The language industry is being very quick off the mark to slate AI as a result of a recent faux pas in China.
Tencent was showcasing the latest in transcription and simultaneous interpreting technology at the Boao Forum. Unsurprisingly the incident made The South China Morning Post, Hong Kongs’ Most popular English newspaper owned by Alibaba (Tencent’s main rival in China) More details about what happened can be found here in a report by Slator
For me, however, this opens up a much greater conversation. For the Language industry to be so vocal about something that is purported to be so useless, leaves me thinking that there’s a lot more fear in this industry than there was when MT first came out. So much so that they are now actively chastising such failures as much and as loud as they possibly can. It leads me to believe that the first real disruption to a once thought of “human only” industry might be just around the corner. Whilst the translation industry has come to terms with the fact that MT has its uses and can replace a human linguist in some cases, (If you doubt this statement then just take the time to read your last instruction manual for electrical or white goods coming out of the Far East) in the interpreting world this has never really been thought possible.
Step into the ring AI
Then along comes AI! In the last 5 years alone, AI has made so many significant advances that the thought of going to a conference and listening to a robot simultaneously interpreting the speaker’s keynote into 30+ languages is now a reality. In fact, I would wager that within the next 3-5 years AI interpreting will be used in a large portion of the day to day requirements and by the end of the next decade, it will be the mainstay at all major conferences and events.
It will NEVER replicate a human translation fully, but it will be accurate enough for the majority of the receiving parties to be able to understand all they need to know. And if we’re being truly honest about all of this, our ability to listen has been deteriorating at an alarming rate of knots in the last 3 decades anyway. Statistics show that we spend 60% of our communication time listening and retain just 25% of what we hear, watch AND listen to Julian Treasure’s Ted Talk from 2011 to find out 5 ways to help you listen more.
So taking all this into consideration and the fact that the Tencent huge AI engines got it a bit wrong the other week what’s the point of all this uproar? Unless, of course, your job is at risk and that listening out for every single mistake made by AI is so important in order to get a message out to the wider public about the potential dangers thus causing fear amongst the wider users of these services in a last minute attempt of self-preservation.