This is soooo weird: Krebs on Security, Lorem Ipsum: Of Good & Evil, Google & China.
A Personal Blog
by Michael Froomkin
Laurie Silvers & Mitchell Rubenstein Distinguished Professor of Law
University of Miami School of Law
My Publications | e-mail
All opinions on this blog are those of the author(s) and not their employer(s) unelss otherwise specified.
Who Reads Discourse.net?
Readers describe themselves.
Please join in.Reader Map
Recent Comments
- Brooks Fudenberg on I Voted
- Jermaine Chad Ingram on Some Thoughts about the Downballot (Voters’ Guide Part II: Judicial Retention Elections)
- C.E. Petit on I Voted
- Jane Moscowitz on I Voted
- Ally Figueroa on Some Thoughts about the Downballot (Voters’ Guide Part II: Judicial Retention Elections)
Subscribe to Blog via Email
Join 52 other subscribers
I think the explanation is quite simple.
That text, in it’s full version, is used in (among many other places) Latex typesetting software, which is used by all kinds of people that might be writing things about Nato or the Internet. When used, you often will make mock documents, fill it out with that text to see how they look, etc. Then you add in your REAL text as needed. I have numerous examples of this on my computer that I’ve used over the years.
It is 100% conceivable, if not inevitable, that examples of the above-described before and after, have been scanned by Google. maybe they were posted on line. Maybe they were emailed. Maybe they were in an on-line how-to page for typesetting. Maybe maybe maybe.
Then Google’s autonomous, apparently self-teaching software linked a default text to some completed documents and “learned” how to translate that text – leading to the odd results you cite.
The specific details are unknown, but that’s almost surely what happened.
The bigger issue is that Google IS scanning this stuff. This isn’t proof of anything going on with Google Translate, it’s proof that nothing on-line slips by Google’s gaze – and that’s far scarier.