AI Threatening the Careers of Interpreters

Which industry will be wiped out by the robots? From the current situation, at least it will not be the interpretor industry.

Bell Wang, a co-translator, accused iflytek of using human translators to disguise itself as AI co-interpreters. In response to the human-machine coupling, which is the focus of the controversy, guoku held an exclusive dialogue with li shipeng, co-director of iflytek AI research institute, and listened to the voices of several translation insiders including li changshuan, vice President of the high translation institute of beiwai university, and translators zhang wei and Chen (pseudonym).

Was it a misunderstanding? On September 20, Bell Wang came to Shanghai international convention center to do a live transmission for the high-end equipment technology and industry branch of the 2018 international conference on innovation and emerging industry development (IEID). During the meeting, he found that iflytek recognized the translation spoken by the interpreter and then put the translation on the screen and live broadcast. Translation is not done entirely by AI. What's more, iflytek used their translation under false pretenses without informing Bell and his partner in advance, suspected of infringing their intellectual property rights.

At 8:00 p.m., sunfly held a press conference via teleconference. Senior vice President jiang tao and product manager liu chenxuan, who was in charge of translation business, attended the press conference.

At the press conference, iflytek denied the accusation, saying it was a "misunderstanding" by its interpreter: Bell did not know where he was in a sub-conference, and ai worked in a human-machine coupling mode. According to the introduction, IEID USES IEID intelligent conference system of hkust, which has two working modes. One is automatic translation, which recognizes speakers' speech, converts them into text and translates them in real time, and then projects them on the big screen. The second is human-machine coupling. As the name implies, the machine recognizes speech and then transcribes it into text for translation, reducing the difficulty of human co-interpreters.