Microsoft and Skype may have just invented the world’s first real universal translator


Universal translator

Universal translators are a common trope in science fiction. In The Hitchhiker’s Guide To The Galaxy, they come in the form of the babelfish, a tiny crustacean you jam in your ear. In Farscape, they are bacteria injected into your body. In Star Trek, they take a less squishy form as a wand or tiny computer pinned to your lapel.

In all incarnations, though, a universal communicator is seen as alien and futuristic. But Microsoft wants to change all that. The Washington-based company has just revealed a new real-time speech translation tool that is set to be built right into Skype, and which can translate any foreign language into English in the blink of an eye.

Currently in the early stages of development, the new tool is called Skype Translator. Microsoft’s Skype VP Gurdeep Singh Pal and communications manager Diana Heinrichs demoed the feature at the Code Conference. Using the tool’s live speech translations capabilities, Pall was able to have a conversation in English while Pal spoke German.

Here’s how Microsoft is explaining the new tool:

Skype Translator results from decades of work by the industry, years of work by our researchers, and now is being developed jointly by the Skype and Microsoft Translator teams. The demo showed near real-time audio translation from English to German and vice versa, combining Skype voice and IM technologies with Microsoft Translator, and neural network-based speech recognition. Skype Translator is a great example of why Microsoft invests in basic research. We’ve invested in speech recognition, automatic translation and machine learning technologies for more than a decade, and now they’re emerging as important components in this more personal computing era.

Although Skype Translator only supports a few languages right now, it is well on its way towards release, with Microsoft promising a beta app released for Windows 8 by the end of the year. But imagine when this technology comes to Skype for iPhone, or better yet, got built into Siri! As a traveler currently typing these words overlooking a picturesque valley full of hot-air balloons sailing over Cappadocia, the possibilities of just pulling out your phone and talking with anyone make me giddy. The future of technology is the future of communication, and the future of communication is to be able to talk to anyone, no matter where they’re from.

Source: Microsoft

  • Jan Zegers

    All due respect, but the German translations are very ungrammatical. It makes me presume that the English ones aren’t that spontaneously translated, rather pre-modified for this demo purpose.

  • Andy Shorrock

    Hmmm, nice but if all of the emotional meaning is lost it’ll be of little value, it’s not the words you say, but how you say them that count.

    • Kevin Sereni

      Is this the first time you’ve seen text communication? Ever read a book? Or comment on an article, maybe?

      • justin

        du hast, and du hast mean 2 entirely diffrent things buddy :p

      • Tha Jonster

        But this is where the patents are made, recognition of specific speech patterns and content references can help technology to recognize the difference between “here and hear or even they’re vs their or there. Computers are starting to gather a lot of knowledge very quickly. I see google android doing this as I text. Making guesses that are contextual based.

      • hod0r

        You probably mean “du hasst”, because “du hast” always means the same.

      • Andy Shorrock

        Amazingly enough no, it’s not the first time I’ve ever seen text based communication – you and I are doing that now… Now are you being funny (humorous) or clever (smart) or just plain old “I know it all & you know nothing” because all 3 are given away by tone of voice & inflection that mere words alone can’t convey.
        I don’t know about you Kevin, but I’ve never done a “creative writing” class so I’ve never been trained in turning emotions into plain text so others can get what I mean, rather than assuming from what I say.

      • Kevin Sereni

        Maybe a little of each, Andy. All I’m really saying is…This problem was not invented by Skype. Any type of text communication requires a little reading between the lines, or context.

      • Andy Shorrock

        True enough.
        Sounds roughly like some kind of AI will be needed to catch all of the (pardon the pun) nuances of communication in ‘plain text’ if it’s to successfully heuristically ‘learn’ the difference between say ‘la chat’ & ‘la chat’ & exactly what it means in a certain circumstance.
        When I started looking at computers 40 odd years ago, “4th generation computers would be artificially intelligent” – well IMHO we’ve a long way to go on natural intelligence yet, so don’t hold your breath.
        You have to know a thing, to be able to successfully describe a thing, to then turn it into algorithms that fast calculating IT can apply. That’s where the intelligence lays…
        IMHO of course ;)

    • Karolis Makrickas

      It’s harder to translate from english to german, and it’s easier from german to english just because in english you can change word places a little bit, for example: “i had brekfast” and “brekfast had I”. The last one isn’t right writen, but you can undestand. But in german, if you missplace the word you will get “brekfast had me”. And to translate other languages is even harder. For example Lithuanian language have more declension than human has fingers :) and even more exseptions…

  • dawson

    Why would Microsoft put it in Apple iPhone when Microsoft have Windows Phone and all of Nokia market …???

    • Microsoft thinks software, not hardware. That’s the why.

  • Adrian Kóska

    That automatic German makes me cringe, and I only learned German at high school and don’t remember much. Germans don’t speak so slow and clearly. I’m sure it would produce utter garbage if the lady was to speak in a normal fashion.

  • RandomScribbler

    Um, sorry, a Babelfish is not a crustacean. It’s

  • lastguru

    Why does she open her eyes so unnaturally? I thought they would pop out of sockets..

    • Karolis Makrickas

      ’cause she stage 1 zombie :)

  • JSintheStates

    Well, I have to admit, as a first pass, that ain’t too shabby! Universal translator? I don’t think so! And as for all the IT jargon, and double talk—this is not magic! This is a lot of software code! As for the German to English demo, English is a root Germanic language!

    Of course, you can’t buy this! You can license it through your local IP provider, after you sign the non-disclosure agreement (everything is their’s, including the content of your conversation) and purchase the premium upgrade from your Cable TV Megacorp.

  • Richard Patient

    It’s early tech and obviously has a long long way to go. (as anyone that has ever used Google translation or any other machine translation can tell you) but it still exciting to see such tech being invented and developed. Science Fiction is becoming fact.

  • U Mad Bro?

    Ana amazing start! I approve of anything that brings us closer to Star Trek Tech.