Skype Shows off its German…and…”Meh”
And yet, after 15 years (and countless dollars), the German audience members who heard the translation commented that “it wasn’t very good.”
As Nataly Kelly writes in The Huffington Post, “technologies don’t typically replace whole fields.” Rather, and especially in professions which can be described as both art and science, such as translation and interpretation, technology enhances the profession, expanding the capabilities of the professionals and generally raising the bar.
Technology also brings awareness. Consider your own experience with Google Translate, and the parodies you may have come across where people have thrown text through the translator a few times to very comedic ends.
What we begin to notice is how much we need professional translations. Sure, if you’re stuck in a country where you don’t speak the language, Google Translate can help you order dinner, or find a doctor.
But it isn’t going to replace the United Nations any time soon. And I don’t think any Human Resource managers want to see Google Translate touching their employee handbook or helping to explain the pension fund to prospective employees, either.
Futuristic Natural Language Processing Projects
Still, the profession is evolving, and the best translators and translation services are going to evolve with technology. Some of the most inspiring natural language processing projects of late offer hysterically funny but also surprisingly brilliant language. Here are a few of our favorites:
- Cleverbot – This AI app is hovering at around 60% human, and has been around for a few years now. Try having a conversation with it. It’ll get you thinking about how very difficult human language really is.
- Eugene Goostman – From Princeton Artificial Intelligence, this “chatterbot” earned the 2012 Turing 100. Try out a conversation with Eugene and you might begin questioning what rational language is.
- Grammar-Link – This project will parse sentences for you…which may not be that exciting, but can make you start thinking about how intricate our languages are. Try it out, and see if you notice how complicated language can be.
- ELIZA – If you’re up for some machine-based therapy, ELIZA was originally written by Joseph Weizenbaum at MIT between 1964 and 1966. ELIZA is a parody of a Rogerian therapist, and is considered parody. As one of the first “chatterbots” you can see at once how far we’ve come, and how far we need to go to reach anything even close to an intelligent human “chat”.
- SYSTRAN – Ending with the oldest machine translation service, SYSTRAN was behind Google Translate until 2007, and still helps to power Yahoo and Bing
The profession is evolving and we’re looking forward to what’s in the future. One thing the Microsoft/Skype team noticed is that, when their machine learned two languages, the first language improved. And when the machine learned three languages, again, the first languages improved as well. No one knows why. All we know is that learning languages and being interested in language is one of the most human things we do.
Take some time to learn a bit more about Language Tran and how we integrate Computer Aided Translation into our work, to provide the best translation possible for you, whatever your needs are.