Páginas no tópico: < [1 2 3 4] > | Poll: Do you think that machine translation will significantly reduce the need for human translation? Tópico cartaz: ProZ.com Staff
| Mario Chavez (X) Local time: 08:47 inglês para espanhol + ... Not happening | Aug 7, 2012 |
Gianluca Marras wrote: unless a machine can think like a human being. Very eloquently put, couldn't have said it better myself. | | |
James_xia wrote: I also voted "No, absolutely not" , only to the word like "significantly" used in the poll. That's the point! Machine translation is a common thing and most of us are preparing or have ever tried to use it in daily translation work. It is, due to the ceilings of the modern technology development, indeed able to reduce part of the workload. Nevertheless, it is not what they called "significantly reduce...". That's not a few years away to reach, but quite a long way ahead of us professionals. Agreed. Moreover, in my case, it will not be reduced at all - not significantly, not merely, not a little bit, because there is no way for a machine to be aware of the correct context, gender, word case and which one of the synonims to use. And I am speaking here of the rare occassions when you don't have to paraphrase a lot. However, English and Latvian languages are so distinct in semantics, use of words and phrases, that most of time, translation has a totally different wording than the original text, sometimes there is not a single word directly translated from the original text in order to make the translation sound natural and fluent in Latvian. It's a creative process which only a human mind is capable of. I don't believe there will ever be a machine with a human-like creative mind, even after zillion years. This is why they are *machines* and we are humans. As for those who don't agree and has an opinion that machines will significantly reduce the workload, I assume your source and target languages are quite similar in semantics and historic roots. And maybe there are no particular word cases and genders either. It's understandable why you argue. But I am speaking only about my source and target languages and similar cases where machines will never be able to make a proper translation without making obvious mistakes [a human would never make but a machine would never be able to overcome] and with correct cases, genders, prepositions, etc. | | | and what we could do is | Aug 7, 2012 |
starting from now, say NO when someone asks us to review a machine-made translation. Simply say NO, or charge 100% more of our translation rate. | | | Yaotl Altan México Local time: 06:47 Membro (2006) inglês para espanhol + ...
It should be banned if we don't want the end of our beloved profession. | |
|
|
Simon Bruni Reino Unido Local time: 13:47 Membro (2009) espanhol para inglês
Gianluca Marras wrote: starting from now, say NO when someone asks us to review a machine-made translation. Simply say NO, or charge 100% more of our translation rate. It takes longer to fix a machine-generated pseudo-translation than it does to make a new one from scratch. I politely explain this to my clients and they are usually more than happy to send me the original text and pay me my full translation rate. Then again, I've done the same with substandard human translations too, more often than not by non-natives. | | | neilmac Espanha Local time: 14:47 espanhol para inglês + ... Yes, it's already happening | Aug 7, 2012 |
We are seeing it already. What it gives rise to is an increase in demand for revision or "proofing" of crappily "translated" machine garble. Decent translations nowadays seem to be pretty thin on the ground. | | | neilmac Espanha Local time: 14:47 espanhol para inglês + ...
Gianluca Marras wrote: starting from now, say NO when someone asks us to review a machine-made translation. Simply say NO, or charge 100% more of our translation rate. I agree. It doesn't stop clients trying to wheedle you into accepting these jobs though... | | | Mario Chavez (X) Local time: 08:47 inglês para espanhol + ... The game of predicting the future | Aug 7, 2012 |
Chris S wrote: Mario Chavez wrote: If you think that MT will improve in the future to the point that it will make human translation in some areas obsolete, then you fail to see the difference between translation as a process unique to the human mind and translation as a matter of speed or "productivity." But the question is not whether machines can replace human translators altogether. I agree that machines will probably never be able to translate poetry well, but machines with a little human help can already replace most translators. Not just MT but Trados too. I'll give you that, the question is referring to some future event or set of circumstances. I am still answering no to the question because it is based on a faulty argument, and I've explained the reasons why. Let me draw another analogy: will high-tech machines ever be able to run for office? Machines and computers are tools that help humans with repetitive or dangerous tasks. That's why automakers and other industries have industrial robots. "Machines with a little human help can already replace most translators"? That's a tall claim. Prove it. Trados has nothing smart or intelligent about it. It is a piece of software. It automates some of the mechanical or manual tasks we do on a daily basis, it does not help in the thought processes for translation or writing in any way, shape or form. | |
|
|
Yes. And it has already started | Aug 7, 2012 |
Chris S wrote: Obviously machine translation is a bit rubbish but it can only get better. Exactly, the problem (for us!) is that then comes... a post-editor!, and the final result is "marketable" (and an obvious threat for us, language professionals). For now, MT needs post-editors, but as time goes by, fewer and fewer post-editors will be needed "thanks" to the MT+TM combo. My position is simple: I will leave the profession if the only option left is to post-edit a machine. No matter the rate, of course. The concept of post-editing is harmful for us (unless you post-edit your own documents, of course, without feeding any vendor's corpus). MT cannot help a company save money and, at the same time, be a "good opportunity" for us, translators. Chris S wrote: every text translated by machine is a text not translated by a human Amen! Kind regards, Au
[Edited at 2012-08-07 22:54 GMT] | | | No - because it opens up more translation possibilities and fields | Aug 7, 2012 |
A translator colleague says to me: "MT reduces my work: clients who once asked me to translate text X now get what they need from a machine translation." I say to my colleague: "MT increases my work: clients can now use MT to look into text X, Y and Z and realise they need a good translation of not just text X but also parts of text Y. The more MT gives monolinguals instant access (however approximate) to foreign-language texts, the more they will a) realise how much va... See more A translator colleague says to me: "MT reduces my work: clients who once asked me to translate text X now get what they need from a machine translation." I say to my colleague: "MT increases my work: clients can now use MT to look into text X, Y and Z and realise they need a good translation of not just text X but also parts of text Y. The more MT gives monolinguals instant access (however approximate) to foreign-language texts, the more they will a) realise how much valuable information there is out there that just happens to not be in their language and b) identify among that material what is most valuable to them and thus deserving of the time and expense required for a quality translation. Or commission 2 or 3 "gist translations" (post-edited MTs) instead of 1 publication-quality text. MT increases the inter-language flow. Increase in inter-language flow means increase in the demand for my skills. Melissa ▲ Collapse | | | Yaotl Altan México Local time: 06:47 Membro (2006) inglês para espanhol + ...
Mario Chavez wrote: Gianluca Marras wrote: unless a machine can think like a human being. Very eloquently put, couldn't have said it better myself. Yes, but there are already some research projects on the brain field, neuron connection trying to emulate beings, first rat, cats, and then human beings. To emulate a mouse's brain using several processors is trying to be a kind of demigod, modificating food and life, in general.... | | | Allison Wright (X) Portugal Local time: 13:47 Thought processes | Aug 8, 2012 |
Mario Chavez wrote: Trados has nothing smart or intelligent about it. It is a piece of software. It automates some of the mechanical or manual tasks we do on a daily basis, it does not help in the thought processes for translation or writing... While we know that Trados is not "pure MT", Trados still has to rely on statistical probability - unlike humans, who have "instant recognition". For example, even an 85% match returned by Trados is often seen by my human brain as no match at all. As humans we process "similar" things differently, in "jumbled" fashion, and not by means of linear progression, which is the way mechanised translation has to behave. As humans (as I presume everyone in this forum is) we also use an incredible amount of extralinguistic data to achieve a proper translation. Google and the like are useful to us in that these tools point out (the gobbledygook factor) where we as humans can best apply that wealth of extralinguistic knowledge, and non-linear thought. | |
|
|
Michele Fauble Estados Unidos Local time: 05:47 Membro (2006) norueguês para inglês + ...
Ty Kendall wrote: ... the technophiles keep saying "oh it's only a few years away" - the trouble is they've been saying that for years. The boy who cried wolf springs to mind. Eventually the wolf did actually show up. | | | Ty Kendall Reino Unido Local time: 13:47 hebraico para inglês Not in my lifetime he won't | Aug 8, 2012 |
Michele Fauble wrote: Ty Kendall wrote: ... the technophiles keep saying "oh it's only a few years away" - the trouble is they've been saying that for years. The boy who cried wolf springs to mind. Eventually the wolf did actually show up. I use that analogy because I'm not so naive as to think it could/would never happen (I won't be caught with my head in the sand) but I seriously doubt it will happen in my lifetime, maybe even my grandchildren's lifetime.... Basically, I think the wolf will show up (at some unspecified point in the (perhaps distant) future), but I doubt I'll be around to greet him. | | | Clearly we're never going to agree, but... | Aug 8, 2012 |
Mario Chavez wrote: Let me draw another analogy: will high-tech machines ever be able to run for office? Machines and computers are tools that help humans with repetitive or dangerous tasks. That's why automakers and other industries have industrial robots. But most of a translator's work is basic and repetitive! Even with arty-farty jobs, most words are straight, literal translations - take "chien" and replace with "dog". This is perfect fodder for machines. Mario Chavez wrote: "Machines with a little human help can already replace most translators"? That's a tall claim. Prove it. It's an opinion. For most technical and commercial texts, I think MT plus editor will be quicker and so cheaper/better than the average translator bashing slowly away at a keyboard. My dictated first drafts can also be pretty dire. Mario Chavez wrote: Trados has nothing smart or intelligent about it. It is a piece of software. It automates some of the mechanical or manual tasks we do on a daily basis, it does not help in the thought processes for translation or writing in any way, shape or form. No, but it does a lot of the work for the translator, which is the point. Every phrase it matches or half-matches is work the translator no longer needs to do and probably doesn't get paid for. Again, technology is reducing the need for human translation. Don't get me wrong, I am not a fan of CAT or MT and I'm keenly aware of the value that a good human translator can add. Unfortunately my experience (which is exclusively of qualified, tested, professional translators in the developed world) is that most translators are not very good. I cannot see how machines can fail to do a better job than these translators on the majority of texts where accuracy is more important than style. | | | Páginas no tópico: < [1 2 3 4] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Poll: Do you think that machine translation will significantly reduce the need for human translation? Wordfast Pro | Translation Memory Software for Any Platform
Exclusive discount for ProZ.com users!
Save over 13% when purchasing Wordfast Pro through ProZ.com. Wordfast is the world's #1 provider of platform-independent Translation Memory software. Consistently ranked the most user-friendly and highest value
Buy now! » |
| Trados Studio 2022 Freelance | The leading translation software used by over 270,000 translators.
Designed with your feedback in mind, Trados Studio 2022 delivers an unrivalled, powerful desktop
and cloud solution, empowering you to work in the most efficient and cost-effective way.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |