In case you missed Frank McPherson’s comment on my previous blog rant on both speech and handwriting recognition losing to thumb keyboards…
Frank mentioned a web service I had not heard of before: Jott.com (not to be confused with JotSpot which has a jot.com URL).
Founded by two ex-Microsoft-ies, this web service lets you call it from your mobile phone, transforms your voice message into text, and then emails the text back to you.
I’m tempted to try it except for one little problem… I’m always leery of giving out my phone number to any web service. The thought of a security breach that reveals all phone numbers stored by a service or an errant piece of software that starts dialing madly concerns me a lot (obviously). The Jott service requires you to provide your phone number. It probably uses Caller ID to identify you to determine which email address to send text of your voice message.
If you are less paranoid than me and have tried this service, let me know how it is working for you.
BTW, you can find Frank’s own blog covering mobile technology at: Pocket pC Hints and Tips.
Comments
5 responses to “Jott.com: Speech to Text via Email”
I’m not so worried about publication of my phone number as I am about the conversion process. Reminds me of a Jerry Seinfeld episode and the translation of mute signing. I wouldn’t lean to heavily on the service if World Peace were at steak, — er stake.
I went to Jott.com.
regarding the cautious nature of the originator of this thread, re: phone numbers, I feel similarly. BUT, I read through their agreement/contract.
based on that alone,I cannot sign up for this. If I have to come back to continually read agreements, and be subject to unspecified ads, etc, ten this is not for me. the service sounds good, but ther is no free lunch…
Jott just keeps getting better and better. I really like the Reminder feature.
If it does misjott you at some stage, please let us know at http://misjott.com.
Here’s another use for this. We put up this just to test it. http://res.realestateshows.com/jott
I have used Jott, and it works well (if you leave it in the default setting). Basically, by default, they use a mix of automated speech recognition and if that doesn’t work, they use human transcriptionists. If you turn off the human option, it’s very difficult to have a message transcribed properly. I’m not sure how their service will scale, but I love the idea.
I have switched back to using VoiceMinder, by WebIS instead. I have it assigned to a key on my MDA and after I create a voice note, it gets emailed to me automatically. I love that!