I actually bought a speech recognition program for email. I have a large family and continually replying to emails. I've not tried it for writing a ms, but wonder if anyone else has? It seems like the flow of would be hampered, but with practice, perhaps make it easier.
Yes, a strong accent could be a problem, but it's rather clever. One goes through a long [and annoying] testing period where the s/w learns the voice and patterns. You repeat a series of statements and it "learns" your speech. I am quite amazed at how simple and accurate it is for my email needs. But have not tried it for a ms.
I'd be apprehensive to use it. It seems like it would just be too disorienting to never touch a single key and have words still appear, and I don't think I can come up with sentences to write by first speaking them. I think I'll stick with writing on paper.
I bought one when it first came out back in the late 1990s - don't know if they've improved it - I hope so because when I first used it my paragraph turned into some surreal story about a purple hippo dancing in moonlight. I showed my brother the final paragraph. He said - You're writing about a hippo?! I said no, but apparantly it assumes I am. He and I still joke about this. I had to talk really slow and form each word. Doubt this has changed. And I found that it could never keep up with my flow of thoughts. I've tried the usual gimmicks other than just typing - speaking into a mini tape recorder ( I got the idea from a writer in the movie Troop Beverly Hills! Ha ), then the speech recognizer. But I don't think any of them save much time.
Speech recognition software is evolving and getting better. However, I don't think it's anywhere near being a tool which a dedicated writer can use effectively.
You are probably right. Yes, it is surprisingly simple and accurate, but the flow is difficult to manage. However, at 72 with increasingly stiff fingers, typing has become a major effort. So, I will practice.
I use speech recognition software from time to time. It generates fewer mistakes than my crappy typing sticks (fingers), so I can create a lot of output in a session. Drawbacks: cumbersome headsets and/or having to work in a quiet room. I work best with background noise, and even the best headsets are uncomfortable after a while. (I insist on a wireless headset, so I am free to move around). I'm a naturally clear speaker, so that's not a problem for me, and the current generation of speech software is pretty good at choosing the correct spelling of homophones from the context. Also, the more you use it, the better the precision. I too keep a digital recorder for dictation away from the computer, and feed it through the speech recognition software.
