Code Monkey like Fritos
Code Monkey like Tab and Mountain Dew
Code Monkey very simple man
“Code Monkey,” Jonatahan Coulton
Some very ubiquitous people in the MT Crowd are fond of saying that translation costs are going down. And many translators parrot them despite the absence of hard data to confirm this. Recently, the head of an agency formulated a version of this belief in a blog post: “[M]any of the translation services will continue to get cheaper and cheaper until eventually, MT will be able to do them for free.” And he proceeded to rattle off the usual McLocalization suspects: machine translation, crowdsourcing, Death Star agencies…
I am pretty sure somebody swallowed a caveat. And a rather large one at that. Nonetheless, if you analyze this belief about falling prices carefully, you will come to see that it is nonsense.
Let us suppose, purely for the sake of this thought experiment, that the entire annual output of the entire global translation industry is one billion words. All languages, all companies, all activities related to translation, from the lowliest hamster to the fattest “MLV.” Let us further suppose that the average cost of translating those one billion words was $200 million. That comes to an average per-word rate of $0.20.
Imagine now that I copied and pasted another one billion words into a single Word file and pushed that sucker through Google Translate. Imagine, furthermore, that even mighty Google’s servers don’t balk at chugging down that massive ton of garbage and, after, say, a few weeks, I get back an MT’ed final file. My cost per word would be zero per word.
(Of course, there would be a cost to Google for the Biblical amounts of resources I have wasted on my little antisocial effort, but that is irrelevant for this thought experiment. Let us further assume that Google can’t track me down even with those Google Maps cars that go around the world with little cameras on them.)
(I go AWOL with a new identity and end up eating Fritos while watching Sopranos re-runs in some fleabag motel.)
The fact remains that a massive drop would have occurred in the cost per unit translated. If you average out the cost of $0.20 for the original billion plus the zero pennies per word of my GoogleT experiment, the average, overall cost of translation would now be $0.10 per word.
But what does that have to do with anything?
What would my massive sophomore prank have to do with the cost of producing the original billion words?
Ipso facto, presto change-o, thanks to my machinations, translation costs would immediately plunge to $0.10 per word. I would have halved the cost of all translations with my little prank. In the entire world. In one fell swoop.
But what actual impact does my little act of cyber-piracy have on the first billion words?
It would still cost $0.20 per word to produce those original billion words.
Miguel Llorens is a freelance financial translator based in Madrid who works from Spanish into English. He is specialized in equity research, economics, accounting, and investment strategy. He has worked as a translator for Goldman Sachs, the US Government's Open Source Center and H.B.O. International, as well as many small-and-medium-sized brokerages and asset management companies operating in Spain. To contact him, visit his website and write to the address listed there. Feel free to join his LinkedIn network or to follow him on Twitter.