Posted by: patenttranslator | March 1, 2016

The Inevitable Unforeseeability of the Foreseeable Future

“I tell in my book that Bill Gates in ‘98, when I asked him what he worried about, he didn’t say the obvious, which is “My competitors, Netscape, or Oracle or Apple.” He said “I worry about someone in a garage inventing something that I haven’t thought of.”

Ken Auletta

Sometime in the 1970s, when IBM decided to subcontract the operating system design for its new computer to Microsoft, (to be called “personal computer” or “PC”),  the company did so because it erroneously determined that a computer’s hardware would naturally be more important and thus more profitable than its operating system. This turned out to be a very shortsighted decision for IBM because once the new PC design was out in the open, there was no reason to continue using IBM hardware.

As journalist Ken Auletta tells it, Bill Gates was more circumspect than IBM executives and bean counters because he knew that he needed to constantly worry that somebody somewhere might be inventing something that he could not foresee himself. Except for the fact that instead of one guy working in a garage, it turned out to be two guys working in a dorm who eventually came up with the idea of an Internet search engine, Bill Gates was quite the visionary to worry about the inevitable unforeseeability of the foreseeable future.

The expression “foreseeable future” is in fact an oxymoron because the so-called foreseeable future has never been foreseeable.

At the end of the 1980s, nobody expected the Soviet Union to implode within a few short years. The only person who did predict it was Andrei Amalryk, in his book “Will the Soviet Union Survive the Year 1984?”, which was published in 1970. But implode it did and very quickly, along with its vassal states in Eastern and Central Europe, because even a tiny dose of hesitant democratization is extremely destructive if you unleash it on a calcified, bureaucratic system that is held together by lies and a lot of violence.

Who would have expected the son of a Jewish immigrant from Hungary to become president of France, or a year later, for white people in America to vote a black guy whose middle name is Hussein into the White House?

Things change all the time, usually quickly and when we least expect it – although it is also true that the more things change, the more they tend to remain the same, especially when it comes to politics.

Are Future Trends in “the Translation Industry” Foreseeable?

If what I just said here is true (and it just might be), is it even possible to try to foresee and predict the future of our profession, or what is often referred to as “trends in the translation industry”?

I believe so, provided that we keep in mind that many things, often the most important ones, cannot be foreseen or predicted. It’s easy to predict that the demand for translation will continue to grow on this planet, from and into many languages. But it is not that easy to foresee how exactly this demand will grow, in which fields and languages, and how that demand will be met.

Because I have been working as a translator since 1980, both in-house and freelance, in Europe, Asia, and the United States, I have already seen a lot of changes to the profession in my lifetime.

A Language Can Become Less or More in Demand with the Passage of Time

One of the inevitable and largely unforeseeable future changes is that over time, the demand for individual languages can change, as some become less important and others become more important.

Starting in 1987, I translated mostly patents and the majority from Japanese. I still translate a lot of Japanese patents, but while Japanese patent translation used to represent about 80 percent of my workload some 20 years ago, last year it was probably only 30 percent, and I am pretty sure that I must have translated more patents from German than from Japanese last year.

And while I was told that there used to be very little work for translators who could translate patent applications from Chinese and Korean, there is so much work for patent translators in these language combinations now that the rates paid to these translators are as high or higher now than the rates that were paid 20 years ago for patent translation from Japanese. This is especially true about Korean patents, because there are many patents in Korean that need to be translated and relatively few translators to do it well, while the competition among legions of Chinese translators keeps the rates for translation from Chinese at a lower level.

So I am predicting, in my wisdom, that the rates for German patent translation will remain relatively steady, and will probably even go up over time, while the rates for Japanese, which have gone down during the last two decades, will probably not go up significantly. They may even still dip a little lower over time, unless Japan wakes up from its post-Fukushima stupor and starts cranking out new patents at the pace it was cranking them out from the 1960s until about the year 2000.

If that happens, a few years later we will all be surrounded again mostly by Japanese TVs, computers and smart phones in every big-box electronics store, instead of seeing mostly Chinese and Korean hi-tech products on display in every store.

Incidentally, if you are a Japanese patent translator who does not seem to be getting enough work these days, it might make sense to start cooperating with a Korean or Chinese translator, or both. That is what I have been doing for quite a few years now and it is working quite well for me – although something like that can obviously only be done if you mostly work for direct clients.

Is Technology on a Collision Course with Human Translators?

Many prophets of doom for human translation and human translators on the Internet have been expecting the ultimate demise of our profession for a long time. The change that many prophets of future trends have been anticipating and foreseeing for at least the last 20 years is the ultimate switch from translation performed by humans to translation performed at lightning speeds by machines.

Based on how many people perceive translation, especially those who don’t understand much about it, human translation should have been replaced by machine pseudo-translation a long time ago. Many people firmly believe that if this hasn’t happen yet, it soon will, as soon as machines are fast and powerful enough, or have enough “corpora” (comparable translated data), etc. The general consensus among translation ignoramuses is that human translators will disappear from the face of the earth within about five years. Incidentally, this has been the general consensus of people who don’t understand how translation works since about 1990.

I don’t believe that computer tools, including machine translation, are on a collision course with human translators. As a translator, I find them very useful, as long as it is up to me whether I use a tool and which tool I want to use. If technology were on a collision course with human translators, we would already have been replaced by these tools. But as I have written in many posts on my blog, although at first glance the product of these tools may look just like real translation, appearances are deceiving, and machine translation is just one example of such a deception.

The “translation industry” is on the forefront of the movement to replace human translation as much as possible by automated translation output generated by computer tools such as CATs, machine pseudo-translation, and based on what I’ve read on social media lately, also by “translation quality checking and validating software tools” such as “X-Bench” and “Verifika”.

Human translators are protesting about being abused by “the translation industry” in this manner, but nobody seems to be paying much attention to them. Here is a translator’s comment on LinkedIn who is being forced by a translation agency to use this software “to ensure translation quality”:

“I translate into and from Hungarian, too, and I know exactly what you are talking about [this was in response to a comment of another translator who was being forced by a translation agency to use a tool that she despised and considered a complete waste of time]. What I do now is that I tell upfront to my clients that if they want to use X-bench, they should not bother asking me to work for them because I will not work with it. Period. It is a complete waste of time, especially with Hungarian, but I suspect this may be the case with other languages, too, because it does such a crude, mechanical comparison that the number of false positives must be high, no matter the language. If you proofread your translations before submitting them, this tool will not tell you anything useful.”

The tools don’t work. In fact, they have the power to kill human translation if used as directed by “the translation industry”. But who gives a damn when the expectation on the part of the great “translation industry leaders” is that these wonderful tools will speed up the translating process and reduce the need for expensive human translating, thus leading to higher and higher profits.

Is “The Translation Industry” on a Collision Course with Its Customers?

Yes, I believe so.

I am predicting, in my wisdom, that the push to replace human translation as much as possible by mechanized and automated output generated by machines and software, to be later only lightly checked (as long as it’s not too expensive) by humans who will be completely subordinated to these machines and software and who can no longer be called translators, will dominate in a certain part of what is called for lack of a better name “the translation industry”.

I am certain that major translation agencies, often referred to as “mega-agencies”, will definitely jump or already have jumped on the bandwagon, lured by the expectation of ever higher and higher profits.

As the reliance of “the translation industry” on mechanized and computerized output and the insistence that all human translators adhere to the same tools prescribed by the boss (often on the basis of ignorance and always to save money on human translation) is forcing educated and highly experienced, specialized translators to abandon agencies that they used to be working for, “the translation industry” is eagerly embracing new, inexperienced translators who are often referred to as “newbies”.

Since this trend for maximization of mechanized and computerized translation output and minimization of participation of human translators in the translation process inevitably results in very poor translation quality, “the translation industry” is on a collision course with its customers, especially among those customers who need highly specialized translations requiring a certain amount of expertise, for example patent translations.

I see this as an opportunity for small, specialized translation agencies, (often referred to as “boutique agencies”) as well as for individual translators who are able to establish a continuing relationship with direct customers and who can deliver what “mega-agencies” don’t seem to be too interested in. Namely, highly specialized translations in fields and subjects that are limited only by the capacity of human brain to absorb and process new knowledge rather than by the storage space on a hard disk and the processing speed of a CPU.

To the extent that that the future can be predicted or foreseen at all, this is something that I definitely foresee in the foreseeable future.

Advertisements

Responses

  1. I agree about the positive outlook for translators with a real specialism in demanding subject areas. I am frequently amused by digital romantics who assume that computer algorithms will soon be able to handle the convoluted German legal and construction texts which I regularly get to translate.

    Liked by 2 people

  2. Below is a response to my blog post that I am reposting from TheOpenMic.

    “I followed the whole conversation on https://www.linkedin.com/groups/3415770/3415770-6107811562578333699 as well. My answer on LinkedIn was: “QA can only be done by people; Software at best can raise attention to what might be an issue, but only people can tell if it is an issue indeed. If they are using a QA tool and they assume every flag is a red one, they are simply not using the tool correctly. If they come back to you with question, ask to be paid per question. For that amount of money they can also call a second translator to do a review, and it will be clear to them you did do a good job!”

    There are many very good posts on TheOpenMic, worth checking out:

    https://theopenmic.co/

    Like

  3. Agree the collision course is inevitable, which means that as more and more clients realise the shortcomings and risks inherent in the “computerised output optionally with subsequent cheap pseudo tarting up” model, they’ll increasingly look to human translators for their important materials.

    And that’s not just their technical or legal stuff. Its anything where a poor translation could come back to bite them or cause a negative consequence. Marketing materials, company and product information, their website, reports, dealings with customers & suppliers, etc, etc. For most organisations I’d imagine it would be most of their translation volumes. Or put another way, I think organisations would assess that most of their materials would be too important to go with a cheap and nasty, a few mistakes here and there won’t matter, translation.

    It’ll be a gradual process of client self-education leading to much greater care in who they select as their translators.

    Liked by 2 people

  4. Right, it’s not just the technical and legal stuff, but there are also many different kinds of technical and legal translations. When it comes to patent translation, it is possible to use cheap, low quality translation for what is called “prior art” just for basic information, and sometime even machine translations, if the prior art (existing technology) is not directed exactly to the same technical issue.

    But when it comes for example to prior art once a published patent has been attacked by opposition, new patent applications for filing, lawyer’s briefs, etc., these translations must be precise and reliable, otherwise they will be useless or counterproductive.

    I just wonder at what stage in what you call gradual process of client self-education we are now.

    Are we only at the beginning of the process, or are clients hopping mad already at what “the translation industry” is throwing at them?

    Liked by 2 people

  5. We’re a generalist translation agency, meaning we deal with both niche and demanding work (which calls for expert translators and careful preparation) as well as large-volume and simple content (often done with MT or MT + post-editing).

    Our approach is to offer options to clients and tell them: Listen, we can do this using this process or that process, that’s how much each costs and here are samples of what you’re going to get.

    From my experience clients are quite smart making their own decisions about what quality is acceptable vs the price. The ones who are not so smart are usually new to localization and become smart quickly after getting burnt once (usually going for a cheaper option against our advice).

    Like

    • Thank you for your comment.

      I’m curious to know what simple content is being dealt with by MT or MT + post-editing.

      Could you give us an example?

      Liked by 1 person

      • For something to be suitable for MT, it needs to be:

        1) large volume (at least hundreds of thousands of words)
        2) well-structured (text for-translation clearly identified and marked-up)
        3) limited vocabulary (something like Simple English is ideal)
        4) factual information (not creative)

        A good example is hotel descriptions for companies that allow world-wide booking.

        #1 – volumes can be in the millions of words
        #2 – clearly tagged: property name, description, room type, facilities, etc.
        #3 – vocabulary is limited due to the nature of the content. You’ll find similar phrases repeat: Queen Bed, lobby, Italian restaurant, etc.
        #4 – visitors don’t expect this type of text to be beautifully-written, they want to know the facts: Where’s the hotel? How many stars does it have? Is there a swimming pool?, etc.

        I have personally worked on such projects and clients were happy with the results, because they understood the limitations and knew that they wouldn’t be able to pay for human translation. I know others have had similar successes.

        I have also read about good results with MT of patents and I believe it’s possible, but haven’t had direct experience myself.

        Like

      • “I have personally worked on such projects and clients were happy with the results, because they understood the limitations and knew that they wouldn’t be able to pay for human translation. I know others have had similar successes.”

        Thank you, that makes sense. What I am wondering is why use a translation agency for an MT project of this type at all when anybody can simply run similar text through Google Translate or Microsoft Translator for free. Many people, such as hotel owners, probably do it by themselves because most of the information will be preserved, regardless of the occasional mistranslation bloopers.

        (I saw a good one recently, it said “Living nanny and not living nanny available here”).

        Liked by 2 people

      • Oh, I’m not talking about Google Translate, Bing Translator or any other free online MT tool. Using that would not have brought results nearly as good.

        What we used and what is normally used by professional providers are customized MT engines built and trained for the specific customer.

        What’s the difference?

        1. The engine is not a one-size-fits-all like GT, but rather it’s customized for a specific domain. For example, GT wouldn’t know whether “bank” in the source refers to a financial institution or a river bank. With a custom engine you can select “finance” as the domain of the text which will prioritize the former translation.

        2. You can train a custom engine by feeding it pre-translated bilingual corpora similar to the actual project content, which will result in translations similar to the training material. Another common way to train an engine is forcing it to follow a glossary/termbase of key terminology that’s been translated by subject-matter specialist (and human of course!) translators.

        3. You can run several rounds of training, meaning you translate a sample first, review the results to look for pattern of errors, fix those, and translate the sample again and again until you’re happy with the result; and only then translate the rest.

        There are plenty of other ways to train and tweak an engine to improve the output, but they’re not available with free online solutions like GT.

        So to answer your question: a client would work with an agency or directly with an MT provider to take advantage of the above benefits of a custom engine. Secondly, an agency can provide professional linguists who can help train, review and edit the MT output.

        Like

      • Thanks for your detailed explanation. It makes perfect sense to me to replace human translation in cases such as the one you described, if human translation is too expensive in the client’s opinion.

        But only if the clients understand and agrees with this solution.

        However, as I wrote in a post a few months ago linked below, there also seems to be a new business model type for translation agencies that is based mostly or only on machine translation edited by human sub-contractors who are paid 1 cent a word for editing chunks of text that were simply run through GoogleTranslate.

        I believe that in most cases, the clients are not even told that they are buying machine translation, once they are lured by the low, low rates into the agency’s trap. That is just one example of where I see an inevitable collision between the interests of the clients and “the translation industry”, and I also believe that this clash is an opportunity for highly specialized “boutique agencies”, as well as for highly specialized individual translators.

        https://patenttranslator.wordpress.com/2015/12/18/so-now-i-know-how-much-they-want-to-pay-us/

        Liked by 1 person

      • I’m glad if you found any of my explanations interesting or useful. There really is more to MT than Google Translate and Bing Translator.

        As for agencies that pay very little for post-editing very low-quality MT output and then offer it to clients saying it was human translation (I hate this term), that’s not OK. I don’t mind any translation provider using any process they wish (including MT), working with whoever they wish (including people with no translation experience), but only as long as they are transparent to what they are doing.

        I’ve read the other post you’ve linked to and can understand why you’d feel personally offended that someone would offer you such a low rate. You’ve worked hard to get the education and professional experience and you know what your time is worth.

        You are like a provider of fine wines and can sell your product at a premium price, but should cheap wines be banned from the shelves? I don’t think so. For some occasions, a cheap wine is perfectly fine and I believe it’s always better to give customers a choice.

        Like

      • I like your allegory very much.

        Let me offer my own, inspired by what you said.

        I think that companies selling edited machine translation are pretending that they are selling basically the same thing as real wine, which would be translation by an educated and experienced translator, but in fact, they are selling a whole range of potions ranging from jug wine to poisoned lemonade.

        Liked by 1 person

  6. Why has there been no sign at all of a “collision course” over the past five years? MT will also keep improving over the next five years just as it has over the previous five and ten.

    Like

  7. 1.

    “Why has there been no sign at all of a “collision course” over the past five years?”

    Are you sure about that? That’s not what I hear. But even if it were true, why were there no signs at all of any iceberg during the first part of Titanic’s journey?

    2.

    “MT will also keep improving over the next five years just as it has over the previous five and ten.”

    Yes, it will keep incrementally improving, just as it has over the previous five, ten and fifteen years. But there is nothing to indicate that it will achieve the stated goal, “machine translation that is as good as human translation”, within our lifetime and beyond.

    And there are very good reasons for that that, some of which are discussed in many posts on this blog.

    Liked by 1 person

  8. Hello.

    It is refreshing to read your view on this issues. Mainly on the third topic – I really do hope there is some sort of collision in the future to give new perspectives to the “translation industry.”

    I do a good amount of work to a relatively good agency. They are big but not huge and don’t offer cheap services as their way of attracting customers, which is a positive. But sometimes, more often lately, I see them trying to cut corners and biting more than can chew. Accepting big projects with silly deadlines, having little care about translation quality. I mean – I also do editing and proof-reading for them, and the translations that I have to work with are sometimes atrocious. Bad grammar, poor vocabulary, mistranslations all over the place. And then I wonder – how can they really assess quality if they don’t have a PM or quality checker on my language pair? One of those projects was a booklet for a huge global company, marketing material. I asked beforehand – “is this going to be from the same [crap] translator from the last time (if it is I will have to charge it as heavy editing)?” And they replied, “no, the client really trusts this translator, it should be ok.” It wasn’t ok, I was quite shocked that the client “trusted that translator” because I would not trust anyone to work for me with that poor level of language skills, mainly for communicating my brand at global level. So, was the agency lying or does the client really not care about how their marketing comes across (or even: neither has a clue about the work)? I sincerely don’t know.

    I guess the big issue is that this business model that sells translation as an affordable [cheap] commodity might wear thin when some people realise that cheap can become expensive and “there’s no such thing as a free lunch.” I know some companies have learnt this the hard way and probably have turned to professionals or boutique agencies* for a better experience, but there is still a lot of aggressive rhetoric out there about translations tools and cheapness of a specialised service. Of course, there will always be cheap costumers for cheap agencies or service providers, but I hope that the better clients start to see translation more like what it really is – an important communication tool, not an afterthought that can be a bargain.

    I also hope, as with everything that has too many competitors fighting for the same slice, that a natural selection process will take place and reorganise the market in an organic way. I hope, maybe I am being naive.

    *When searching for boutique agencies that I might try to work for I have realised that some not so specialised ones are selling themselves as such without actually being so – offering too many language pairs, too many specialisms in a ‘small scale.’ I guess this shows that this business model might be gaining ground, but then, of course, there are people misusing this label to take advantage of it.

    Liked by 1 person

  9. 1. “So, was the agency lying or does the client really not care about how their marketing comes across (or even: neither has a clue about the work)? I sincerely don’t know.”

    I don’t know either, of course, but my guess is that since the client has not complained so far, probably because they don’t know what is in the translation either, and the translator is really cheap, the agency told you something that they consider a white lie. How could they possibly say to you – yeah, well, maybe it’s crap, but we can’t really tell, the translator is cheap and the client pays on time, so everything is fine.

    That is what I mean by “collision course”, because at some point, the client may realize that a poor translation has done harm to the company’s bottom line based on feedback from their clients, etc.

    2. “When searching for boutique agencies that I might try to work for I have realised that some not so specialised ones are selling themselves as such without actually being so – offering too many language pairs, too many specialisms in a ‘small scale.”

    They all try to do that. But it is a pretty transparent gimmick: If you take a look at what an agency specializes in and it is literary everything, it’s clear that they don’t specialize in anything, including to clients who have a well functioning brain.

    Liked by 1 person

    • What I am wondering about, is what the previous reviser did with the previous translation done by this terrible translator.

      Either the previous translation needed heavy editing, the reviser reported it to the intermediary – in which case the intermediary is lying to the second reviser.

      Maybe the first reviser refused to revise a second time such a bad translator’s work, which is why the intermediary needed another reviser, by the way!

      Or the first translation a) was not revised or b) was incompletely revised AND the customer has not read (or understood) the first translation and has not received feedback from business partners yet.

      Another possibility is that the first translation was revised but did not need corrections, but this is not very likely.

      Like

      • “Another possibility is that the first translation was revised but did not need corrections, but this is not very likely.”

        Actually, I believe that unnecessary, counterproductive and outright harmful corrections and revisions are pretty much guaranteed based on the way the work is organized by translation agencies who do not have any particular knowledge about the field or the language of the translation.

        If you are a proofreader who is proofing five thousand words of a really excellent translation and all you do is correct two or three typos, how do you justify a substantial amount on your invoice, relatively speaking, if you have not made any real changes in the translation.

        So, naturally, most proofers will make changes to make sure that their invoice will be accepted, and that they’ll get paid and next jobs will be sent to them.

        Liked by 1 person

  10. Andovar stated above that at times a client makes the wrong decision in choosing MT or MT + editing “become smart quickly after getting burned once.” This prevents any widespread “collision course” scenario. If it hasn’t happened by now, I don’t see why it will happen in the future.

    By the way, Google Translate is turning ten this year but is only eight years old as a statistical machine translation tool.

    Like

  11. I think I know who you are, Zepher.

    You need a new pseudonym again.

    Like

  12. I have read (and tweeted) this interesting and reassuring article dated December 6, 2015 and entitled “7 Fastest-Growing Industries to Invest in for 2016” : http://www.gobankingrates.com/investing/7-fastest-growing-industries-invest-2016/

    So these are 2 pieces of advice to investors seeking to invest (create an agency) in the translation industry:

    1) “For 2016, “investors should also keep these two things in mind: ***innovators appear all the time trying to leverage machine translation, but breakthroughs are elusive*** since at its core, [the translation services and technology industry is] ***a very human-intensive industry***,” said Stempniewicz.”.

    2) The sentence that follows is interesting too, even if it is not in line with the present subject:

    “And the developed world is reaching saturation, but emerging market languages are under-served.”.

    Like

    • “A very human-intensive industry”, interesting formulation.

      Liked by 1 person

      • Yes, it is used as opposed to “capital-intensive” (http://www.investopedia.com/terms/c/capitalintensive.asp) and is a synonym of “human capital intensive” and, more common, “labour intensive”, cf http://www.tutor2u.net/business/reference/capital-and-labour-intensity

        If you google these expressions, you will find plenty of hits.

        They are normal economic expressions to describe the different types of industries.

        Labour intensive sectors have higher costs than capital intensive sectors, normally.

        Translation is a labour intensive activity.

        Lots of capital has been injected into CAT tools and machine translation, but translation remains primarily a human/labour intensive industry.

        You might want to use these expressions again in your posts!

        They are in the public domain: no charges! 😉

        Like

    • Hi Isabelle,

      Since you’ve quoted me, I’d like to clarify what I meant in that article.

      1) “investors should also keep these two things in mind: innovators appear all the time trying to leverage machine translation, but breakthroughs are elusive since at its core, [the translation services and technology industry is] a very human-intensive industry,”

      While I am a firm believer that MT has its uses (an example of which I gave in my previous comments), I don’t consider the latest developments to be breakthroughs, but rather evolution. MT itself is getting better all the time and people are getting better at customizing the engines, which results in improved output overall.

      As for “very human-intensive industry”, I meant to say that regardless of the progress in MT and language technology in general, most translation requires human knowledge, skills and input. Even in the case of MT, rarely the raw output is published without post-editing or other involvement from linguists and translators. So I used “human” to differentiate from “machine” in that expression. I did not look up the term in Investopedia or anywhere else. It’s just the expression that came to my mind. Since I’m not a native speaker of English, I fully accept it may have not been the most accurate wording.

      2) “And the developed world is reaching saturation, but emerging market languages are under-served.” I stand by this statement. My company has offices in Southeast Asia and Latin America and we work with many emerging market languages, and I see the following. The “old” languages that most translation has been done in (English, FIGS, CJK, etc.) are well-served, and supply and demand is growing steadily.

      On the other hand, languages of emerging markets (SEA, Latin American Spanish and Brazilian Portuguese for example) are growing faster and in my opinion will continue to grow along with the economies of their mother countries. I wrote a blog post about this if you’d interested: http://www.andovar.com/localization-next-one-billion/

      Like

      • Dear Andovar – Jacob (@andovarglobal), a) I did not quote you, as far as I remember, and the quoted article is not from you. The author is indicated in my quote. b) I do not intend to lose more time reading your apologies of MT. I do not consider you a professional of the translation sector, but rather one of those too numerous sharks trying to make a buck out of enslaving intellectuals. So long. Isabelle F. Brucher

        Like

    • Oh my…

      Anyway, just to clarify, the quotes in the article http://www.gobankingrates.com/investing/7-fastest-growing-industries-invest-2016/ are indeed from me:

      “Jacob Stempniewicz is the vice president of marketing at Andovar, which offers translation and localization services. He agrees that the industry is fueled by international business, which calls for communication with customers in their languages.

      For 2016, “investors should also keep these two things in mind: innovators appear all the time trying to leverage machine translation, but breakthroughs are elusive since at its core, [the translation services and technology industry is] a very human-intensive industry,” said Stempniewicz. “And the developed world is reaching saturation, but emerging market languages are under-served.”

      Have a nice day,

      Jacob

      Like

  13. I think that “a very human knowledge-intensive industry” would be a better description.

    When I hear the term “a very human-intensive industry”, prostitution or smuggling of illegal aliens comes to mind.

    Liked by 1 person

    • Yeah. The standard expression is “labour intensive”, though. You might want to use this expression instead! 😀

      Like

  14. Yes, I know the “labor intensive” formulation, but did not know the “human intensive” one. It is kind of more intense that saying just “labor intensive”, I think, that’s probably why it is used now.

    Like

  15. Last but not least, I think that “the translation industry” (so-called LSPs & MT software producers) have given up trying to produce perfect translations with MT. Yet they keep perfecting the system because it does serve as a pretext to huge savings – whether they put the difference in their pockets or make their customers benefit from these savings.

    The more intermediaries save on translation, the more they can invest into their salesforce… This way they get even richer and richer and richer (and richer)!

    Like

  16. The latest trend is to offer specialized MT services, by business sector: all the standard terminology of a given business sector has already been fed into the machine, so as to improve the blablabla it spits out.

    Even the European Commission is acting like this. This morning I read a recent article explaining that cross-border commercial complaints are very numerous but are blocked by the language barrier between the customer and the supplier.

    So the EU Commission has just started to provide a multilingual platform where negotiations will be translated by their own machine translation system.

    Only the final outcome will be translated by their abundant internal translators teams, on their standard budget of 330 million euros per year.

    This is the article: https://slator.com/demand-drivers/eu-commission-to-pay-for-translation-when-online-shopping-goes-wrong/ : “EU Commission to Pay for Translation When Online Shopping Goes Wrong” :

    “On February 15, the European Commission launched its Alternative / ***Online Dispute Resolution (ADR / ODR) platform***.”

    “Should the ODR catch on, the numbers might be staggering. According to an article by the BBC, ***Ebay’s dispute resolution system settles 60 million disputes between small traders every year***. That would be a lot of work for the [[EU Commission]] Translation Centre.”.

    I recommend to activate Google Alerts (https://www.google.com/alerts) on key words like “translation” and “translator” in all the languages you understand and to follow the latest trends weekly or even daily.

    It mostly talks about the latest machine translation inventions…

    Like

  17. It’s money that makes the world go round …. not common sense.

    Liked by 1 person

    • The translation market is infested by non-translating intermediaries.

      I suggest to systematically call them like that – “non-translating intermediaries” -, in order to put them back to the place where they belong (mentally) (and they do obviously read your posts and the comments below them, since one of them lead to a major change in Trados Studio: SDL did add AutoCorrect to the 2015 version!), instead of their pompous/bombastic “LSP” appellation (which makes us what: “non-language service providers”, by any chance ??).

      I have to catch up with your other recent posts… but it’s already 9pm in Belgium. And I have more time to read on week-ends. And since your blog posts are quite long, I guess I can only read one per week-end (including reading all the comments and maybe commenting too : takes half a day !). 🙂

      Liked by 1 person

  18. I just call them translation agencies. I don’t have anything against the honest ones, I do believe there is a need for agencies in this business, especially since part of my income is from working as an agency.

    I may be a non-translating intermediary too sometime, but I do know what I am doing, how to pick the right translator, how to proofread the translation, etc.

    And unlike many translation agencies, I don’t need to pretend anything.

    Thanks so much for reading my silly posts, Isabelle!

    Liked by 1 person

  19. Well, I meant “non-translating intermediaries” as “intermediaries who do not have internal translators” and thus often do not have the faintest idea what translation is all about.

    You might not translate each project, but your company is a “translating intermediary” in the sense that it has an internal translator (yourself), and so your firm’s people (you) know what translation is all about.

    Liked by 1 person


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

%d bloggers like this: