Ray Kurzweil is interviewed here by Nataly Kelly who works for Common Sense Advisory, a company focused on language in the context of globalisation, which of course includes translation. While her questions are informed by her work context (and sometimes lead to repetition in Kurzweil’s answers), they nevertheless make Kurzweil reflect on whether and how computers will replace humans as translators.
Kurzweil already predicted more than eleven years ago in his book “The Age Of Spiritual Machines” that computers will reach a high degree of almost human proficiency in translating written and spoken language by 2029. While he was formulating his prediction, Franz Och started to work on language translation algorithms; Och today is the man behind Google Translations, which still is very crude but nevertheless has reached mainstream. Considering how rapidly technological progress happens, I wouldn’t be surprised if 2029 will turn out to be a fairly conservative assumption.
But the video is not so much focused on time frames, as it is on questions based on the effects of computers on translation, like for example whether we will live in societies without language barriers because computers will comprehensively convey meanings from one language to another in an instant. Kurzweil answers that question with a categorical ‘no’. Even if computer translations will be able to reach a level of proficiency that will allow us to use them for day-to-day conversation and communication, Kurzweil predicts that people will still gain immense benefits from learning other languages, the same way literature (in particular poetry) will still need humans for translation. Words do not just have literal meanings but are also carriers of cultural messages, are steeped in historical contexts and are the products of unique individual creative expression. Unless we create artificial intelligence that at least fully matches our own human one, machines relying on databases for translation won’t reflect and produce the complexity and richness that humans are able to tap into when extracting meaning from words in another language.
Another question Kurzweil addresses in this context is whether computer translations will wipe out the profession of human translators, for example in non-literary fields like that of commerce. Again he answers with a ‘no’. While making references to the just mentioned translation quality argument, he also suggests that new technologies bring diversification to traditional professions by providing people with new skill sets (and Kurzweil refers in this context to the introduction of synthesisers in the 1980s and the consequent changes computers have made to music creation).
While Kurzweil’s historical perspective seems verifiable over longer time periods, it certainly won’t hold true in the short-term. After all: car factories are not filled with masses of multi-skilled workers but with robots and a hand-full of engineers and technicians, and those workers’ experience was just another repetition of what for example the weavers’ trade went through in the 19th century when the steam engine eliminated it. Therefore and despite different avenues being opened for a new generation of language agents, countless translators will lose their professional career prospects forever – the same way people are losing their jobs in bookshops and record stores without being re-employed en-masse by online companies. But I guess, if you’re a futurist with a fascination for technology, social justice aspects are outside your general frame of reference – which still makes Kurzweil’s reflection within its own albeit limited context quite interesting.
melange 