Real-time IM translation

The wired population may be able to zap messages around the planet in an instant, but whether those messages are understood is another matter. Language barriers still exist, and that’s where a new service called SpeakLike plans to make its mark. SpeakLike bills itself as the first instant messaging service for accurate, real-time translation chat across multiple languages, making it possible for users to type text in their own language and have others see it in theirs, accurately and within seconds. Whereas most current translation options rely either on machines, which are notoriously inaccurate, or on human translators, which are costly and time-intensive, SpeakLike uses a combination of both to improve accuracy and reduce costs. Users begin by downloading some free software from the service. Then, when they send some text, a human translator for SpeakLike checks and corrects the machine translation of that text in real time, allowing those on the receiving end to see it in their own languages, quickly and correctly. While not accredited translators, SpeakLike’s translation staff are bilingual and capable of conversation-quality translation, the company says; they are also bound by strict confidentiality agreements and a code of ethics. New York-based SpeakLike was launched into beta at the end of January, and is currently still for use by invitation only. Its services are available only in English, Spanish and Simplified Chinese at the moment. More languages are coming soon, however, as is an option for users to get transcripts of their conversations. Premium, enterprise and integrated web services—including support for legal and medical translation specialties—will be added later, SpeakLike says. Following its beta period, SpeakLike’s introductory pricing will start at USD 0.10 per short translated message. Cultural homogenization notwithstanding, the world is not that small when there are barriers to communication. By breaking those barriers down, SpeakLike—and the many minipreneurs that may follow in its footsteps—stand to win. Spotted by: Ozgur Alaz