FOOD FOR THOUGHT Google STT Reboot (Voicemail Transcription)

[deltaend]

Hey guys, I'm back and I have some great news for you! I am working on a new version of our Google STT which is faster, cleaner, better, and comes lemony fresh. This new version will accept all lengths of voicemail up to 5 minutes and will break up your pager text messages into up to 5 different messages (like Vonage) if you want.

The system will come in two flavors.

A free version which will simply be an email address that you send the voicemail to. The system will then email your target email address/pager address with the MP3 and transcription. This system will be limited to 10 transcriptions per day so for the hobbyist it will be perfectly fine.

The paid version will be $(TBD) for unlimited use in your company/household. You'll be able to put this on as many servers as you want to as long as you don't give it away or sell it to anyone else (funding us helps to keep this project going). Minor revisions will be free, major revisions will cost $(TBD)/2.

Do you think this is a fair model? Please let me know.

 

[wardmundy]

Kinda depends on what TBD turns out to be.

 

[Hyksos]

I'm not your lawyer nor your business advisor but you're planning on selling a product that is based on a Google developer API that is not Public nor stable nor guaranteed to remain available?
It's pretty clear up to now that Google prohibits this kind of things.
It would also be required to advise anyone leaving a voicemail that it's going to be uploaded to Google Speech API systems.

The Speech API allows developers to access Google speech-recognition services. It is only available for development and personal use.

Google Speech API Terms of Service

Last modified:September 1, 2012
By using this API, you consent to be bound by these terms in addition to the Google APIs Terms of Service ("API ToS") at http://code.google.com/apis/terms/index.html.

1. Privacy Obligations
   
   In addition to the privacy obligations in Section 3 of the API ToS, you agree that:
   • You will comply with all applicable speech and biometric-related laws and regulations, including any applicable voice consent and interception laws and regulations.
   • You will disclose to users that you are using the Speech API, and that Speech API content will be controlled by the applicable Google privacy policies.
2. Additional License Grant
   
   The license granted in Section 5 "Submission of Content" of the API ToS is extended to include usage for any purpose, including but not limited to, improvement of other products and services.
3. API Clients
   
   You and any API Client you create may only access this API for your own development and personal uses.

Courtesy Limit: 50 requests/day

I won't quote the entire Google API ToS but it's also full of limitations and...well, ToS.

I'm not against your endeavour at all, but I do have some reservations if this is taking a different route then "developers toying with the API while understanding the whole story".

And like I said, I'm not a lawyer but I'm almost certain what you're describing is 100% illegal at this point, at least for some usage scenarios that you said it could be used for.

Hope this helps!

 

[deltaend]

I'm not your lawyer nor your business advisor but you're planning on selling a product that is based on a Google developer API that is not Public nor stable nor guaranteed to remain available?
It's pretty clear up to now that Google prohibits this kind of things.
It would also be required to advise anyone leaving a voicemail that it's going to be uploaded to Google Speech API systems.





I won't quote the entire Google API ToS but it's also full of limitations and...well, ToS.

I'm not against your endeavour at all, but I do have some reservations if this is taking a different route then "developers toying with the API while understanding the whole story".

And like I said, I'm not a lawyer but I'm almost certain what you're describing is 100% illegal at this point, at least for some usage scenarios that you said it could be used for.

Hope this helps!

This seems to be the highly relevant:

You will disclose to users that you are using the Speech API, and that Speech API content will be controlled by the applicable Google privacy policies.

This makes me believe that this is designed to be offered as a packed service for applications sold or not to other users as long as you clearly let people know. I read through the rest of the ToS and there doesn't seem to be anything that would seem to make this illegal in any way.

Am I missing something? Please let me know.

 

[Hyksos]

im sure you havent missed anything, like i said I'm not a lawyer
but for developper and personal use, doesn't sound like a commercial product to me. It would appear now that you need to sign up for google chrome dev mailing list to have access to the API, which also point to the notion that its not released yet. but that's just my understanding and I agree I could be wrong!

 

[wardmundy]

Plain meaning of "for your own development and personal uses" contemplates non-commercial use of the platform. There is some ambiguity as to whether a person has to BOTH develop the app AND use it personally or whether an either/or situation is OK, i.e. I can use it to develop or I can use it personally. If a user downloads a free app that is modifiable and the app only facilitates use of the API and then the person only uses it personally, that's probably OK. But, if a user downloads a commercial app that you've developed and then pays you a fee for using it (which is really a disguised fee for using Google's API ), then that user is really a customer. Any other interpretation would render the restriction meaningless.

My WAG is you'd probably be in hot water regardless of notification to potential customers. This is not legal advice, but you probably should get some before moving forward.

 

[deltaend]

True to all of that Warmundy and my thoughts exactly, however, I have some thoughts on this that you may have not considered:

1. The version of the product that uses my API is for personal/development use and not for a production environment (my previous terminology of "hobbiest" was simply different).
2. The version of the product that people may purchase is for their use and from my perspective, licensed to be used on as many machines as they want. If they break Google's ToS, that's not my fault as they will separately have to agree to the Google ToS for the API and create a new API for each machine that they use this software on.
3. All of Google's API's are designed for eventual commercialization. Just because Google doesn't want people using more than 50 queries per day and just because they say that you can't use it in a corporate environment doesn't mean that it will always be this way. In fact, following Google's general development and product cycling style, I find it likely that they will change this API's restrictions by end of the 3rd quarter this year (if not sooner). If there was never going to be commercialization of this API, Google would never have created it in the first place. However, it is Google so tomorrow they might decide that Chrome and Voice are both not cool anymore and can both of them.

No matter how I see this, I don't see my proliferation of this add-on script as being anything more than enabling users to use in their own private environment. What choices that they make with it is up to them, not me so I don't see how I can be held responsible if someone acts responsibly or irresponsibly.