Posted by: patenttranslator | February 22, 2011

Another Missed Opportunity to Say Something Meaningful About Machine Translation

Every time I read another article about machine translation, I have the same feeling of futility that I experience also when I listen to John Lennon sing “A Day in a Life” (I read the news today, oh, boy, the English army has just won the war …..).

All of these articles have interesting anecdotes about unlucky users of MT who perhaps expected too much, and many of them start with one such anecdote as did this article by Konstantin Kakaes in Washington Post. His blog has the following introduction: I’m a freelance journalist. I have many interests, though these days my focus is writing about nuclear proliferation, science, technology and the world. I also write about Latin America. It does not say anything about what he knows about foreign languages and linguistics. Perhaps he knows some Spanish. That would be an improvement.

All of these articles in what I call corporate media, because I think that is the proper term for it, are written by journalists who have various backgrounds, often impressive ones. But none of them has a background in languages and linguistics. These journalists then interview users and developers of machine translation, but they are not interested in talking to people who translate for a living. I was once interviewed by phone on this subject by a journalist in Canada about 8 years ago, but that is the only exception to the rule that I can think of. Apparently, human translators have nothing of importance to say about machine translation.

The recent crop of articles tells the readers that everything changed with the advent of Google Translate, which is now here, as Kakaes puts it, “to remove humans from equation”. Right. Let’s use mathematics to find the magic algorithm that will eliminate the need for human brain. Kakaes says among other things in his article that the late Frederick Jelinek (which is a Czech name), who pioneered work on speech recognition at IBM in the 1970s, is widely quoted as saying: “Every time I fire a linguist, my translation improves.” I suspect the late Frederick Jelinek was firing linguists because they were telling him 40 years ago something that he did not want to hear, namely the same thing that I am saying in this blog.

“I am sorry, but machine translation will never work, Herr Jelinek.”

“Vhat did you say? You are fired, you damn ingrate.”

It does not take a genius to figure out that the statistical approach pioneered by Google will not really work either. The way the commercial propaganda machine writes and talks about it, real machine translation that will get rid of people like me and the readers of this blog is just around the corner. It has been just around the corner for quite a few decades now.

But machine translation is not really translation at all, although it may look like one, and never will be because no algorithm will obviate the need for the concept of meaning in translation. Trillions of words in a database are really nothing more than a huge haystack hiding what is missing and always will be missing in machine translation – the meaning of the words. You need a human to make sense out of things. Google is such a great engine because thousands of very smart human programmers analyze and update links to information every second of every hour of every day. If Google stopped doing that, it would be out of business within a few weeks.

But you can’t really put everything that humans say now and will say in the future into a huge database that could be used for machine translation by Google Translate in the same way that Google the search engine can be used. People who got used to the miracle of Google the search engine naturally expect this to happen one day soon with Google Translate. Only it never will. We are all unique. We all say things that nobody else has ever said and possibly never will …. we don’t do it all that often, but we all do it. The human brain is not a database. I don’t know what it is that makes it work the way it does, nobody really knows, but I do know that it’s not a database that can be updated just like a search engine. Because language is what it is, the most likely equivalent to a sentence in another language can be correct … or completely incorrect. Probability is not a replacement for meaning. And machine translation will never break the barrier of meaning.

30 years ago, you had to pre-edit and post-edit every machine-translated sentence, otherwise it would make no sense. In the second decade of the 21st century, you still have to pre-edit and post-edit everything, except for really simple sentences, see the example of an international lawyer who uses MT for translation from English to Chinese in the Washington Post article. But when you make a conscious effort to use only short and simple sentences, this is really pre-editing too.

Machine translation is getting better at simple tasks like this, and the statistical approach may be more instrumental than linguistic analysis. I don’t really know that much about it as I prefer to spend most of my time doing the real thing …. translating.

This blog post is too long, I am tired and I am going to finish it now. But I would like to pose a question here.

Do you think that it is possible to create software, similar to machine translation software, that would write steamy romance novels that women would actually be buying and reading?

And if not, why not?

It’s just stupid words on a page. Just like a translation.


  1. Where to start….

    1) At some point you might want to discuss how a computer just beat the best humans in the history of the quiz show, “Jeapordy!” You have argued that language requires thinking, and computerrs can’t think, so doesn’t this imply a computer couldn’t beat the two elite humans in a game that contains riddles and puns? Yet it did beat them.

    “It’s just stupid words on a page” or a screen in this case. So how did the computer beat the two champions?

    2) People have not been saying MT that will replace humans is right around the corner for decades. Two computer scientists said that decades ago.

    3) Your test was just one sample. You also tested before Google gained access to 1.5 million patents at the EPO.

    4) You haven’t explained how Google Translate passed what a Portuguese translator said was the equivalent of an ATA exam 1 out of 3 times in 2010 without arguing that either a) she wasn’t good enough to judge or b) the test must have been too easy. They were not “simple sentences” that MT translated at that level as you claim is only possible with MT. Or did you mean only J>E?

    5) There was no Chinese MT good enough to use pre-editing and post-editing in 1981 as you infer with the example of the lawyer in the article.


  2. Jeb, quiz shows are the sort of thing that enable computers to show their strengths. Anything that can be put into database format is ideal for computers. My computer has a database containing the source and target sentences for almost all of my translations for the last 11 years. If I ask it whether I have ever translated words such as “Satteldach” or “Verfügung”, and how I translated them in each context, it has a fantastic memory. Far better than my human memory. But it is no good at telling me what is best in my present sentence – it can only guess because it hasn’t a clue what the sentence actually means.

    I am not familiar with the quiz show you call “Jeapordy” (Google seems to prefer “Jeopardy”, but hey, that’s just a computer), so I don’t know how the questions were phrased and what algorithms the computer could have used in its responses, but I am not surprised at a computer beating expert humans in anything that can be formulated in databases plus the appropriate algorithms. Chess is another similar pursuit. But human language is more than another data set to be solved by database plus algorithm. Sure, Google Translate and others can go some of the way by pattern matching, but they still haven’t a clue what the text is actually about.

    On another note (@PatentTranslator)
    I asked my computer about the software-generated romance novels you mention at the end of the post. She said she can’t wait to read them.


  3. “On another note (@PatentTranslator)
    I asked my computer about the software-generated romance novels you mention at the end of the post. She said she can’t wait to read them.”

    As long as your wife does not find out, it should be OK.


  4. Another matter too seldom mentioned as a potential problem for MT is the evolution of the language and its use. I find that many records from a decade ago in my TMs use language that is simply not current enough for certain fields.

    But sounding stilted or archaic is a minor problem compared to the gibberish and more dangerous sentences that sound right but aren’t. However, unlike most translators, I actually enjoy the spectacle of MT and hope it is widely adopted, especially for safety-critical applications. This would be of great benefit to my friends in the legal industry and might finally get me to revisit an old ambition of becoming a tort lawyer.


  5. Translators may be the only people who really, really like lawyers.

    I know I do even though sometime they are a bit overbearing.

    Without lawyers, I would have to have a real job!


  6. Hello.

    What do we mean exactly when we say human language is more than a dataset”? What exactly do we mean with the word “meaning”? I guess only men are working on TM development ;>))), men who try desperately to push the uman language “use” in a scientific, countable dress (in economic contexts). Language does not exist on itself. It exists only in the presence of at least two uman beings that want or must communicate with each other. Language use has it’s place amongst playing music, dancing and painting. Children do all these things, including “speak” and later “write”. The difference with language use is that growing up, we all “speak” and “write”, while not all umans continue to make music, to dance or to paint. Adults push dance, music and paint on an adult and professional “competence” level. In very few circumstances they recognise the adult “competence” of language uses: written or spoken. They accept that writing is a competence in the context of “book writers”; speach writers, song writers are recognised to have or to need a specific competence in writing. For all the rest, they believe that they “can speak and they can write” for the very fact that they are grown up and learned to speak and to write when they were children. So all the adults in the world, in every kind of professional context, are daily “writing”, assuming that they are as “professional” when they write as when they “perform” their specific profession…This is absolutely untrue, but nobody “likes” to admit that. The very fact that a manager is a manager, makes him believe that as easy it is for him to write his own things down, as easy it is for a translator to translate what he has written.
    Even letting aside the problem of correct grammar and word choices, when we are writing, we express what is “hidden” in our mind for others. “Writing down” something what you have in your head is extremely risky. You will “forget” to write for instance, about what for you is “obvious” (you assume that it is visible in the head of the other). Wanting or not, writing is also expressing ones feelings. Even in “feelingfree” contexts. Writing, you are communicating your own feelings, your own thoughts and what “you” retain necessary to write. You will know what you have expressed, or not expressed, only when your reader tells or asks you. Thus, when I read, “I” will give a meaning to the words I’m reading; a meaning based on my own thougts and feelings and the context in which I must read a text. This means that untill now, we have two meanings… the one of the writer, expressed and NOT expressed, and the meaning given by the reader. So language becomes an ever “failing” mean of communication between umans that use other language (words and gestures and so on) to explain the text that they have written and read…. Who can reasonably hope to obtain from a computer to do what umans can’t without thinking, choosing words, letting out some and putting in others, trying not to create interferences such as feelings, misunderstandings when we “write”? Hoping that a computer can translate, forgetting the basic failure of language as a communication mean? As long as writing in each professional context is not considered as a specific competence, translating will be considered in the same way: it can be done by a computer, in the same way as “each uman that has learned to write can write”. Waiting for Godot… only a translator knows that when he translates, he is as much as possible in the head of the writer, in the collective head of a culture and in the head of the reader with another collective culture head…. but if you say that to a “writing” manager, he will not even listen to you. Unless the “translation” in his business is becoming “a problem” (= costing money) in a way he did not expected. I have one customer who is admitting right now that the “problems” with his foreing partners are caused by “his texts” (= not by the translations!) And another who has started to consider as much as possible, the different structures of the texts in the languages in which he needs his texts to be translated. Admitting that this is a complete surprise for him. ;>)))


  7. Hi Frauke:

    Thank you for your poetic contribution.

    Yes, you are right, it’s mostly about money, like just about everything else.
    But it’s not just men who has made this world into what it is.

    Women are not exactly angels either. (I know you were mostly just kidding ….)

    So I take it you would probably not buy and read a romance novel written by a computer.


  8. Notice that Steve won’t respond to any of my points. Just as he said on his May 2010 post “I’m not going to respond to your questions 1 to 4”

    Because he can’t.

    Full stop.


  9. Nobody can respond to your points, Jeb, because you don’t have any.

    And what is “Jeapordy”?


  10. Reading between the lines, my robot says tread carefully here, coz the atmosphere is hotting up.


  11. That’s what it’s all about.

    I don’t know if you have the same thing in your country, but I used to love watching my kids play the hokey pokey at the skating ring:

    You put your right foot in,
    You put your right foot out;
    You put your right foot in,
    And you shake it all about.
    You do the Hokey-Pokey,
    And you turn yourself around.
    That’s what it’s all about!


  12. I realize it is easier to snipe at a spelling error than address the issues, but I think most translators understand something powerful is evolving if Google Translate scores as well as a human on an ATA type exam 1 out of 3 times with Portuguese.

    I could take 1 out of 3 and show a translator , and he would not be able to tell which was the computer and which was the human. Yet Steve claims one is a translation and the other isn’t.

    One is free and the other isn’t. Most will take the free one.

    But the computer isn’t thinking, so what’s goign on?

    (Notice I kindly added a typo for people to snark over)


  13. You and your ATA test.

    Give it a break already, Jeb. I would not touch an ATA test with a ten foot pole, OK?
    Real translators have real accomplishments, like a university degree in Japanese studies. A diploma like that is a real test, as is the test of time. The ATA test is a joke. I am not particularly surprised that a machine passed it.

    If you have something worth discussing, I will respond. Otherwise, I will not. There is no law that says that I have to respond to every comment.

    And if you make pretty imaginative typos in your posts, don’t be surprised when people call you on it, man. It’s a mean, mean world out there.


  14. It’s up to you if you prefer to make petty comments.

    You wrote last May, “By definition, a machine or a piece of software will never understand the meaning of anything. And you cannot translate if you don’t understand the meaning.”

    Despite you claiming the ATA test a joke, Google still translated at a human level. So how did that happen without it being able to think?

    By the way, a Japanese degree in no way indicates that someone is ready to translate, and many good translators do not have Japanese degrees.


  15. @Jeb: “Notice I kindly added a typo for people to snark over.”

    Thanks for your kind consideration. I über-snarked this by introducing a couple of deliberate typos into the penultimate paragraph of your later reply, just to see if Google could be smart enough to catch me out.

    So instead of “Despite you claiming the ATA test a joke, Google still translated at a human level. So how did that happen without it being able to think?” my version now reads:
    “Desipte you cliaming the ATA tset a jkoe, Gogole stlil transalted at a huamn leevl. So how did taht hapepn wiothut it bieng albe to thnik?”
    As you see, I have switched no more than one letter pair per word, and only internal letters and only in words of four letters and more. A human reader can read my version, perhaps with mild surprise, but the mistakes would not throw a human reader off the scent.

    Google actually called me out on four of the 15 typos and asked if I mean:
    “Despite you claiming the ATA test a jkoe, Google stlil transalted at a huamn leevl. So how did taht hapepn wiothut it bieng albe to thnik?” (That is almost the “one in three” rate you quote). It then proceeded to translate the 4/15 corrected sentence into German thus:
    “Trotz der ATA Sie behauptet Test eine jkoe, Google stlil bei einer huamn leevl transalted. Wie kam es, dass hapepn wiothut bieng albe thnik?”
    I challenge any German speaker to make head or tail of that.

    Perhaps this was too harsh, so I gave it the paragraph as you wrote it, and it came up with:
    “Trotz der ATA Sie behauptet Test ein Witz, Google noch in einer menschlichen Ebene übersetzt. Wie konnte das passieren, ohne sie in der Lage zu denken wird?”
    Oh dear, two sentences, and neither of them make sense in German (even without the inteligence test of spotting typos). But then, you said one in three would be a success. Any idea of the third sentence that Google would “understand”?


  16. I see translations of my blogs into various languages, I presume from Google, on my dashboard quite frequently. They look really impressive when these are languages that I don’t understand, such as Chinese, Italian and Portuguese. But if it is for instance in German, Czech or Polish, I can see that the result is more often then not really ridiculous.

    But MT is still very useful, of course. My brother uses Google to read my posts once in a while in Czech because he does not speak English, although he keeps complaining that he can’t figure it out. But without MT he could not read them at all.

    Incidentally, there was one sentence that was translated perfectly and he was very excited about it. The sentence said: “Although Germans are very civilized and polite to each other in Germany, they keep yelling at each other for some reason when they are abroad”.

    “That’s exactly right!” said my brother when we were talking later on Skype.


  17. @ Jeb
    I’m afraid you missed the point, Jeb–humans are creative, computers aren’t, and never will be.

    I promise never to yell at you, should we ever meet abroad.


  18. @Volkmar:

    I promise the same. However, Germans (and possibly Austrians too) only yell at their compatriots in German, I think. They seem to quiet down when they start talking to foreigners in English and I would be a foreigner who speaks English.

    Incidentally, the Japanese have a proverb, which says something like “If you do something away from home, it stays away from home”.

    I think that a good translation into American English would be “What happens in Vegas stays in Vegas”.


  19. Portuguese, Victor, Portguese. Not German. Portuguese.

    Computers aren’t creative yet it translated at a level in PORTGUESE (not German) that was equal to a human. Pretty cool, eh? German is coming, then Japanese.

    Funny, if MT was such a joke, why are translators talking about it so much in the last couple of years. It isn’t just the media.


  20. Hi Jeb,

    Throughout your posts, you just repeated and relied on that one piece of evidence from the Portuguese translator. I have no access to that presentation and so cannot judge for myself the validity of the claim. Even if it is valid, it is quite a jump to say that Google Translate can translate at the human level in Portuguese. Google may happen to have those particular translation stored in its database, which had been proved to be the case in previous MT quality contests. If you work in science, you know that any scientific findings have to be empirically verifiable. Have you seen any other independent reports supporting that claim? If not, please do not place all your hope, argument, enthusiasm and prediction in that one piece of “evidence”. There is actually counter-evidence to that claim: if Google Translate can translate into Portuguese at the human level, why are there still Portuguese translators? Should they have all lost their jobs or work as editors? You can email Cris Silva and see whether she is doing more translation or editing.

    You are not the first person to be so optimistic about Google Translate. A localization guru predicted at the end of 2009 that Google Translate would replace human translators in two years. Well we are pretty close to the deadline now, but Google Translate has remained, for the most part, a joke for commercial quality translation. Can you use Google Translate to translate “the chicken is ready to eat” into a language you know other than English? It is complete gibberish in Chinese. Even the worst Chinese translator can do better than that.

    It is amazing that, of so many MT systems, you picked Google Translate as your source of optimism, while other MT practioners are distancing themselves from it because of its poor quality. They play down its bad quality by claiming it is not customized, and therefore cannot represent the state of art in MT. They have already admitted that even highly customized, domain specific MT systems are no match in quality for human translation at the moment or in the near future. You can read more on the LinkedIn Automated Language Translation group discussions, where people discuss with their real identity instead of hiding behind a web alias. Unless you have something more up your sleeve than that unverified piece of “evidence”, your argument and predictions are very weak, to say the least.


  21. Hi Frank,

    1) It is just common sense to say that if Google Translate put out Portuguese>English that a qualified Portuguese translator says meets the criteria of an ATA exam pass level, then Google Translate is translating at the human level 1 out of 3 times. That was two years ago. It will rise to 1 out of 2, then 2 out of 3, then 3 out of 4, then 9 out of 10, and 99 out of 100 very soon as computers get exponetnially powerful. For free.

    2) Actualy, I didn’t raelly choose GT as my source of MT optimism. I simply used it as an example. Then again, Google is a $200 billion company. Pretty deep pockets for improving GT. Oh, did I mention GT is free? Some other system could be better tahn GT in 2013, but I’d bet agaisnt that.

    3) Google Translate doesn’t customize yet. My guess is that it doesn’t feel the need to at this point. But Google got those 1.3 million patents from the EPO, so it looks like customization is coming.

    4) GT turned “The chicken is ready to eat” in Japanese to 鶏は食べて準備ができています。 That is not gibberish.

    5) You wrote: “There is actually counter-evidence to that claim: if Google Translate can translate into Portuguese at the human level, why are there still Portuguese translators? Should they have all lost their jobs or work as editors?”

    But this is as nonsensical as Steve claiming he should be out of a job by now if MT is a threat. It is like saying in 1900 “If people can fly, why can’t we?”

    And translators are already turning into editors in Spanish, Portuguese and French. We are at the cusp of this, so I presume Chris still trasnlates. Let’s see if she is in 2012 and 2013.

    One person complained to the Translator’s Journal that he thought he had a $5,000
    tech translation job lined up but the company decided to use GT and an editor. Of course, the TJ staff told them that the company was probably lying. Perfectally natural to rationalize, but it is still rationalization, right?

    6) Japanese/Egnlish is further out, but not at all immune to MT. The J/E translators will turn into editors within a few years. We are close to the end of an era.


  22. (Reposted with permission)

    Hi Steve,

    Nice to see that you’re still alive and kicking.

    I follow your very enjoyable blog, and I’m currently following the discussion under “Another Missed Opportunity to Say Something Meaningful About Machine Translation”. Here’s something you might like to try:

    It’s very easy to verify Google Translate’s current performance in Portuguese > English for yourself, even if you don’t speak Portuguese. Go to to get, you guessed it, a list of newspapers in Portugal. Pick one that puts its content online, then run an article from the latest edition through GT into English. Such an article is likely to be too new to have been translated anywhere yet.

    Here’s a sentence from an article in yesterday’s (February 25 2011) edition of Público:

    Os magistrados de Milão decidiram dar seguimento às acusações e o julgamento de Berlusconi irá começar a 6 de Abril, depois de a juíza de instrução Cristina Di Censo ter considerado que há “provas evidentes” de que Ruby manteve relações sexuais com o primeiro-ministro quando era menor de idade em troca de dinheiro.

    And here’s GT’s translation:

    Magistrates in Milan decided to act on the charges and trial of Mr. Berlusconi will begin April 6, after the judge of instruction Cristina Di Censo considered that there is “clear evidence” that Ruby had sex with the prime minister when HE was minor in exchange for money.

    The capitals for emphasis are mine. The rest is pure GT.

    Feel free to repost if you like.

    Keep up the good work,


    Hi Marc:

    It must have been about 12 years, right?

    What are you up to these days?

    By the way, the GT sample from the previous comment: “The chicken is ready to eat” was translated to Japanese to 鶏は食べて準備ができています。 That is not gibberish.”

    was translated into Japanese as “The chicken is eating and the preparation is done.”

    With Babel Yahoo fish, the MT translation into Japanese was 鶏はインスタントである (The chicken is instant).

    That is gibberish in both cases. Typical, stupid MT gibberish, the same as it was 30 years ago and it will not be much different 30 years from now, because computers only replace words by other words without any understanding of the meaning of these words.

    If a computer can pass the ATA test in any language, it only proves to me that there is something really, really wrong with the ATA test.


  23. Hi Jeb,

    If you choose to hang on to unverified evidence and place all your hope in it, please go on. A decade ago, MT enthusiasts made the same claim and, guess what? Translators are still translating. It is MT products by companies like L&H and IBM that faded away. That localization guru who is more optimistic than you are is going to eat his own words 10 months from now. I wish you all the luck while waiting.

    Your last post reveals what (little) you know about MT, besides your blind faith in it. GT is not the only free MT. So is Moses, a more advanced system with much greater flexibility and customizability. Who in serious MT solutions cares about GT? Google has a deep pocket, but it does not prevent it from flopping time and again. Still remember Goolge Wave? Where is it now? “Some other system could be better tahn GT in 2013, but I’d bet agaisnt that.” Go and check out Asia Online, or Language Weaver. I am sure they will tell you they are much better than GT in output quality at this moment, right now. And they have proof for it.

    Your example of customization of GT “with 1.3 million patents from the EPO” shows you have no idea what customization means in MT, or the trends in improving MT quality. It is not stuffing the database with more raw data. Follow this link and you will see how a professional MT company customizes its MT engine. Here is some more expert’s view on how adding data may not further improve GT quality: I guess you are unware of such issues. But, if all you care for are the two anecdotes cited time and again through your posts, you can go on living in your imagination. I will stop here. No meaningful discussion is possible with a closed mind. Discussing MT with Kirti Vashee, Jeff Alan and Alon Lavie is much more educational and interesing.


  24. Whistling past the grave yard.

    Can we stop with the “30 years ago…” and ” a decade ago…” bullshit? Steve, with
    no science or computer science (or logic, apparently) background is telling us what computers will be able to do in 30 years. What a joke.

    The Portuguese example above has one error — he instead of she. No MT was capable of anything close to that in 2005.

    Google does have deep pockets… $200 billion deep. They are a top 10 company in the world, and they are very serious about MT. Translators don’t have a prayer.

    Steve’s translation of the chicken example was wrong. Bable Fish was gibberish, but those who know Japanese understand that Google was off, but far closer. It does not say “the chicken is eating” Stick with patents, Steve.

    Translators are being turned into editors in the romance languages. By the way, Language Weaver was bought last year by SDL. Interesting to see what will happen as GT steamrolls ahead.

    I still haven’t heard Steve’s response to Watson beating the two all time Jeopardy champs. Computers dont understand language. So I guess Jeopardy isnt a language game.


    Oh, Jeff Alan seems like a really nice guy, but he is selling MT. Im not, so I dont care if GT, Moses, or Mother Jones is the best machine translation. But Jeff is worried about Google, or he is in denial. I doubt his company is worth $200 billion.


  25. “鶏は食べて準備ができています。 That is not gibberish.”

    It is gibberish. It does not mean “the chicken is ready to eat”.

    I asked a native Japanese speaker whether 鶏は食べて準備ができています makes sense to her and she said “意味が全然通じない (it makes no sense whatsoever) and started laughing.

    Either you don’t know Japanese, or you are simply lying.

    You must think that people reading this blog are total morons, including myself.
    That is why I don’t respond to your posts, although the entertainment value of your numerous and really creative spelling errors cannot be denied.

    You are either very ignorant, or you simply lie.

    It’s a waste of time to talk to people like you. I’ll try not to do that any more.


  26. Wait, you are the one who translated that into “the chicken is eating and the preparation is done” even though that is clearly wrong. Reread my post. I didn’t write that it was “The chicken is ready to eat” My point was that GT was getting much closer than BF that put out “The chicken is instant.”

    You are the one who belittled Chris’s Portuguese abilities when I doubt you have knowledge of her skills.

    You are the one who belittled the ATA exam, when you have never taken it in Portuguese. I doubt you’ve taken it in Japanse, either.

    You are the one who wrote that 30 years ago people pre editing and post editing Chinese MT when that didn’t exist in 1981.

    You won’t respond to the deeper issues of MT and human language, so you just point out spelling errors. For example, if computers can’t understand language, and they can’t for now, how did Watson beat Jeopardy’s all time best players? The human players weren’t thinking either? Those weren’t English questions and answers? English isn’t really a language? What?

    I don’t think you are a moron, but you have to admit your question to anonther person: “Then how would you answer the question in my article, how is it possible that I am still in business?” was pretty moronic.

    Still, I can tell that you will be a fine J/E editor in coming years.


  27. […] This function of MT as a very useful tool is an aspect that is almost never mentioned in articles that are published every now in then in newspapers because these articles are written by authors who do not really understand the issues involved, and who usually simply describe juicy anecdotal evidence obtained from users of MT, seasoned with unhealthy propaganda from  MT developers, without bothering to talk to real translators who actually may know a lot about MT because they understand the issues and because they are using MT all the time in their work. A typical example of such article is linked here in my previous post on this subject. […]


  28. […] neck-deep in science fiction. I think Steve Vitek said it quite eloquently, however, on his blog when he asked “Do you think that it is possible to create software, similar to machine […]


  29. […] I’m neck-deep in science fiction. I think Steve Vitek said it quite eloquently, however, on his blog when he asked “Do you think that it is possible to create software, similar to machine […]


Leave a Reply

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

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

Facebook photo

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

Connecting to %s


%d bloggers like this: