Exactly... voice recognition for texting != conversational artificial intelligent assistant. The frustrating thing is that my viewing their website gave them some amount of ad revenue. I want that money going to accurate reporting and insightful blogging.
To diffuse some of this, the comparison to Siri is that it's a conversational voice interface.
We're actually surprised that people are picking up that we're optimized for texting while driving and that Siri is a personal assistant (perhaps it's more obvious in retrospect given these comments). Our first expectation was that people would think "oh, how is this different from Siri again, except worse?" and that we'd have a hard time differentiating ourselves.
"Conversational?" Conversational like the automaton I get when I phone my bank? It looks to me that the system is a canned response / limited expected response / state machine that every motherfucker on earth can do. If you want to do marketing bullshit, HN is probably not the best place to do it.
We're actually surprised that people are picking up that we're optimized for texting while driving
You are? You are surprised that people are picking up on your core selling point? Why? Did you not feed this to them?
We're actually surprised that people are picking up that we're optimized for texting while driving and that Siri is a personal assistant
Oh, I see. "Gosh, we're surprised that <some really obvious thing> while <some other thing that we wish to be perceived as true but isnt>".
It would be great for you if the press would frame the discussion as "sonalight is for texting while driving while siri is a personal assistant", because it then sounds like siri doesn't totally wipe the floor with you when it comes to texting while driving. If only saying such tripe online would make it true - but maybe HN isn't the place to try it.
When I said conversational, I meant that our app talks to the user, waits for a reply from the user, then does that cycle again, like Siri does. There are a lot of other ways to do voice interaction, like voice actions does (you press a button and then talk and then something pops up) or Vlingo (where the interaction is mostly hands free, but sometimes requires you to press a button).
I'm also serious that we thought that people wouldn't understand the differences between us and Siri. I'm not talking about people commenting on HN, I'm talking about someone who casually hears about our app as a "voice app" from a friend, or who glances at the headline on TechCrunch. The differences are in the details of the execution, and they can be subtle, especially if you haven't gotten a chance to use both apps and do a comparison on how they perform at texting-while-driving.
Come on, this is HN indeed, can't you assume a little good faith? They're here answering questions, trying to learn from potential customers.
Have you launched a product? It's easy to call it "marketing bullshit" and move on with your day. Those people have spent time and energy designing their app, and they didn't post an ad, someone (maybe them, who cares?) posted a link on a blog about their product. Just ignore it if you don't care for it.
If he'd come on here and talked about their marketing strategy, such as how they are having to deal with a vastly superior product but which fortunately their customers don't have access to, then I'd be all over it, and HN would be a great place to discuss it. But to come here and actually attempt such transparent bullshit gets such transparent bullshit pointed out.
Text by Voice allows you to safely text while you drive entirely through voice. There's no need to touch or look at the screen at all. Text by Voice will automatically read incoming text messages aloud and give you a chance to respond. You can also set an auto responder for incoming text messages, and have Text by Voice start automatically when you start driving. Text by Voice lets you compose texts on the fly. It runs in the background so you can use other apps at the same time. Drive more safely with Text by Voice.
Search Keywords
Text by Voice, TextbyVoice, Voice Recognition, Speech Recognition, Hands Free, Speech, Text, SMS, Drive, Driving, Text While Driving, Sonalight, Sonalite, Sonilight, Sonilite, Sonolight, Sonolight, Sonarlight, Sonarlite, sona soni sono sonar light lite text by voice textbyvoice
Email dev@sonalight.com
PHONE SPECIFIC NOTES:
-Tested fully with no problems on Motorola Droid, Motorola Droid 2, HTC Evo, HTC Droid Incredible, HTC Droid Incredible 2, Nexus S, and T-Mobile G2
-Samsung Galaxy S microphone has lower quality audio
The limit is only for composing a text. You can receive an unlimited number of texts. Did you get the impression that you could send unlimited texts for free? (serious question- it's not our intention at all)
Sorry, I definitely meant to write 'composing' not 'incoming'.
That is really the impression I got as it's not mentioned anywhere in your description. I was wondering what your monetization scheme would have been. I predict that you will open this up to unlimited texts and charge for other actions.
We're going to be more flexible with our monetization in the next few weeks. $20/year for unlimited was a last minute addition to the launch so that we could get feedback like this. We don't want to pigeonhole ourselves into one model just yet. Opening it up for unlimited sending and charging for other actions is definitely on the table.
My other piece of advise is that I would stay far away from being compared to Siri. You will by definition always lose, as you are not processing contextual statements. Just bill this is a (big)improvement on Google Voice Actions.
1. Call <Name> <Loc>
2. Call <Number>
3. Send Text <Name or #>
4. Send Picture <Name or #>
5. Lookup <Name>
6. Go to <App>
7. Check <Item>
It only understands these 7 commands. The voice processing wouldn't be good enough to compose a text message, #3 just sets that up for you to key in, but it's doing the job on a 50 MHz single-core ARM9 CPU instead of uploading it to a server somewhere. I have no beliefs that the above task is anything like Siri.
Siri is not about turning speech into text. That's been done already, everyone else has a button on their phone that does this, and you're offering them this familiar (and often broken or bad) functionality in an app for Android (Note: I'm not implying that your app is broken or bad). As long as your app is like what people are used to, you'll have a hard time of sales. Once you cross over to the functionality of the well-marketed and Apple-branded Siri, you'll do much better. Based on the description of your features, you're not there yet.
Our sense is that people don't want to use speech for everything ever, they'd much rather use their hands and eyes when they're available. We're laser focused on texting while driving right now, it's the one big use case for voice commands.
We've done a lot of user testing and have found the biggest problem with voice products now is that people try them twice or three times, and when they don't work they decide to give up on using that product. No one wants to spend time learning a user interface. So we've optimized our product for the first time user experience for texting. We're offering what voice what Dropbox did for cloud storage. There's nothing novel technology wise per se, only that the product is optimized for ease-of-use.
With regards to Siri, we're interested to see how it'll perform once it gets released into the wild.
We're laser focused on texting while driving right now, it's the one big use case for voice commands.
Regardless of whether your app is same/better/worse than Siri, please please please don't pursue a use case around texting while driving.
I'll skip my rant about the fact that texting while driving is dangerous even if you remove the keyboard input (you still gotta read the reply messages @ 65mph).
(EDIT: Ok, I see that your app will read texts too. Still doesn't negate the following:)
The issue you face is that as a second-class citizen app (ie not Google or the device manufacturer) on my phone, here is my workflow to text while driving using your app:
1) I decide I want to text
2) While driving, I have to feel my way around my phone to press the fiddly "power" button at the top of the handset (let's assume it is in a cradle)
3) Oh, but wait I have screen-wipe security enabled, so I now got to dance my finger across the screen - again, while doing 75mph (yes, I like to drive fast up the 280)
4) Ok so now I'm in. But damn, I'm at the home screen. I got to find your app - we'll assume I put it on the backdrop of the home screen. I press it and wait for the app to load.
This all assumes that I put your app in a primary positon on my phone, that I wasn't already in another app and so I had to press the home button, that the home screen wasn't on a different 'pane' then the one your app icon is on, and that your app starts recording voice commands as soon as I open it up.
Sorry, but that's still A LOT of steps, is dangerous, may not be legal in California due to it being more then one button and makes a lot of assumptions.
Google and the handset manufactures, who give themselves first-class citizen access, have the advantage because they can add extra buttons for one-press access. They can also use private APIs to potentially remove the screen lock (although that has other issues associated with it). Apple, when they integrate Siri into iPhone 4S, will presumably have similar advantages.
Sorry, I'm all for the little guy and the 3rd party ecosystem - but I'm also for safety. As a driver (and motorcycle rider) I really fear people using tools like yours while driving.
For that reason, I'm sorry, but I really hope your app is a failure.
We agree with you completely, we are very much for safety.
We are against people touching their phones while driving (it's illegal in most states) and designed our product specifically so you don't have to touch your phone.
It auto starts up when it detects you've begun driving via your speed from GPS (go to Settings in the app to enable the auto start up). We also designed this GPS detection to not suck battery life (it will take about 1% of your battery every hour).
It works on phones even if the phone is in the passenger seat next to you- you just have to make sure to talk loudly enough.
It will even pick up your voice if your phone is in your pocket (again, just speak loudly enough) and you can hear it talk back to you.
No touching your phone or taking it out of your pocket required.
edit: also, no need to downvote this guy, it was an honest misunderstanding.
If you have the "turn-on when driving" setting enabled, then our app will acquire a CPU wake-lock that keeps it running (this is not a big drain on the battery, about 1% per hour). Every three minutes we'll poll the GPS to see if you're driving and then start the application up if GPS indicates your speed is above 10mph. When it starts up it'll also notify you that it started up by speaking aloud to you.
If your phone is entirely off (and not just sleeped/screen locked) then there's nothing we can do about that... :)
Also if you have more questions, start a new comment thread in this post- the reply timeout makes replying super slow.
