Reports indicate that Google plans to offer translation from any language to any language. “Historically, the approach to building machine translation systems is to have expert machine linguists write down dictionaries and rules on how to translate, say, from Chinese to English” said researcher Franz Och. “Trying to write down all the rules on how to translate from Chinese to English is very hard.” Google will use statistical MT technology to circumvent this need.

Why is Google interested in the translation business? Global business requires local content. Since Google’s mission is to “organize all the world’s information and make it universally accessible and useful,” getting that information into the right form makes a lot of business sense.

If Google’s engineers can figure out how to align different language versions of websites and use statistical-based machine translation to translate between any two languages, the company could accelerate international business by making gisted translations easily available from a trusted source.

Would such translations hurt the small translation agencies that provide much of the world’s language services? At first, projects such as Google’s will be a productivity booster for translation companies, but later, once companies trust simple machine translation for e-mail and internal documents, language service companies will have to turn their attention to more complex projects that require legal approvals, local adaptations, complex desktop publishing (DTP), and other labor-intensive and difficult-to-automate tasks.

The next frontier for Google and other machine translation suppliers: translating speech on the fly — “machine interpretation.” Until then, mainstream attention to any kind of translation can only help companies focus on the importance of localizing information to consumer needs.