Whatever humans can do…

…computers can do it better!


産業界では2年後に第1弾を迎える消費税増税に戦々恐々としている。 =That we fear in the industry and we welcome war to the consumption tax hike in two years after the first bullet.

Eh…maybe not.

For this DS106 assignment, “Google Translate Fail”, one must “find something in a foreign language and use Google Translate and laugh about how awful it is.” Oh, and while you’re at it, “if its a language you know, give an explanation of what it should be and list possible reasons it got messed up.” For this assignment, I chose to translate something from Japanese to English. The results, as you have seen, are quite..um, remarkable (kind of, at least).

For my line, I took it from a news page after typing in 戦 々恐々in Google search.

First, in regards to why the explanation is “messed up”, is because Google translate does not have the ability to understand the complex nuances of the language. Taken from a video (produced by Google presumably) found on the blog entitled Google Operating System, “these computers use a process called ‘statistical machine translation’ — which is just a fancy way to say that our computers generate translations based on patterns found in large amounts of text.” That means that there is really is no true “understanding” of a language to a computer. Translating comes down to, in simple terms, numbers or statistical correlations. So, in terms of quality, the translations are usually more wrong than correct. In terms of speed, however, it is certainly much faster.

So while Google translate isn’t the greatest in terms of quality, the speed it is able to translate at is incredible and will no doubt have us coming back to it again and again regardless of the lack of quality.