Machine translation improves but still falls short in literary art of human translators
-
Machine translation has improved tremendously since 2006, becoming widely used for tourism and basic communication, but still struggles with literary translation.
-
Literary translation requires making compromises between literal meaning and preserving voice, spirit, etc. which AIs currently aren't skilled at.
-
Recent research shows that post-edited machine translations can compare well to human translations in terms of reader perceptions, but may lack creativity and consistency of voice.
-
Some European publishers are releasing post-edited translations, especially for commercial fiction, though there are concerns about impacts on translators and translation quality.
-
The continued limitations of AI in literary translation highlight the artistry and value of human literary translators.