I have a bit of news…
I’m trying to learn to enter information in a new way, hopefully faster and easier to edit than my previous fumble-fingered keyboarding. I don’t know how well this will work. I’ve only been at this since late last Wednesday. (Today is Monday, February 15, 2010, although you will be reading this somewhat later.) If all goes well, I hope to be able to create many more documents faster than I ever have before.
What I’m doing is learning to dictate. Using a program called MacSpeech Dictate and a USB microphone from Plantronics (how interesting that the program knows both its own name and the microphone company), I can both order the computer around—give commands—and dictate text right into word processing programs and even text fields in, say, web browser pages.
I know I’m just kind of wasting space and time by saying this, but I’m hoping to get the computer to learn to understand me effectively and efficiently. Things went adequately last week. I did create most of the posts for Friday, Saturday, Sunday, Monday, and one other post orally. I started dictating two different stories that I’d already started writing longhand—one about The Tourist (this one set in San Francisco) and the first several chapters of my François Villon novel. I also started a brand-new Tourist short story, hopefully almost flash fiction, this one set at O’Hare international in Chicago.
Let’s try a test: François Villon. Last time I tried to do that—before “training” the program—it came up with “false wall” and something I don’t even remember in place of Villon (since it just screwed it up, I found out the phrase that it likes to insert is “the home”). Now you’d think that it would know how to interpret that name, since I dictated about five pages on Friday. However in starting up this morning, I suddenly had more than one “profile” from which to select for the speech recognition program to use to recognize my voice and my words. Deciding to have MacSpeech Dictate delete the extras, the program suddenly froze and quit, and in restarting, I suddenly had to start all over again from the very beginning. So the software is having to relearn some of the things that we had gotten started on last week. Frustration, frustration, but at least I’m not typing!
Let’s see if it’s learned anything: François Villon. The software inserted it perfectly (after I tried training it from the first two uses of the name), and so we advance.
Other issues: I have to be careful not to make extra noises, which the software interprets as words like “the,” “and” or other short, simple, one-syllable words. Although I’m trying to work in silence except for my voice, I was playing music last week and it still heard my voice very clearly. Right now, I have a heater running to try to keep the temperature above 54°. It’s still doing very well hearing only my voice (I’ll have to give it a shot with music later today). I also have to behave in a well-trained manner.
I think we are doing well today because the software got me well-trained in those now lost, earlier sessions last week. I have learned some of the commands it wants me to use; I have also started working in a way that is suitable to the software (not very exhilarating and fairly demeaning to my humanity). And I’m becoming much more conscious of my punctuation. After all, I have to tell the software each and every punctuation mark I want to insert as I’m dictating, or else I must tell it in painstaking detail where to go back and click to insert.
The principal frustration for me is that the software wants me not to use the keyboard at all (at least while dictating text). MacSpeech Dictate maintains a “cache” document of everything I’ve said and all the punctuation marks and, I guess, all the edits that I have done. If I try to make any edits manually—using the keyboard—, the cache gets all screwed up. Screwing up the cache makes me a very bad boy: it breaks the so-called “Golden Rule” of computer dictation—“when dictating, have the voice do everything.”
I guess, except for the software dominating me, I’m having fun.
Since I’m not making any money from doing this, I guess I should get back to “writing” some more fiction. After all, I still have a lot more of “Mantorville” (a word the software did know last week, but which it doesn’t know now) to make up for you. I’ll make sure some near-future post does indeed get us back to Quetzal County. So I guess it’s “Hi ho hi ho it’s off to work we go” for now.
Adding a bit, later, shortly before it’s going to post: recent “auditory errors,” such as the one that cropped up in the original posting of Mantorville, part nine, are the result of dictation. For tomorrow and Sunday, I have a new—well actually quite old, but recently rediscovered—bit of writing that I was excited to find (even though I have already told you it was long-lost). I dictated four whole chapters (nearly 12,000 words) from a xerox copy in just a few hours, which is a vastly superior input record to retyping stuff, and I hope I have located all of the “auditory errors” (which are becoming fewer but still surprise me) before the posts appear. You judge.
I used to tell students that keyboards would some day—probably in their lifetimes—have become antiquated technology. I never realized the keyboard might be out of date while I still sucked breath here on earth. But I am finding I can do and write many things without my ineffectual fingers (and I begin to dream about a Bluetooth headset/microphone someday liberating me to pace where I wish while I write).
Sometimes I love living in the future.