As a first entry in this blog, I would like to draw attention to the prediction of Ray Kurzweil (The Singularity Is Near, 2005) that machine intelligence will overtake human intelligence around mid-century. This point has been named the “singularity”. “The technological singularity is the hypothetical future emergence of greater-than-human superintelligence through technological means. Since the capabilities of such intelligence would be difficult for an unaided human mind to comprehend, the occurrence of a technological singularity is seen as an intellectual event horizon, beyond which events cannot be predicted or understood.” – Wikipedia.
The documentary “Transcendent Man” is a thought-provoking presentation of Kurzweil’s ideas. However, a related and more entertaining video caught my attention, and serves to raise the question of man’s relation with machine intelligence: Singularity: How Scared Should We Be?
If Kurzweil’s peredictions are true, the kind of technical translation I do should eventually be performed as well if not better by computers, even if I don’t live to see it. Literary translation is another question as far as I am concerned. Manipulating information is one thing, plumbing the human soul is another.
Post Scriptum: An Interview with Kurzweil on Translation
After writing the above I cam across an interview with Ray Kurzweil precisely on the question of the future of machine translation technology. The interview was referenced in the book Found in Translation: How Language Shapes Our Lives and Transforms the World. In fact, it is one of the authors of the book, Nataly Kelly, who conducts the interview available here:
Kurzweil is more specific here, predicting that machines will achieve human levels of translation by 2029. On the other hand he cautions that there is no substitute for reading literature in the original language. As he explains, even with human translations, something of the specific culture of another language is lost in translation. And furthermore he doesn’t see human translators being out of work, anymore than musicians are out of work because of the power of present-day computers to generate music. Human translators will just be able to do more while working differently.
In this interview, Kurzweil characterizes translation as “the most high-level type of work one can imagine …. The epitome of human intelligence is our ability to command language. That is why Alan Turing based the Turing test, which is a test of whether or not a computer is operating at human levels, on a command of language.”
And it is precisely because of the complexity of translation that humans must harness the power of machines to improve it. “These tools are going to increase our ability to use, create, understand, manipulate and translate language,” Kurzweil explains. “The idea is not to resist the tools, but to use them to do more.”
I agree with Kurzweil’s fear (fear, not prediction of a doomsday though) that AI constitutes a serious danger. I add that we translators are unwilling contributors that accelerate the taekover by machines. When we translators proofread MT output, we actually teach the machine to translate better. Our feedback is ploughed back into the machine, either right then, or when our clients collect our translation memory.
That feedback loop, as all loops, will at one point exponentially sharpen machines’ ability to squeeze us out of the equation. If/when machines develop a sense of ego (us/them) and enough autonomy, we’re fried.
But then again, there’s no stopping progress. Let’s keep our fingers crossed.
Unfortunately the progress of technology has until now far outstripped humanity’s progress in learning to live peaceably with each other. Technology is essentially neutral. It can be put to good or to bad use. The potential for a race of super machines dominating humanity is real. Also is the potential for these machines to solve problems of hunger, sickness, and even good governance. I agree with you, we need to keep our fingers crossed, on both hands!