As time goes by, people are amassing larger and more diverse collections of digital belongings.
The emergence of Cloud computing has enabled people to move their personal files to online places, creating new digital commodities through online services. However, little is known about how this shift might shape people’s orientations toward their digital possessions in a not so distant future.
Consumer behavior patterns have changed drastically over the years, modifying the way they relate to the products or services they acquire daily, moving gradually from physical ones, to locally kept digital materials, now stored in the online world.
Suddenly a new layer of reality emerges, sprouting into existence. And we wonder once again, questioning the nature of our existence and how we perceive it.
Are we coming out of Plato’s metaphorical cave at last? Or are we dwelling into an even deeper, darker one?
“Topos ouranios”: The ideal world of pure ideas. A place unreachable, ethereal, where everything exists in its own true form as opposed to our material daily sphere of mundane objects and banal endeavors.
The arrival of Cloud computing and its heavenly immaterial associations only helps me fuel my bizarre associations even further. Through the recent introduction of neural based network technology applied to many areas, of which translations are but a minor portion of it, this “ethereal space” seems to be evolving into some sort of living entity, progressively capable of learning and understanding by its own means.
But how does this relate to the translation industry, you ask?
How does Cloud Computing integrate with Neural MT and how does this affect the future of translation?
Well, Google has previously taught its artificial intelligence to play games, and it is now even capable of creating its own encryption. Now, its language translation tool has used machine learning to create a ‘language’ of its own, which enables what they now call “Zero-shot translations”: The possibility for a machine translation engine to be able to translate into an unknown language pair by automatically comparing and virtually “learning” from the interaction of other associated language pairs that contain at least one of the original languages as part of the process.
How does it work? Basically, the machine learning system analyses and makes sense of languages by looking at entire sentences – rather than individual phrases or words.
Following several months of testing, the researchers behind the AI have seen it become able to translate languages blindly, even if it’s never studied one of the languages involved in the translation before.
The team in charge of this research project said the system had become able to make “reasonable” translations of languages it had not been taught to translate in the first place.
According to their reports, the AI was taught Portuguese→English and English→Spanish and it was then able to make translations between Portuguese→Spanish through the Zero Shot Process.
The developers of this technology now speculate that this system’s translation capabilities will be greatly empowered once they upload it and let it “feed” itself from the cloud.
A little bit too much anthropomorphization for my delicate taste…
So, if machines seemed to be behind their human competitors in matters of precision, this new aspect of its capabilities might shorten the distance quite considerably in favor of machine translation.
Of course, at this stage, it’s all nothing more than wishful thinking or perverse speculation (depending on where you stand on the subject). If you ask me, I see different, ominous shapes taking form in those clouds.
But that’s the magical thing about clouds, isn’t it? We all see different things in them.
The truth lies in their pure potential.