Google have announced that they are currently developing a language application which will allow the instant translation of speech. Whilst such a thing has been attempted before, head of translation services at Google, Franz Och expressed that the technology needed should provide a reasonably well working device within a few years.
However, whilst this may well be true, the long history in such attempts is notable for one thing, its failures. In 1954 IBM said that they had successfully translated words of Russian to English using an artificial, electronic brain. Georgetown scholar Leon Dostert revealed that though they could not yet “insert a Russian book at one end and come out with an English book at the other”, the process would only take about three years.
56 years later and Google have revealed that the translator phone will provide instant language changes and will require individuals to use voice recognition on devices. Experts in the field have hit upon this latest revelation as the major flaw in the possible plan, with voice recognition currently being far from the smooth, instant transaction that it should be. Whilst activated systems currently allow for simple vocabulary and commands to be used, resulting errors often cause huge aggravation. In addition, the discovery that a leading voice translation service in the UK which used a computer to convert voice to text was largely relying on call centre staff, has not done the industry any favours.
However, whilst there may be many sceptics, Google is not a firm to make rash decisions and untested announcements, so only time will tell if the universal translator is only around the corner.
