Skype translator is one of the new softwares innovations. Skype one of the most used video-call softwares, came out with a very innovative idea; the skype translator. Clay Dillow, the author, explains the usefulness of the skype translator and its’ biggest purpose which is “to make it possible for any human on Earth to communicte with any other human on Earth, with no linguistic divide”.  Dillow uses images and videos of new users enjoying this new features.

The image shows two users speaking in spanish and english through the chat windows. This features translates voice call between different languages in realtime, turning English to Spanish and Spanish back into English on the fly. Clay Dillow’s main argument is all the technocolgical arrangements that make the skype translator work, and how little of awareness and recognizitions users of these features have of the technological arrangements. In this article, he argues that the Skype translator was designed more like the human brain than a classical computer. With the development of Deep neural networks (DNNs), softwares are more capable of building more knowledge about the data sets that they were trained on, in this case, language.  This development reflects Lee Manovich’s media principles of automation. According to Lee, the high level of automation of media creation ” requires a copmuter to understand, to a certain degree, the meanings embedded in the objects being generated, that is, their semantics”. Similarly Dillow argues the automation of the Skype Translator is needed for the translation to be effective. Dillow uses the example of saying ” I like Ice cream”, he states that this is a much easier sentence for the software to understand and translate into another language. But what if the phrase was ” oh, that fumble was the last straw that broke the camel’s back”, much more complicated phrase that will need automation. Automation will look for probablistic matches, selecting the best solution from a number of options phrases based on what it thinks is most likely to be correct.

Another characteristic discussed in Lee’s new media principles is Transcoding, which is also found in Dillow’s article. Transcoding according to Lee, is “how the computerization of media turns into computer data”. Which reflects in all aspects, what DNNs were able to do. By Speech recognition DNNs were able to store certain tones, syntax, structure and meaning of languages, then DNNs went through a statistical matching process to be able to translate at it’s best accuracy. The statistical process shows how DNNs allowed the Skype Translator to perform transcoding. The idea that media, in this case languages’ tone and stucture was converted into computer data for the skype translator then be able to use a statistical procedure to pick the best educated words for translation. While doing research on Skype Translator I came accross opinions of users, this is what they said. The comment of Keval Padia, reminding of one our previous readings. His first paragraph acknowledges how skype has transformed media using speech recognition, automated translation and machine learning. this idea also reflects Couldry’s idea.Couldry states on page 10, “The internet has brought a shift in information production from a limited number of discrete forms”. It is evident by Skype Translator what Couldry’s arguing. The internet has connected more people, creating the production of information, that enable software like Skype Translator to be able to use that information to destroy language barriers.

Dillow  does a really good job stating all the difficulties that arise from using the Skype Translator, which makes his argument, conclusion and finding more valid and convincing to me.  His finding shows us how improved and developed machine learning has become, and also reflects some of the new media principles that Lee argues about. His conclusion make his argument more legitimate, he is aware that the Skype Translator is not perfect, that it still gets stuck on some problems. But he has hope that the technology will advance sufficiently for projects like this to work.

After reading this article, I started thinking what if our Television does the same. I come from a spanish speaking family, who sometimes has difficulty watching movies or shows, more importanlty news  on television were there aren’t any subtitles. What if television companies start using new principles such as the automation and the transcoding principle to able to translate and delivered a well translate message to their non-enlsigh speaking audience? Or what if this principles are used in class for students who hav e ahard time understanding english?

photo source