James Chandler wrote:
>I really hope it doesn't happen, because I can't imagine anything >that
will more completely and permanently kill interest in language >learning
than the ability to translate texts immediately from most >foreign
Don't be so sure! Most adults I have encountered learning a foreign
language seem to do so for the pleasure of the challenge. Human lives
are fairly mundane on the whole, and are likely to become more so with
advances in technology. Mundane, that is, unless they choose to
challenge themselves in some way. Why does the ape ancestor always come
out in men who want to 'test their mettle' by fighting other men?
Technology has given us far more efficient ways of killing each other
than fist-fighting, but the popularity of this 'pasttime' amongst some
quarters never ceases to amaze.
Back to the point - maybe it will, maybe it won't... Sometimes people
learn language out of interest, sometimes out of necessity. If it were a
necessity, I would probably take the 'babel fish' option - but what if
I wonder if people will ever find listening to a translation of a song
(even a 'perfect' translation) compared with the original. Machine
translation will greatly reduce the drudgery of language learning for
those who must, and have little effect on those who just enjoy.
As a pointless aside, I have often dreamt of the wonders of a
'translation chip' which is installed directly into the brain and which
gives 'perfect' (probably 'technically accurate' might be a better
phrase, but it's too long) translations, yet still there were a few
languages I would like to have excluded from my chip so that I may learn
them for myself! Strange, isn't it!
And surely there will never be a match for the feeling of 'thinking' in
another language, as compared with just knowing what is being said.
>[In fact, if they are going to ban human cloning and other kinds of
>genetic engineering, isn't there a case for banning machine
>translation research? Perhaps mankind can't in the end be prevented
>from inventing the technology to wipe out all that is valuable in >its
existence, and eventually also itself.]
Such technophobia, James! Still, maybe you'll be proven right in 100
years when machines run the earth. OTOH, I can't imagine they'll do any
worse than us! :-)
Something I read on machine translation says that the accuracy and speed
of real-time machine translation should improve dramatically in the next
25-50 years. Still, the cost and availability of these technologies will
be limited at first. Perhaps something phenomenal will occur in this
field in the next 10 years, perhaps it won't be significant for 100
years. Is it worth worrying about?
I think the point about this field's relation to AI overall is true.
Translation has much in common with other artistic fields of human
endeavour. Will a machine ever truly understand concepts like
'subtlety', or will they just be programmed to recognise them?
The long and short of my discussion here is this : We don't need to pack
it in *just* yet!!
Cheers, and Hail to our Machine Masters!
Get Your Private, Free Email at http://www.hotmail.com