If you’re like me, one of your favorite pastimes is plugging some of the hilariously besnargled (new word) language on Engrish.com into a text-to-speech program and prank calling your friends with it. For example:
…and this beauty…
But why do the robot voices sound so silly? Science fiction told us we would have HAL-like computer voices that sounded like native speakers — rather than clipped guttural noises from the bottom of a tin can.
Over at Slate, Farhad Manjoo looks into it and finds that pronunciation is only part of the challenge for robot voice designers:
Because human speech is extremely varied, too complex and subtle for computers to understand and replicate. Researchers can get computers to read words as they appear on the page, but because machines don’t understand what they’re reading, they can’t infuse the speech with necessary emotion and emphasis.
Consider this simple exchange:
I’m going to ace this test.
A human reader would understand that the second sentence is meant sarcastically. So would a duplicitous machine like HAL 9000. But today’s computers wouldn’t get it; a robot would think the guy really was going to ace that test.
In other words, robots simply aren’t sarcastic enough to communicate effectively with us humanoids. Which I’m sure they’ll be really upset about as they’re harvesting our spinal fluid for lubricant.
Read Me a Story, Mr. Roboto [Slate]