Recently my wife and I were talking with a friend who teaches in the area of special needs students and special education in mainstream classrooms. During that discussion I discovered Google docs has a voice to text feature. So naturally I had to try it. And while I can’t say I’m throwing away my keyboard for a microphone there are times that this feature could be worthwhile.
Why I like it
There are times, like doing video scripts, that I want to write things that sound more like someone speaking naturally. In these cases it is easier to say it than to type it. For some reason when I type or write long hand I get into a mode that is more rhetorical and “printed word/texty” than I want. When I talk into a mic it’s easier to avoid that.
Also, there are times when I want my hands free while I’m writing (like when I’m trying to write up a craft project and need my hands free to do making stuff). Google docs voice feature is a free tool that helps with this.
Actually when I tested it the voice to text feature worked surprisingly well. It was able to have a fairly good level of accuracy in translating what I was saying. It was actually able to translate phrases like four in one chainmail without stress.
Of course the feature did have quirks and it wasn’t entirely a solution to my problems.
What I’m not happy with
The core engine driving the feature seems to be the same one that converts phone calls to texts for google voice. If you have ever seen the ‘creative’ resolutions that happen with that google feature you can imagine what happens if you cough, mumble, or pause mid word. You may also run into problems with more unusual idioms, phrases or words.
You will definitely need to do your editing because the software also occasionally swaps words for other words. In my test case the voice to text feature kept using the word ‘cloths’ for the word ‘close’…
The system is also light on punctuation and formatting options so you may need to put those in later. I was able to get a period by saying the word period, or a comma by saying the word comma. But if I said the word semicolon I got the word semicolon and not a punctuation mark.
Similarly you get a new line by saying “new line” unfortunately if you want a blank line between paragraphs you have to add it in later or say “new line”… wait… and then say “new line” again. This does tend to slow things down a bit. So you may be better off just accepting the fact that you will have to do your formatting later.
I found myself wondering what would happen if I wanted the word ‘period’ instead of a punctuation mark. I experimented and my results were mixed. If I talked about a woman’s period I got the word period. If I said “periods” the software would write “periods”. But if I talked about a trial period, or the colonial period, or said that the program “was the best software period” I got a punctuation mark.
And… Things kind of went downhill from there…
After finding the program’s selectivity about the word period I found myself wondering what would happen with other words.
Naturally the first place my mind went was the old F-bomb…
When you use that particular word you get f***.
You will also end up with c*** and a variety of other similar items when you use words that might be offensive to women. But oddly enough the word ass is apparently ok.
At this point I was feeling a bit wierd swearing at my computer, but hey if you’re going to go you might as well go all the way…
Oddly enough while Google docs voice tool seems to want to “bleep you out” if you say something insulting about women; it seems to be just fine with racial epiphytes against blacks, Jews, Italians and others.
At this point I was both surprised and offended and decided to stop.
There are times that the voice to text feature is really useful. And I will use it in those cases; however, the feature does not have a full range of punctuation and text formatting features (and you know how much I love those parentheses…). This part makes it even more important to take the time to reread and edit the things you write using this feature.
And of course watch out for those wrong word situations…
I also find it odd that the software actively filters things that might be insulting to women, but seems to be fine with users insulting blacks and Jews (as long as they’re male of course…).
It’s also a good idea to remember that Google docs are stored online and may be more easily observed or pirated than things you keep resident to your machine.
In the final analysis I would say that if you think the idea of talk to text could be worthwhile for you, then give the Google docs voice tool a try. But realize that it does have its quirks and will increase your editing load. You may want to move on and try other speech to text software but the google docs voice to text feature is a starting point.
I would also love an explanation of what they will and will not censor (it’s a free tool so they can make their own choices… I just find the choices that have been made interesting).
That’s it for this one. See you next week