We all know that texting while driving is not only extremely dangerous, it’s also most likely illegal (depending on where you live). However, your Windows Phone offers a nice alternative.
By visiting Settings -> Speech in your phone, you can turn on a setting called “Play audio confirmations”, then tweak when your phone reads incoming text messages to you. Once you’ve turned on that feature, your phone will audibly announce any incoming messages to you, then ask you if you want to read the message or ignore it.
If you choose to have your phone read the message, the friendly Windows Phone “Speech” voice will read the message aloud to you, then ask what you want to do next: “Reply, Call or I’m Done”. If you say “Reply”, you’ll be prompted to speak the response you want to send. Your message will then be transcribed and read back to you. If it’s correct (which it seems to be about 80% of the time in my experience), you can tell your phone to “Send” the message; if it’s not, you can tell it to Try Again.
To send a message without first receiving one, you can use the standard “Speech” feature by holding down the Windows button on your phone. When prompted, say “Text…” followed by the name of the person in your contact list that you want to text.
For me, the speech-to-text/text-to-speech features are invaluable while I’m driving. I keep my phone up on my dashboard while I’m driving down the road, and rarely have to touch my phone (let alone look at it) when having an SMS conversation with someone.
There are some minor downsides to turning these features on. If you have your phone set to always read incoming messages (which I do, since I don’t have a headset), it will always announce when you have an incoming message (and the name or number of the person that sent it). This can be a little annoying when you have your phone plugged in on one side of the room and you’re working in another area. It can also be a bit embarrassing when you’re in public.
You can stop it from reading out this information to you by simply putting your phone on vibrate, but that’s not always an ideal solution.
You also have to speak very deliberately when talking to your phone, otherwise it will most likely misinterpret what you’ve said. Some of the transcriptions of my commands are pretty humorous. One of my standby responses when my wife texts me is “I guess”, and my phone rarely gets it right on the first try. It’s interpreted it as “high gas”, “hi dad”, “i gas”, etc., but rarely actually gets it as “I guess”. You also can’t really use abbreviations, as it doesn’t understand them. For instance, whenever I say “K” instead of “Okay”, it does strange things with it. Sometimes it will transcribe it as a word that sounds similar to “K”, and other times it will just interpret it as various symbols like “@$#%” (presumably because it thought it was a cuss word).
Finally, you can only use one sentence per message when responding by voice. The speech-to-text feature doesn’t seem to have any concept of punctuation (though the text-to-speech feature works just fine with punctuation), so you have to think of a way to say everything you want to say in a single phrase.