3 ms·

The thing is, that would require that your smartphone be sufficiently aware of the context of your words that it could correctly map one ambiguous statement in

The thing is, that would require that your smartphone be sufficiently aware of the context of your words that it could correctly map one ambiguous statement in one language to another in another language with a different set of ambiguities. For example, the German 'ich mag sie' could mean 'I like her' or 'I like them' or 'I like you' in formal settings, while the English 'at least' could be translated to German as 'zumindest', or 'mindestens', or 'wenigstens', with different emotional significance. I'm not sure you could really do adhoc translation at the language level with any real fluidity: you would probably need your phone to monitor your brain before you spoke to get at the pre-language meaning of what you are saying, if that is even possible.

sho_hn 12y ago
Not to detract from your point (which is very true; for machine translation to match human translations they need equal access to and understanding of context), but as a native German speaker, your example seems wrong to me :). "It is pleasing to her" = "es gefällt ihr","It is pleasing to them" = "es gefällt ihnen". No doubt it's possible to construct examples that support your agument in just about any language, though. I'd recommend East Asian ones, since they're less heavy on the pronouns than Indo-European languages, and you can generally often omit subject or object if it's clear from context, or are even expected to for good style. Korean poetry even relies on intentional ambiguity a lot of the time, forcing the reader to consider the options, for their enjoyment or to challenge them.
- sbmassey 12y ago
  true, damn, let me fix that... I was thinking of using sie in the accusative rather than the dative ... hopefully 'ich mag sie' works better, right?
  - sho_hn 12y ago
    Yep, 'ich mag sie' could be informally referring to a woman, formally referring to a person of either gender, or just any group. Machine translation is of course aware of this problem, and doing the (rough) equivalent of "whole program analysis" is an active concern. I.e. you can try to source info elsewhere in a longer input to decide between multiple possible translations for another bit. The problem is really these short inputs where no context is easily available to the program - it's an input problem in some sense; it's just much easier for a human translator to guess from cues the program simply has no input channels for. One of those cues is who exactly they've been hired by, though, and that can be replicated with user profiles, too.

sho_hn 12y ago

Not to detract from your point (which is very true; for machine translation to match human translations they need equal access to and understanding of context), but as a native German speaker, your example seems wrong to me :). "It is pleasing to her" = "es gefällt ihr","It is pleasing to them" = "es gefällt ihnen". No doubt it's possible to construct examples that support your agument in just about any language, though. I'd recommend East Asian ones, since they're less heavy on the pronouns than Indo-European languages, and you can generally often omit subject or object if it's clear from context, or are even expected to for good style. Korean poetry even relies on intentional ambiguity a lot of the time, forcing the reader to consider the options, for their enjoyment or to challenge them.

sbmassey 12y ago
true, damn, let me fix that... I was thinking of using sie in the accusative rather than the dative ... hopefully 'ich mag sie' works better, right?
- sho_hn 12y ago
  Yep, 'ich mag sie' could be informally referring to a woman, formally referring to a person of either gender, or just any group. Machine translation is of course aware of this problem, and doing the (rough) equivalent of "whole program analysis" is an active concern. I.e. you can try to source info elsewhere in a longer input to decide between multiple possible translations for another bit. The problem is really these short inputs where no context is easily available to the program - it's an input problem in some sense; it's just much easier for a human translator to guess from cues the program simply has no input channels for. One of those cues is who exactly they've been hired by, though, and that can be replicated with user profiles, too.