Could AI translators change the language of business? by Sophie Charara
With the help of parallel text datasets such as Wikipedia, European Parliament proceedings and telephone transcripts from South Asia, machine-learning has now reached the point where translation tools rival their human counterparts.
According to Ofer Shoshan, chief executive of One Hour Translation, neural machine technology (NMT) translators will be capable of handling 50 per cent of the global translation market in one to three years.
NMT translators might not be mainstream yet, but the biggest companies in the world, including Google, Microsoft and Amazon, are engaged in a fascinating race to close the communication gap between their software and skilled translation workers and agencies.
AI translation has flourished most in business
Business has been a key use-case from the start as technology giants position their Skype tools, translation apps and smart speaker skills as essential to any companies looking to expand in foreign markets.
Wearable tech startups too are still able to make an impact with Waverly Labs’ Pilot, Timekettle’s WT2 and Mymanu’s Clik+ all vying for the space left open by last year’s Google Pixel Buds, which were criticised as being badly designed and too reliant on connected Android smartphones. Actions like interacting with Alexa or popping in a smart earbud can make much more sense in a client meeting than for an everyday tourist.