Recently, Google released its Translation Toolkit enabling the translation of HTML and DOC files via an easy to use web portal that offers a WYSIWYG translation environment integrated with its Machine Translation engine.
If you don’t mind correcting machine translations, or retyping the English source text when you need to revert back to English, then the environment will be a plus for the average translator that does not currently use computer aided translation (CAT) tools.
Google smartly predicted that professional translators, which are a key crowd that Google wants to attract, will not be satisfied with just the machine translation option. So they included the ability to upload glossaries and translation memories. This helps produce more consistent and professional translations that machine translation cannot accomplish.
The WYSIWYG translation environment is helpful to visualize the source and get a better idea of the overall context; it is better than translating HTML at the text level and deal with all the tags and code. But the limits in language pairs and supported file formats (I was very surprised that they are not supporting XML and that only English source was allowed), and the fact that you need to work online, quickly offset the gained benefits for the professional translator, who is used to using similar CAT tools without these limitations.
So what will the impact be on the translation and localization industry?
In the short term, less experienced translators will be drawn to use the Google translation environment and will quickly learn the limits of machine translation and the benefits of translation memories and building accurate glossaries. This will help them mature and become more experienced translators. As more translators use the system to produce quality translations, it will help improve the Google Statistical Machine Translation accuracy, enabling Google to facilitate international markets reach, improving its bottom line.
Established CAT publishers will have to work harder to outdo Google or integrate with it. What Google built so far is not very hard to duplicate; Microsoft already has a similar environment to handle translations of its knowledge base. But long term, it will be very hard for others to keep up with the bilingual corpora that Google will amass. It is the corpus that the statistical machine translation industry is after to improve its Statistical Machine Translation engine.
Google still has to work on ensuring the protection of intellectual property to their rightful owners (the ones who pay for the translation in the first place). But as statistical machine translation quality improves, translation efficiency may significantly improve to possibly make this issue mute, or at least not as pressing. But this will not happen for many more years.
Some are cautioning that Google’s offering may not be free forever. That may be the case. But so what? We pay for other SaaS tools like CRM, online meetings, telecom, email… No one will mind paying a fee when the ROI is well justified!