk flipped it to English again! They simply couldn't believe the foreigner could read it.
> And I've read stories of people who didn't act like a "gaijin" (foreigner) and people didn't know how to interact with them, and the person finally just accepted it and acted like they expected a gaijin to act, and then everything was fine.
Not trying to be rude but this pretty much tells me your first claim is likely wrong. Plenty of the exact same stories abound about Korea, especially online, yet they're both outdated and very cherry-picked. They're real, they happen (I've experienced them myself), but 99% of people would strongly prefer it if you were fluent in the language and followed every procedure as "normally" as possible.
They are, but they have in common that a huge portion of things written about them online in English - especially about these cultural subjects - are very outdated or otherwise unrepresentative. Like I said, most of that comment is written all the time about SK too, and they're effectively untrue. There would need to be strong evidence that such reports about Japan are any different.
Never heard someone saying 님 sounds creepy, it's very commonplace in tech companies, Kakao was far from the first to introduce this. I've worked at a place where everyone was ..님 and no one cared.
Just so you know, the system has been adjusted a while ago so that new 빠른 no longer occur.
And it wouldn't matter anyway; the changes from the top are already having their effect, elementary school kids are starting to consistently use 만 나이 with each other.
That's only viable if the quality of the outputs can be automatically graded, reliably. GP's case sounds like one where that's probably possible, but for lots of specific tasks that isn't feasible, including the other ones he names:
> write poetry, give me advice on cooking, or translate to German
Certainly, in those cases one needs to be clever and design an evaluation framework that will grade based on soft criteria, or maybe use user feedback. Still, over time a good train-test database should be built and leveraging dspy will do improvements even in those cases.
If by reasoning you mean showing CoT, Gemini and OA are the same in this regard - neither provides it, not through the UI nor through the API. The "summaries" both provide have zero value and should be treated as non-existent.
Anthropic exposes reasoning, which has become a big reason to use them for reasoning tasks over the other two despite their pricing. Rather ironic when the other two have been pushing reasoning much harder.
reply