AI Translated.

Base +6, randomsomething -6, Poe mk.II +6, flansc +9, ernie12 +16, Master Sigiward -3, LikewiseLikewise +7, Videlicet Rex Rgis +7, nmop_apisdn +6, kishudark +6

Super, old school, I waited a long time for the translation.
Even though it’s AI, we’ll see what the quality is like.
Thank you.

Base +6, Asphen +7, LikewiseLikewise -7, fdcx -10, SomeLazyDouche -7, ackmanb -6, Videlicet Rex Rgis -7, nmop_apisdn -6, Shino22 -7, Dune69 +14, kishudark -6

Thanks for the translation.

Base +6, LikewiseLikewise -7, fdcx -10, SomeLazyDouche -7, Videlicet Rex Rgis -7, nmop_apisdn -6, Shino22 -7, kishudark -6

I don't get it, didn't we already have reliable translations (after years of pretending they were reliable but they weren't at all) that don't involve AI but just running shit through google or deepL or whatever? What is the point of adding in the randomness of AI where it suddenly shifts to prose?

Base +6, Videlicet Rex Rgis -7, Lady Laila +2, kishudark -6

@Kairu_Hakubi Hate to break it to you, but Google Translate and DeepL are both AI too, DeepL is even named after the "deep learning" of neural networks. If you use them enough you'll see them break in classic weird AI ways too. Separating them from AI+being fine with them just means you're fine with AI's output, as long as you don't know it's AI.

Base +6, Videlicet Rex Rgis -7, Dcsuper +6, kishudark +6

Love how Cousalino got down voted for stating factually correct information.
Twitter and its consequences have been a disaster for the human race.

Base +6, Dune69 -14, kishudark +6

@Kairu_Hakubi those systems are accurate enough in every limited context. Specifically, they work decently if you want a very literal translation of small pieces of text that don't need to be strongly connected to each other; like signage or simple instructions. However, they'll still fall apart if you try to expand them to more complicated and/or longer pieces of prose and, to be clear, the bar for "more complicated" is very low. New systems also have problems in those context, but a slightly different set of problems and they can be better at faking accuracy. (I.E. if you were trying to translate straight text, rather than speech bubbles, most of the text might look fine at first glance, even if it fails to convey the source material properly or falls apart in more nuanced ways that are more apparent when you take the work as a whole.

Either way, though, automated translations are worthless for this kind of thing and the attitudes of the people who put them up do them no favors.

Base +7, kishudark +6

tagging rtl (title)

Base +39, kishudark +6