When displaying how long ChatGPT “thought” for, the Norwegian version currently says: “Resonnerte for 7 sekunder.” This is incorrect. It should be “Resonnerte i 7 sekunder.”
The mistake likely stems from a direct translation from English, a common issue with prepositions in Norwegian. To illustrate the error for native English speakers, we can reverse the situation: Imagine if the English version said “Resonated in 7 seconds” instead of “Resonated for 7 seconds.” (Whether “Resonated” is the actual term used is beside the point.)
It would be great if OpenAI addressed this, as it makes ChatGPT look unprofessional and feel amateurish when they confidently display in another language than English.
Good catch—this is a great example of how direct translations can sometimes miss native fluency in other languages. While ‘Resonnerte for 7 sekunder’ is a literal translation, I see how ‘Resonnerte i 7 sekunder’ is more natural for Norwegian speakers.
I wouldn’t classify this as a bug since the AI is working as intended, but I definitely agree that localization accuracy is important. These types of refinements would make ChatGPT feel much more professional and native in every language.
It might be helpful if OpenAI introduced a translation feedback mechanism where users could suggest improvements to enhance fluency dynamically. In the meantime, I’d recommend submitting this as feedback rather than a bug, since it’s more about improving AI-generated translations rather than fixing a system failure.
Hope OpenAI considers refining this—it’s small details like this that make AI feel truly multilingual!