C

The Rise of Voice Technology

Voice technology has come a long way. Just a few years ago, it would have been unusable. But now, those who follow the technology know that it has got considerably better.

Writing with your voice raises several interesting questions. How difficult is it actually? Human speech involves a lot more starting and stopping with errors and the need for repairing broken sentences than you may think. Even gifted speakers make mistakes. To turn the spoken word into reasonable writing requires lots of planning. You'll need some kinds of notes or other organisers to make it work.

Another question turning speech into writing raises is the style. How would writing make the change that people speak their writing rather than type? Chances are that it would come up with many more short sentences and more concrete language, which is good. It would probably also rely on prepared phrases a lot more often, which is not available when you are speaking quickly.

To confirm this, a column was not written, but dictated (听写). It was composed from brief notes written down for structure only, and it was edited for length, with all of the original errors kept in. Here were the results. The first was that the literal accuracy was extremely high. There weren't many cases where the software had heard one word incorrectly and written down another. But the other result was that the readability of this column was rather bad. Obviously, the blame is not with the technology at all, which turns out to be rather good. Speaking into writing relies on a better human brain than the one we currently possess. Writing is hard. There's a reason it can't be done at the speed of speech, in real time.

To clarify the matter, this time paragraph breaks were added after the whole writing. Punctuations (标点) had to be spoken aloud, and after a full stop, the first word in the new sentence was capitalized automatically. Some minor punctuation marks were added to make it clearer. To improve accuracy, people "trained" the software beforehand, reading a prewritten passage aloud. Actually it turns out to be more effective. All of these ensure the satisfactory completion of turning speech into writing.

Language is the most important tool for human interaction, and voice is one beautiful part of language. With the maturity of modern technology, it has given birth to a great change in the human-computer voice interaction.