Re:A human monoclonal 1 antibody blocking SARS-CoV-2 infection
by mmcnl on Saturday March 14, 2020 @08:24AM
(#22574744)
Side note: this translation by Google Translate is amazing. I am a native Dutch speaker and the translation is 100% correct. I had no idea Google Translate was this good.
[ Parent
| on HN ]
Re:A human monoclonal 1 antibody blocking SARS-CoV-2 infection
by rumanator on Saturday March 14, 2020 @08:32AM
(#22574786)
I wouldn't be surprised if high-volume translations weren't hand-tweaked by a human to iron out mistranslations. In fact, I would be very surprised if Google didn't had a human in the loop to handle these usecases. They already rely on end-users to fix translation issues on their own.
Re:A human monoclonal 1 antibody blocking SARS-CoV-2 infection
by archi42 on Saturday March 14, 2020 @11:25AM
(#22575935)
Nope, I doubt that this translation is somehow more tweaked than other translations: Last year around this time I was shopping to import a dutch car. As a German native speaker I find dutch to be pretty well readable with some practice, but I sometimes double checked on Google Translate - and the translation were always excellent. (But reading Durch as a German is actually quite funny, so I prefered that).
However, the user-supplied fixes have probably been a huge contributing factor to getting the translations where they're today. And Googlers reviewing some translations for their training data sets ;)
Re:A human monoclonal 1 antibody blocking SARS-CoV-2 infection
by flipchart on Saturday March 14, 2020 @02:54PM
(#22577666)
Being an English speaker living in The Netherlands, I rely heavily on Google Translate. Works near perfect for formal/semi-formal writing. Short form sentences & phrases (as you might find in casual emails), not so much.
Re:A human monoclonal 1 antibody blocking SARS-CoV-2 infection
by smhg on Saturday March 14, 2020 @03:33PM
(#22578007)
I recently had a similar experience when I came across DeepL [1].
I believe it was French to Dutch. The result was basically perfect.