Why translators do not need to be afraid of Google’s neural networks
This article - a great commentary on the news about Google Translate has connected the Russian language to a translation with deep learning . At first glance, everything sounds and looks very cool. However, I will explain why you should not rush to conclusions about "translators are no longer needed."
The trick is that today technology is capable of replacing ... yes, it cannot replace anyone.
The translator is not the one who knows a foreign language, just as the photographer is not the one who bought the big black DSLR. This is a necessary condition, but far from sufficient.
A translator is someone who knows his language well, understands someone else well and can accurately convey the nuances of meaning.
')
All three conditions are important.
While we do not even see the first part (in terms of "knows your language"). Well, at least for Russian, so far everything is very, very bad. That's really something, and the placement of commas perfectly algorithms (Word coped with the year like this in 1994, licensing the algorithm from the local ones), and for the neural network of the existing UN corpus of texts, it’s just above the roof.
Let's return to oursheep translators. By Pareto law, 80% of professional translators are bad. These are people who have graduated from a foreign language course or, at best, some regional pedagogical institute with a degree in a foreign language teacher for rural areas. And they have no other knowledge. Otherwise, they would not be sitting on one of the lowest paid jobs.
Do you know what they earn? No, not in translations. As a rule, customers of these translations understand the text in a foreign language better than a translator.
They sit on the requirements of legislation and / or local customs.
Well, here we have to, that the instruction to the product was in Russian. Therefore, the importer finds a person who knows a little “imported” language, and he translates this instruction. This person does not know the product, has no knowledge in this area, he had “three and a minus” in Russian, but he translates. The result is known to all.
It is even worse if he translates "in the opposite direction", i.e. in a foreign language (hello to the Chinese). Then his work is likely to fall into Exler's “bannisms” or their local counterpart.
Or here's a more difficult case. When contacting the state. authorities with foreign documents need to submit a translation of these documents. And the translation should not be from Uncle Vasya, but from a legally respected office, with “wet” seals, etc. Well, tell me, how difficult is it to “translate” a driver's license or is there a birth certificate? All fields are standardized and numbered. The "translator" needs, in the worst case, simply to transliterate proper names from one alphabet to another. But no, “Uncle Vasya” has a rest, and, more often, thanks to not even the law, but simply the internal instructions of local officials.
Please note, 80% of translation offices live with notaries. Guess from three times why?
How will these translators be affected by good machine translation? Yes, nothing. Well, that is there is hope that the quality of their translations will still improve in some minor aspects, where there is something to translate. Well, that's it. Working time here is not significantly reduced, because they are now most of the time copying text from columns to columns. “There are so many proteins in this cheese, so many carbohydrates ...” The national forms are different in different countries, so they won’t work less. Especially if you do not make an effort.
Intermediate conclusion: for the bottom 80% nothing will change. They already earn, not because the translators, but because the bureaucrats of the lowest level.
Now look at the opposite part of the spectrum, well, let it be the top 3%.
The most responsible, though not the most technically complex 1%: simultaneous translation of very important negotiations. Usually between large corporations, but in the limit - in the UN or similar tops. One mistake of the translator in the transfer is not even a point - emotions, can lead, in the worst case, to an atomic war. In this case, as you understand, the emotional color of even literally matching phrases in different languages can be very different. Those. the translator should ideally know both the cultural contexts of his working languages. Banal examples are the words "negro" and "invalid." They are almost neutral in Russian and vividly emotionally colored, right up to obscenity, in modern English.
Such translators may not be afraid of AI: no one will ever entrust such responsibility to the machine.
The next 1% are literary translators. Well, for me, for example, the whole regiment is highlighted by carefully collected original English-language publications by Conan Doyle, Lewis Carroll, Hugh Laurie - in the original, without any adaptations there or our local reprints. Reading these books is excellent vocabulary, you know, well, in addition to great aesthetic pleasure. I, a certified translator, can retell very close to the text any sentence from these books. But take up the translation? Unfortunately no.
I do not even stutter about translations of poetry.
Finally, the most technically complex (for a neural network, generally impossible) 1% is a scientific and technical translation. Usually, if a team in a country has pulled ahead in their field, they call their discoveries and inventions in their own language. It may happen that in another country, another team independently invented / discovered the same. So there were, for example, the laws of Boyle-Mariotta, Mendeleev-Poisson and disputes on the topic Popov / Marconi, Mozhais / Wright brothers / Santos-Dumont.
But if a foreign team is “completely galloped forward”, the “catching up” scientists have two options in the linguistic sense: to calculate or translate.
Calculating the names of new technologies, of course, easier. This is how algebra , medicine and computer appeared in Russian, in French - bistro , datcha and vodka ; in English - sputnik , tokamak and perestroika .
But sometimes still translate. The voice of the humanities in my head wildly rushes from the term touchpad to denote the argument of the Fourier transform from the Fourier transform, as a translation for querquency . Jokes aside, there are no such terms in Google - but I have a paper textbook on digital signal processing, approved and sanctified by the Ministry of Environment, in which these terms exist.
And yes, the touch screen analysis is the only (known to me) way to distinguish a male voice from a female one. Options?
What I am getting at is that these people have nothing to fear, because they themselves form the language, introduce new words and terms into it. Neural networks just learn from their solutions. Well, not forgetting the fact that these scientists and engineers do not earn money on translations.
And finally, the "middle class", good professional translators, but not the tops. On the one hand, they are still protected by the bureaucracy - they translate, for example, instructions, but not to homeopathic bads, but, let's say, to normal medicines or machine tools there. On the other hand, it is already today modern workers with high automation of labor. Their work now begins with the compilation of a “glossary” of terms so that the translation is uniform, and then, in essence, consists in editing the text in a specialized software like trados. Neural networks will reduce the number of necessary edits and increase labor productivity, but they will not fundamentally change anything.
Total, rumors about the imminent death of the profession of an ordinary translator are a bit exaggerated. At all levels, work will accelerate a bit and competition will increase slightly, but nothing unusual.
But who will get it is translators-journalists. 10 years ago they could easily refer to an English-language article from which they did not understand anything, and write complete nonsense. Today they are also trying, but English-speaking readers dunk them over and over again ... well, you understand.
In general, their time has passed. With a mid-level universal machine translator, albeit a bit clumsy, “journalists” such as alizar or ragequit become just unemployed.
The trick is that today technology is capable of replacing ... yes, it cannot replace anyone.
The translator is not the one who knows a foreign language, just as the photographer is not the one who bought the big black DSLR. This is a necessary condition, but far from sufficient.
A translator is someone who knows his language well, understands someone else well and can accurately convey the nuances of meaning.
')
All three conditions are important.
While we do not even see the first part (in terms of "knows your language"). Well, at least for Russian, so far everything is very, very bad. That's really something, and the placement of commas perfectly algorithms (Word coped with the year like this in 1994, licensing the algorithm from the local ones), and for the neural network of the existing UN corpus of texts, it’s just above the roof.
Who does not know, all official UN documents are issued in the five languages of the permanent members of the Security Council, including Russian, and this is the largest database of very high-quality translations of the same texts for these five languages. Unlike the translation of works of art, where “the translator of Ostap can suffer”, the UN base is distinguished by the most accurate transfer of the most subtle nuances of meaning and perfect correspondence to literary norms.
This fact, plus the absolute free of charge, makes it an ideal set of texts (corpus) for training artificial translators, although it covers only a purely official bureaucratic subset of languages.
Let's return to our
Do you know what they earn? No, not in translations. As a rule, customers of these translations understand the text in a foreign language better than a translator.
They sit on the requirements of legislation and / or local customs.
Well, here we have to, that the instruction to the product was in Russian. Therefore, the importer finds a person who knows a little “imported” language, and he translates this instruction. This person does not know the product, has no knowledge in this area, he had “three and a minus” in Russian, but he translates. The result is known to all.
It is even worse if he translates "in the opposite direction", i.e. in a foreign language (hello to the Chinese). Then his work is likely to fall into Exler's “bannisms” or their local counterpart.
Or here's a more difficult case. When contacting the state. authorities with foreign documents need to submit a translation of these documents. And the translation should not be from Uncle Vasya, but from a legally respected office, with “wet” seals, etc. Well, tell me, how difficult is it to “translate” a driver's license or is there a birth certificate? All fields are standardized and numbered. The "translator" needs, in the worst case, simply to transliterate proper names from one alphabet to another. But no, “Uncle Vasya” has a rest, and, more often, thanks to not even the law, but simply the internal instructions of local officials.
Please note, 80% of translation offices live with notaries. Guess from three times why?
How will these translators be affected by good machine translation? Yes, nothing. Well, that is there is hope that the quality of their translations will still improve in some minor aspects, where there is something to translate. Well, that's it. Working time here is not significantly reduced, because they are now most of the time copying text from columns to columns. “There are so many proteins in this cheese, so many carbohydrates ...” The national forms are different in different countries, so they won’t work less. Especially if you do not make an effort.
Intermediate conclusion: for the bottom 80% nothing will change. They already earn, not because the translators, but because the bureaucrats of the lowest level.
Now look at the opposite part of the spectrum, well, let it be the top 3%.
The most responsible, though not the most technically complex 1%: simultaneous translation of very important negotiations. Usually between large corporations, but in the limit - in the UN or similar tops. One mistake of the translator in the transfer is not even a point - emotions, can lead, in the worst case, to an atomic war. In this case, as you understand, the emotional color of even literally matching phrases in different languages can be very different. Those. the translator should ideally know both the cultural contexts of his working languages. Banal examples are the words "negro" and "invalid." They are almost neutral in Russian and vividly emotionally colored, right up to obscenity, in modern English.
Such translators may not be afraid of AI: no one will ever entrust such responsibility to the machine.
The next 1% are literary translators. Well, for me, for example, the whole regiment is highlighted by carefully collected original English-language publications by Conan Doyle, Lewis Carroll, Hugh Laurie - in the original, without any adaptations there or our local reprints. Reading these books is excellent vocabulary, you know, well, in addition to great aesthetic pleasure. I, a certified translator, can retell very close to the text any sentence from these books. But take up the translation? Unfortunately no.
I do not even stutter about translations of poetry.
Finally, the most technically complex (for a neural network, generally impossible) 1% is a scientific and technical translation. Usually, if a team in a country has pulled ahead in their field, they call their discoveries and inventions in their own language. It may happen that in another country, another team independently invented / discovered the same. So there were, for example, the laws of Boyle-Mariotta, Mendeleev-Poisson and disputes on the topic Popov / Marconi, Mozhais / Wright brothers / Santos-Dumont.
But if a foreign team is “completely galloped forward”, the “catching up” scientists have two options in the linguistic sense: to calculate or translate.
Calculating the names of new technologies, of course, easier. This is how algebra , medicine and computer appeared in Russian, in French - bistro , datcha and vodka ; in English - sputnik , tokamak and perestroika .
But sometimes still translate. The voice of the humanities in my head wildly rushes from the term touchpad to denote the argument of the Fourier transform from the Fourier transform, as a translation for querquency . Jokes aside, there are no such terms in Google - but I have a paper textbook on digital signal processing, approved and sanctified by the Ministry of Environment, in which these terms exist.
And yes, the touch screen analysis is the only (known to me) way to distinguish a male voice from a female one. Options?
What I am getting at is that these people have nothing to fear, because they themselves form the language, introduce new words and terms into it. Neural networks just learn from their solutions. Well, not forgetting the fact that these scientists and engineers do not earn money on translations.
And finally, the "middle class", good professional translators, but not the tops. On the one hand, they are still protected by the bureaucracy - they translate, for example, instructions, but not to homeopathic bads, but, let's say, to normal medicines or machine tools there. On the other hand, it is already today modern workers with high automation of labor. Their work now begins with the compilation of a “glossary” of terms so that the translation is uniform, and then, in essence, consists in editing the text in a specialized software like trados. Neural networks will reduce the number of necessary edits and increase labor productivity, but they will not fundamentally change anything.
Total, rumors about the imminent death of the profession of an ordinary translator are a bit exaggerated. At all levels, work will accelerate a bit and competition will increase slightly, but nothing unusual.
But who will get it is translators-journalists. 10 years ago they could easily refer to an English-language article from which they did not understand anything, and write complete nonsense. Today they are also trying, but English-speaking readers dunk them over and over again ... well, you understand.
In general, their time has passed. With a mid-level universal machine translator, albeit a bit clumsy, “journalists” such as alizar or ragequit become just unemployed.
Source: https://habr.com/ru/post/373239/
All Articles