Low cost, accurate translation of text via the internet is an unsolved problem. Professional translation costs at least US$50 per page, and present computer translation produces poor output with loss of context and meaning. We present a solution based around minimum wage monolingual workers training an adaptive automatic translation algorithm to maintain context and meaning, and provide translations of native speaker quality. Although inappropriate for legal documents, a vast array of business, news and personal material can be targeted. A price-quality trade-off is provided, and the completely automated translation engine itself will quickly become the best in the market.

Shamus Husheer