Hacker News new | ask | show | jobs
by ethbro 2122 days ago
FAQ on Tone component: https://blog.aboutamazon.com/devices/a-new-tool-to-help-you-...

"Once you’ve opted-in by creating your voice profile, Tone will run passively and intermittently in the background so you don’t have to think about it. Throughout the day, it will take short samples of your speech and analyze the acoustic characteristics that represent how you sound to the people you interact with. This gives you a simple way to reflect on your interaction and communication throughout the day."

Aka "Look, Amazon found a reason it needs to record your voice all the time. For science."

Facebook must be fuming.

3 comments

> Facebook must be fuming.

This gave me a good laugh.

> This gives you a simple way to reflect on your interaction and communication throughout the day.

This sounds like a great way to cause social anxiety. It's also a very poor excuse for data harvesting. People are going to love it, this thing is going to sell like crazy.

>> This gives you a simple way to reflect on your interaction and communication throughout the day.

Isn't trying to overthink every communication you had throughout the day going to be unhealthy rather than being helpful? People who are people pleasers are on the risk of being addicted to this thing.

Yep, that's exactly what social anxiety is: you stress out over how you have interacted with people over the day and their perception of you because of your interactions. It's a genuine mental illness, and shit like this is how you contract it.
This device is intended to help with social anxiety issues.

In real-time, while you're currently interacting with someone, it can offer a helpful "WARNING! YOU SEEM ANXIOUS. CALM YOURSELF. OTHERS MAY CURRENTLY BE PERCEIVING YOU AS ANXIOUS."

Problem solved. Thanks, technology!

Holy crap. They are openly saying they will now be recording and listening to (and transcribing presumably) your speech at any time.

BTW, pretty sure Facebook and Google are already listening to speech and serving ads off it, just not admitting to it.

Where does it say transcribing? To me it sounds like it is processing voice on device into emotion markers and only the emotion markers are send to amazon. So the actual voice information would never leave the device. I do think anything else would not be possible legally because you'd need to have consent from every stranger you have a discussion with, moreso what about trade secret meetings. I can't fathom Amazon would go that way.
It doesn't. However for the last nine or more years (since work on Echo began), transcription has been far and away the #1 way Amazon processes speech - not "emotion markers" or whatever else you are dreaming up. I don't see "emotion markers" mentioned or described, and "it sounds like they are doing something else" is not really a strong argument if you're trying to guarantee that they aren't transcribing it.
Not just "your" (customer's) speech but anyone else you're talking with. Piece by piece we are accepting more and more of this kind of invasion of privacy.

Also how is this ompatibe with laws where both parties must consent to a recording being made. Perhaps Amazon would argue this isn't actually a recording because... reasons.

> your speech at any time.

And that of others who happen to be around, presumably.

> it will take short samples of your speech

Unless the microphone is extremely directional, this means that the device will also take short samples of what everyone around you is saying. I assume Amazon's lawyers have passed on this, but it sounds sketchy to me especially since Amazon doesn't mention it.