Translating Charlese
The Defense Advanced Research Projects Agency (DARPA) is putting its not insignificant resources behind funding the Global Autonomous Language Exploitation (GALE), with the goal of “developing a set of software engines that can transcribe, translate, and summarize both text and speech without training or human intervention” and, more specifically, “to address the lack of qualified linguists and analysts who know important languages like Mandarin and Arabic.” As reported on Ars Technica on November 9th, the software is being developed by teams of contractors (including IBM, SRI International, and BBN Technologies):
[T]he transcription engine must be at least 65 percent accurate and the translation engine must be 75 percent accurate at the first milestone. The final milestone in the program is 95 percent accuracy in both modules.
I have been rolling those numbers around in my head: Am I 65 percent accurate in my translation of Charlie’s language—of Charlese? And what about of his non-verbal communication, from hums to the way he holds his shoulders? I hope (hope) to be something more than 65 or 75 percent accurates—85 going on 90 would be nice.
DARPA’s stated goal in developing the GALE is to eliminate “the need for linguists and analysts.” In the case of Charlie, I hope that the need for “human intervention”—for real people, teachers, therapists, translators—never ceases to exist.
Translating Charlie takes a good ear, and an eye for the unprogrammable.
Related Stories
POSTED IN: Language, Military, Money, Technology
0 opinions for Translating Charlese
No one has left a comment yet. You know what this means, right? You could be first!
Have an opinion? Leave a comment: