Skype Translate Reportedly Has a Swearing Problem In Chinese

An anonymous reader writes: Skype Translate was supposed to be Microsoft's attempt at the "Star Trek" universal translator, offering real-time voice and text translation. It launched with one of the most challenging of languages, Chinese. And apparently, thanks to the Great Firewall, it has its problems. An American expat using it in China reports: "A glitch in the beta software misinterpreted the words I spoke. 'It's nice to talk to you' was translated as 'It's f*cking nice to f*ck you,' and other synthesized profanity, like the icebox robot in 1970's sci-fi flick Logan's Run, but with Tourette Syndrome. It was quite funny to me - I couldn't help but laugh during repeated takes, to Yan's exasperation - but the tech team were none too happy about it as they worked late into the night."

Comments

It's the *Star Trek* translator

[Actually, I do RTFA]

Of course it's profanity-ladden when it's trying to translate for people that it thinks are KHAAANNNNN!!

The problem is GAN1/4

[Anonymous Coward]

There is a Chinese character with too many meanings.
http://itre.cis.upenn.edu/~myl/languagelog/archives/005195.html

Re:The problem is GAN1/4

[kamapuaa]

"Gan" would never be used for "it's nice to talk to you." This hasn't been a problem with automatic Chinese translation for the last 10 years, and I only ever saw it with "dry" being translated to "fuck."

Also, we're talking conversation here, not characters. Sure the Chinese word for "fuck" has homophones, but that is just a normal part of the Chinese language. There's no more reason they'd be stuck over "fuck" than a million other homophones, like how "shi" can mean "ten" "is" "stone" "lion" or many other different things, all depending on both tone and context.

Re:The problem is GAN1/4

[hattable]

I imagine it was probably just an error message. Talking with someone from beijing and all skype got was 'er 'er'r'er'rr'er'er'er'er'er'er so the software just said fuck it!

Re:The problem is GAN1/4

[Applehu Akbar]

I wouldn't trust Skype AI to tackle anything like this since the time I wanted to change the credit card I had my World Subscription attached to The Payment Methods page on the site lets you add a card, but you can't delete an old card and you can't tell it which payment method to use. You just have to hope it picks the right one.

Re:The problem is GAN1/4

[Swave An deBwoner]

As noted in the link provided in the post to which you replied, this is not a question of homophones but rather a result of a simplified Chinese character serving triple duty. That discussion also points to another article by the author who states:

I am trying to make sense of how this phenomenon actually came about. It seems that the twenty or so different meanings of the three-stroke calendrical graph that is used to write GAN1/4 (a total of three distinct graphic forms in the traditional script -- , , -- all reduced to one -- -- in the simplified script) in Chinglish have all collapsed into the single meaning of "fuck". Wherever that graph occurs, Chinglish speakers will translate it as "fuck".

The Chinese characters in the above do not display properly on Slashdot but you can read the article at: http://itre.cis.upenn.edu/~myl/languagelog/archives/003205.html

In other words, the problem is with the written character, not the spoken word.

Re:The problem is GAN1/4

[Swave An deBwoner]

I noticed that you wrote Also, we're talking conversation here, not characters. and realized from your first sentence that you were familiar with the simplified character problem with the word "Gan". However despite your statement that "Gan" would never be used for "it's nice to talk to you." we can't know whether it was or was not used in the reported case because, understandably enough, the very short "Network World" article doesn't provide the actual words that were used. Maybe the speaker did use "Gan".

And although "we're talking conversation here, not characters", my first thought was that the translation software did something like:

1. Capture sound of word
2. Determine what character the word corresponds to
3. Translate that character

In which case the "collapsed" simplified character problem would indeed be pertinent.

Obligatory

[the_Bionic_lemming]

ROLLER CAPTION: IN 1970, THE BRITISH EMPIRE LAY IN RUINS, FOREIGN NATIONALS FREQUENTED THE STREETS - MANY OF THEM HUNGARIANS (NOT THE STREETS - THE FOREIGN NATIONALS). ANYWAY, MANY OF THESE HUNGARIANS WENT INTO TOBACCONIST'S SHOPS TO BUY CIGARETTES...
Enter Hungarian gentleman with phrase book. He is looking for the right phrase.
Hungarian I will not buy this record, it is scratched.
Tobacconist Sorry?
Hungarian I will not buy this record, it is scratched.
Tobacconist No, no, no.This ... tobacconist's.
Hungarian Ah! I will not buy this tobacconist's, it is scratched.
Tobacconist No, no, no ...tobacco...er, cigarettes?
Hungarian Yes, cigarettes. My hovercraft is full of eels.
Tobacconist What?
Hungarian (miming matches) My hovercraft is full of eels.
Tobacconist Matches, matches? (showing some)
Hungarian Yah, yah. (he takes cigarettes and matches and pulls out loose change; he consults his book) Er, do you want ... do you want to come back to my place, bouncy bouncy?
Tobacconist I don't think you're using that right.
Hungarian You great pouf.
Tobacconist That'll be six and six, please.
Hungarian If I said you had a beautiful body, would you hold it against me? I am no longer infected.
Tobacconist (miming that he wants to see the book; he takes the book) It costs six and six ...(mumbling as he searches) Costs six and six ... Here we are ... Yandelvayasna grldenwi stravenka.
        Hungarian hits him between the eyes. Policeman walking along the street suddenly stops and puts his hand to his ear. He starts running down the street, round corner and down another street, round yet another corner and down another street into the shop
Policeman What's going on here then?
Hungarian (opening book and pointing at tobacconist) You have beautiful thighs.
Policeman What?
Tobacconist He hit me.
Hungarian Drop your panties, Sir William, I cannot wait till lunchtime.
Policeman Right! (grabs him and drags him out)
Hungarian My nipples explode with delight.

Re:Obligatory

[Anonymous Coward]

https://www.youtube.com/watch?v=G6D1YI-41ao

Re:Obligatory

[jez9999]

Heh. Literally the first words out of my mouth were "Is that something from Monty Python? because it's not at all funny."

Re:Obligatory

[OhSoLaMeow]

You can't expect me to spend 3 minutes of my life reading a convoluted mess of a dialog, only to have no punch line.

A joke doesn't need to have a punch line. Witness:

A teenage boy is getting ready to take his girlfriend to the prom. First he goes to rent a tux, but there’s a long tux line at the shop and it takes forever.

Next, he has to get some flowers, so he heads over to the florist and there’s a huge flower line there. He waits forever but eventually gets the flowers.

Then he heads out to rent a limo. Unfortunately, there’s a large limo line at the rental office, but he’s patient and gets the job done.

Finally, the day of the prom comes. The two are dancing happily and his girlfriend is having a great time. When the song is over, she asks him to get her some punch, so he heads over to the punch table and there’s no punchline.


See?

Re:Obligatory

[KGIII]

That was wrong and you should feel bad.

Re:Obligatory

[tehcyder]

The joke was long and not funny. You, sir, need to work on your comedic timing and understanding of humor. You can't expect me to spend 3 minutes of my life reading a convoluted mess of a dialog, only to have no punch line. And who the **** upvoted this.

your report here says that you are an extremely dull person. You see, our experts describe you as an appallingly dull fellow, unimaginative, timid, lacking in initiative, spineless, easily dominated, no sense of humour, tedious company and irrepressibly drab and awful

So, what caused the problem

[hackwrench]

This is a poor article. What does The Great Firewall have to do with this particular problem.

Re:So, what caused the problem

[jandersen]

What does The Great Firewall have to do with this particular problem.

It is about China - of course the Great Firewall has to be mentioned, otherwise, what's the point of anything? Otherwise we would have to mention things like democracy, freedom of speech or Tibet, and that would be even less relevant. (Note: this was an attempt at sarcasm)

That's not a 'glitch' ...

[gstoddart]

A glitch in the beta software misinterpreted the words I spoke. 'It's nice to talk to you' was translated as 'It's f*cking nice to f*ck you,' and other synthesized profanity

I refuse to believe someone didn't do that on purpose.

That's too damned funny to be by accident.

Please fondle my bum

Re:That's not a 'glitch' ...

[hcs_$reboot]

Exactly the comment I intended to post. This is how marketing campaigns work nowadays.

Re:That's not a 'glitch' ...

[wonkey_monkey]

Hey, it's a serious medical condition.

Its a neat bit of tech

[Karmashock]

... fix the bugs. The whole thing is awesome. Render the language barriers meaningless.

At some point people are going to put a little hearing aid into their ears and auto translate anything.

Re:Its a neat bit of tech

[Nyder]

... fix the bugs. The whole thing is awesome. Render the language barriers meaningless.

At some point people are going to put a little hearing aid into their ears and auto translate anything.

I'd rather have a fish in my ear.

Re:Its a neat bit of tech

[Karmashock]

A fish we can do... a babelfish... not so much.

Re:Its a neat bit of tech

[Kjella]

At some point people are going to put a little hearing aid into their ears and auto translate anything.

Long before that people will have figured out that speaking one of the big global/regional languages is useful. I wouldn't go so far as that we'll all join up on one language, but say one of the top 6 - Mandarin, English, Spanish, Hindi, Russian, Arabic. And I'd likely strike Hindi from that list, since like Portuguese, Bengali, French, Malay, German and Japanese (7-12) it has no significant tendency to spread beyond its current native regions. That is, if they ever get around to learning a second language since many people still aren't literate in their first.

A good example is Europe, the trend is extremely strong that secondary speakers of languages like German and French are down while English is up. India also seems to align on English, not Hindi. Africa is already using English, French and Portuguese as common languages, except much shorter in development. Same with SE Asia, what's a group from Myanmar (Burmese), Thailand (Thai), Malaysia (Malay) and Vietnam (Vietnamese) going to speak in common? Chinese or English. Maybe it's not as clear in the US where 300+ million already speak your language natively, but ours has ~5 million. Yeah, we're going to learn a bigger language.

Re:Its a neat bit of tech

[dunkelfalke]

*facepalm* the stupid is strong in this one.

Hindi and Mandarin aren't even in the same language family. They are as different as two languages can ever be.

There are more than enough Russian speakers in the west. I, for one, speak Russian as a second language. Same goes for Spanish and Arabic, by the way, so you are definitely talking out of your arse. Your provincialism is really showing, except that even in your own country 13% of the population speaks Spanish.

And this "exposing them to modern western culture" is laughable.

Re:Its a neat bit of tech

[Karmashock]

My comment was in reference to this statement:
"I'd likely strike Hindi from that list, since like Portuguese, Bengali, French, Malay, German and Japanese (7-12) it has no significant tendency to spread beyond its current native regions."

To which I said:
"Mandarin is not much different from hindi really."

The point being that I was saying Mandarin is unlikely to spread either much less become a global language. We were talking about language roots. We were talking about the propensity for languages to spread and become dominant.

The mistake you always make is in thinking I am stupid. I am not even remotely stupid. Learn. If you keep making this mistake at some point it will become obvious that YOU are the stupid one.

Re:Its a neat bit of tech

[dunkelfalke]

Oh, but I know you are.

And saying that any language is unlikely to spread beyond its native region is stupid as well. No lingua franca stays forever. Chinese used to be the trade language in the whole Southeast Asian region and it might very well become that again in the next 100 years - even now people in Europe are encouraged to learn Mandarin because China is where the money is. Russian is still a trade language in countries bordering Russia and German used to be the language of science not very long ago, so quite a lot of scientific terms in English are of German origin (say hello to Bremsstrahlung), French is the international postal language. A lingua franca can easily change within a human lifespan.

Re:Its a neat bit of tech

[tomhath]

Language is hard. Even if you aren't translating between two of them.

Re:Its a neat bit of tech

[Anonymous Coward]

You're a sore loser.

Re:Its a neat bit of tech

[kenwd0elq]

When I lived in the Philippines, most Filipinos around the major military bases all spoke English to each other, because they were from many different Philippine provinces and islands, most of which each have their own (mutually incomprehensible) dialects of Tagalog.

Re:Its a neat bit of tech

[Karmashock]

Login and we'll see, troll. Stay in lurk mode and I'm going to waste my time with you.

Re:Its a neat bit of tech

[Karmashock]

You realize that by making trollish AC comments you validate my position. And you're doing little more than making laugh at you because you're so obviously butthurt about how I won't engage with you anymore.

What is funnier one top of that is that you're also admitting that despite not talking to you for ages you're still e-stalking me.

Here's a question... are you a just an AC troll? Or are you a sock puppet of an existing account that is AC trolling?

Just curious. :D

You people are so stupid... why are you so bad at this?

Re:Its a neat bit of tech

[tehcyder]

The point being that I was saying Mandarin is unlikely to spread either much less become a global language.

I saw a TV news item last night about kids in Africa being taught Mandarin. I wouldn't rule out Mandarin becoming a global language if I were you.

Re:Its a neat bit of tech

[Karmashock]

We will have to agree to disagree there.

China is ultimately insular. They're not a people given to evangelism... that is the spreading of their beliefs or culture.

A great many other cultures are evangelistic. The US is the most powerful of these and is strongly evangelistic. Exactly why in the foreseeable future would Mandarin overtake English? Its not credible.

Traditional Chinese Computer Translation Joke

[kenwd0elq]

The "traditional" joke concerning computer translation is about 30 years old - at least, because I've been telling it that long, and I heard it from somebody else.But it's still a classic.

The original translating computer wasn't voice-recognition; you had to type in your statement in English, and it would be translated to Chinese on the screen. So in order to demonstrate how good it was with colloquial English, the programmer typed in a common saying, "Out of sight, out of mind". The computer whirred and chirped for a couple of minutes, and a column of Chinese characters appeared. The Chinese operator looked quite puzzled, but to play along, he typed (in Chinese characters) exactly what he had read on his screen.

Chirp, whirr, beep, and the machine produced the translation back into English.

It said "Invisible Insanity".

Re:Traditional Chinese Computer Translation Joke

[K. S. Kyosuke]

"The vodka is good, but the meat is rotten."

Re:Traditional Chinese Computer Translation Joke

[kenwd0elq]

All _good_ humor has some basis in reality. It's the unexpected disconnections that make humor funny.

How dare they!

[Etherwalk]

Until it can seamlessly change the words I'm saying, as I'm saying them, into the receivers language without so much as a configuration nor without talking over the top of me, it is not the Star Trek Universal Translator.

Yeah. How dare a tech company be aspirational.

"Don't catch any bugs!" --Klingon border sentry to Enterprise

It'll be fixed soon

[ITRambo]

Skype would become the world leader in calls if and when these bugs are worked out by them before another service leapfrogs them. This will be a wonderful technology to commonly use someday.

Re:It'll be fixed soon

[Wycliffe]

Skype would become the world leader in calls if and when these bugs are worked out by them before another service leapfrogs them. This will be a wonderful technology to commonly use someday.

I agree that this has the potential to be awesome and skype is great for remote communication but they need a way to use it in person locally.
If I'm having lunch with someone, we don't want to have to have 2 ipads and communicate over skype when we're sitting right next to each other.

Translate the mind

[lorinc]

'It's nice to talk to you' was translated as 'It's f*cking nice to f*ck you,'

Seems the damn thing is actually translating what's in your mind instead of what your saying...

That explains so much

[antifoidulus]

Here I thought that I accidentally called a really polite phone sex operator.

Re:Who is listening?

[Anonymous Coward]

Who is listening? Mostly you, all of us are being recorded by a foreign government.

From the perspective of an American: 'I object the actions of Hacking Team because they make business with oppressive regimes!'

You hypocrites.

Swearing? That's nothing.

[roesti]

Someone on Skype just said to me, "I seem to be having tremendous difficulty with my lifestyle". Little punk. If I ever find him...

Re:War

[Mal-2]

It is of course well known that careless talk costs lives, but the full scale of the problem is not always appreciated.

For instance, at the very moment that an Earthman, Arthur Dent said "I seem to be having tremendous difficulty with my lifestyle," a freak wormhole opened up in the fabric of the space-time continuum and carried his words far far back in time across almost infinite reaches of space to a distant Galaxy where strange and warlike beings were poised on the brink of frightful interstellar battle.

The two opposing leaders were meeting for the last time.

A dreadful silence fell across the conference table as the commander of the Vl'hurgs, resplendent in his black jewelled battle shorts, gazed levelly at the G'Gugvuntt leader squatting opposite him in a cloud of green sweet-smelling steam, and, with a million sleek and horribly beweaponed star cruisers poised to unleash electric death at his single word of command, challenged the vile creature to take back what it had said about his mother.

The creature stirred in his sickly broiling vapor, and at that very moment the words I seem to be having tremendous difficulty with my lifestyle drifted across the conference table.

Unfortunately, in the Vl'hurg tongue this was the most dreadful insult imaginable, and there was nothing for it but to wage terrible war for centuries.

Eventually of course, after their Galaxy had been decimated over a few thousand years, it was realized that the whole thing had been a ghastly mistake, and so the two opposing battle fleets settled their few remaining differences in order to launch a joint attack on our own Galaxy- now positively identified as the source of the offending remark.

For thousands more years the mighty ships tore across the empty wastes of space and finally dived screaming on to the first planet they came across- which happened to be the Earth- where due to a terrible miscalculation of scale the entire battle fleet was accidentally swallowed by a small dog.

Those who study the complex interplay of cause and effect in the history of the Universe say that this sort of thing is going on all the time, but that we are powerless to prevent it.

Challenging indeed

[jandersen]

It launched with one of the most challenging of languages,...

Nothing challenging about Chinese - it is pronounced like it's spelled, as the old joke goes. Seriously, though, Chinese is relatively easy to learn, even beyond the elementary stage. There are no grammatical inflections in the way we have in Indo-European languages, for one thing, the grammatical rules are simple and regular (unlike in English), and transcriptions like pinyin represent the sound of the spoken language well, unlike in English: there are many words in the English vocabulary that are pronounced differently from what you'd expect from the spelling, whereas there are virtually none in Chinese.

That's not a f*ckin' bug...

[s0litaire]

...it's an f*ckin' undocumented feature!
"Skype translate : Gordon Ramsay Edition"

Re:Plan9 OS pwns it

[juanfgs]

plus it has a cuter mascot.

Icebox Robot?

[FilmedInNoir]

https://www.youtube.com/watch?...

Holy Mother of God and all her wacky nephews

[Erbo]

Obviously, it's not the Star Trek translator, it's the Firefly translator. :-)

Re:WOOPS

[Fortran IV]

Takes me back to the 1970s. "What's a four-letter word for intercourse?" "Talk." Bet your sweet ass.

Re:Stop calling it the ST Translator

[tehcyder]

Until it can seamlessly change the words I'm saying, as I'm saying them, into the receivers language without so much as a configuration nor without talking over the top of me, it is not the Star Trek Universal Translator.

Plus it has to provide free cakes and beer. And address me as "big boy".