Using speech recognition makes sense when you suffer from arthritis. If nothing else it saves wear and tear on your wrists. This morning it’s behaving itself, but over the last few days I’ve had nothing but trouble with it.
To begin with it assumes my breathlessness, which is caused by COPD, is actually saying the word ‘are’.
Typing the words using the keyboard is a comparatively slow process, even at 35wpm, which gives you time to think about what you want to say. As a consequence when speaking into the microphone there are long pauses while you decide which way you’re going next. If this were a conversation between two people these pauses could be considered long and sulky silences, as if they were a divorcing couple in stalled negotiations.
Furthermore, the capacity for misunderstanding is huge. That sentence above, ‘typing the words using a keyboard…’ actually came out as:
‘Typing using the keyboard is a comparatively slow process even the lights died at 35wpm.’
Disjointed, and it makes you wonder what the lights dying had to do with anything.
It means I’m constantly having to go back and correct errors, which is time consuming. Even worse, it doesn’t always understand the corrections, which in turn means I have to resort to the keyboard to type in the correct words.
The speech recognition system also has inbuilt control commands which are supposed to facilitate production and editing, but which can be an absolute nightmare when the bloody software doesn’t understand my Yorkshire dialect.
As I get older and the arthritis takes a more virulent grip on my abilities, the advantages of this kind of system are obvious, but for the moment I have deadlines to meet and the software is simply too slow to match my requirements. This entire post, which runs to over 400 words, was produced using speech recognition and has taken about three quarters of an hour to write. I’m sure I could have typed it faster. But even when the initial post is written I still have to go back to manual editing using the keyboard. This is because I don’t fully understand the editing process using speech recognition, and even in those areas where I do, the software doesn’t always recognise what I’m trying to say.
And as a final example of its frustrations, I’ve just had a coughing fit which the software interpreted as:
The it had a of the border.
Enough said… or should that be enough typed?
I hope you’re enjoying this blog. If so, why not subscribe to by clicking on the Networked Blogs icon in the Sidebar. That way, you’ll never miss a post.
As always, comments are open. Spammers, don’t waste your time, you’ll never get through.
You can also subscribe to my occasional newsletter by clicking HERE where you’ll also find a FREE E-BOOK waiting for you.