From a computer science perspective, what should Google do to train its models in a privacy conscious way?
Use employees, paid testers and, maybe, informed beta customers. Like every other product.
Google employees are listening to Google Home conversations
51–60 of 463 posts
Re: Google employees are listening to Google Home conversations
#52Serious question: how do people think the ML models for Home, Alexa, Siri, etc are trained, if not with human labeling?
Doesn't mean you can use live data for tests. Should be opt-out by default, with the chance to opt-in if you wish to collaborate with your data.
Re: Google employees are listening to Google Home conversations
#53The one thing about these stories that keep coming out about the home assistants... they kind of create the impression that this is an issue specific to home speakers, and you can avoid it, by simply not buying them. That's misleading. Any voice command you use to operate any internet connected tech gadget, from phones to smart TV's, is potentially stored and flagged for human review. You really have to avoid using v…
And not to be anywhere near anyone else's listening devices. Isn't there a law in some US states that there needs to be consent before recording someone? How does this fit in and who would be held responsible, the owner of the listening device or the company behind it?
You already consented by agreeing to the terms of service. If someone else is talking to their smart device while you're talking, it's ostensibly their responsibility. There's no reasonable laws that prevent you from being overheard in the back of a recorded phone conversation
Re: Google employees are listening to Google Home conversations
#54Home, Siri, Alexa, M, they all do. I have friends that work on this field transcribing the audio, and measuring its accuracy. Sometimes it's multiple layers of contractors: An employee hands the task to a contractor, another contractor verifies the speech to text, and they're all managed by a contractor. Search for languages like Portuguese, Swedish, Chinese, etc on LinkedIn and you'll find the jobs posts https://www…
"They all do" ... my understanding is that this expressly does not happen with HomePod conversations.
So I meant "they" in the sense of the companies, not necessarily the home devices. Sorry about the confusion.
I know 100% that it happens with Siri. If Apple is excluding HomePod conversations from Siri's dataset, that I don't know.
Re: Google employees are listening to Google Home conversations
#55From a computer science perspective, what should Google do to train its models in a privacy conscious way?
>From a computer science perspective, what should Google do to train its models in a privacy conscious way?
Install these devices in the homes of google employees, executives and offices and allow the public to listen in. What’s good for the goose is good for the gander and all.
Maybe when google has trained the systems enough to not need to train them by collecting and listening to customers conversations, then they enter them into the stream of commerce.
Re: Google employees are listening to Google Home conversations
#56I mean. Of course they are. Do you expect to be able to do any meaningful level of training on data that hasn't been properly labeled? At some point, a human has to go in and correct the software when the software gets it wrong. If you want services that do what Google Home does, you have to have this. Even with that, I'm sure that the engineers are flagging voice requests that happen more then once, or where some on…
> I'm sure that the engineers are flagging voice requests that happen more then once
As a user, I expect to see feedback declaring that it's having trouble and specifically requesting permission to send my unparseable requests to some review queue. I expect to see feedback indicating what could resolve the issue: was it something I said? was there too much noise in the background? was there a software fault and a fix will be deployed at such-an-such date?
The fact that I do not get any feedback whatsoever when things go wrong leads me to have no faith that the problem will be resolved in any satisfactory manner. When combined with a complete lack of consumer/business interaction options in general (w.r.t. Google), it leads to some very dissatisfied consumers.
You think engineers are flagging voice requests that happen more than once? How would engineers have access to that data in the first place if it's supposed to be anonymous and private?
Re: Google employees are listening to Google Home conversations
#57I mean. Of course they are. Do you expect to be able to do any meaningful level of training on data that hasn't been properly labeled? At some point, a human has to go in and correct the software when the software gets it wrong. If you want services that do what Google Home does, you have to have this. Even with that, I'm sure that the engineers are flagging voice requests that happen more then once, or where some on…
It's precisely knowing how software works, and the limitations of the current technology, what makes it so creepy. With this system there's no way of avoiding humans hearing a clip they shouldn't have heard.
Re: Google employees are listening to Google Home conversations
#58From a computer science perspective, what should Google do to train its models in a privacy conscious way?
It's not really a computer science issue, per-se.
Operationally, though, here are a few thoughts:
- Ensure that the persons doing this review do not have physical access to the systems outside of a secured environment (one in which outside audio and video recording devices are not allowed, and for whose presence is monitored, with physical access controls, etc.) Basically, not remote workers or a typical office environment. Most finserv call centers do this, so it's not particularly crazy to think they can do the same.
- Mask the voices such that they are intelligible but not identifiable. Maintain a limited set of "high access" taggers who can hear the raw clips if there is an issue with the masking.
- Limit the length of the clips (sounds like they already do this).
- Have pre-filters for anything personally identifiable in the audio. The metadata for the audio might already be de-identified, but what if the audio clip consists of the person reading out a phone number, credit card number, username, etc.? They should have their "high access" team building detectors for that and flag those portions of the audio or whole clips and route them to a limited access team.
- Make it more clear to customers that their audio, including accidental captures, can and will be sent to their servers. Make this very explicit, rather than burying it in TOS and using terms like "audio data". "The device may accidentally interpret your bondage safe word as a trigger and send your private conversations to our real-live human tagging team for review."
- Provide a physical switch that can temporarily disable the audio recording capability.
- Pay money, like cigarette companies do, to help fund a public education campaign that informs the general public about these listening bugs and mass surveillance issues so that people are aware of industry practices and how it affects them.
Edit:
I like what others are saying about explicit opt-in, as well as paid end-users. For quality/safety control, I don't know that they can exclusively use paid end-users. They probably need to sample some amount of their real live data. For that, explicit opt-in makes sense.
Re: Google employees are listening to Google Home conversations
#59My biggest frustration is going to a friends house which is blanketed with Echo/Google coverage and realizing that all my words could be used against me at a later time. How does one approach this situation?
I mean, it's roughly the same level of exposure you run into literally every minute you're out in public anyway - all those phones around you have microphones, and more and more, they're always listening in exactly the same fashion. This is how always-on "Hey Siri" or "Okay Google" works. Which isn't to say it isn't frustrating, just that your frustration at your friend is misplaced - they're not exposing you to anyt…
Discussions held in private can be very different from the ones held in public.
I think it's very human to want privacy. Having confidence that someone isn't listening/watching to you can have a tremendous effect on how you act. It's just human and I'm not sure if you can change that by telling people that it's not feasible in today's world. I know I acted differently in private when I was just a kid - it's not something I learned, but was natural to me.
The tech would actually be fine if it was provably contained to the local network. As soon as it leaves the local network, I can't trust it - doesn't matter if Amazon/Google says they don't use it for X purpose.
Re: Google employees are listening to Google Home conversations
#60Earlier quoted context omitted.
And not to be anywhere near anyone else's listening devices. Isn't there a law in some US states that there needs to be consent before recording someone? How does this fit in and who would be held responsible, the owner of the listening device or the company behind it?
You already consented by agreeing to the terms of service. If someone else is talking to their smart device while you're talking, it's ostensibly their responsibility. There's no reasonable laws that prevent you from being overheard in the back of a recorded phone conversation
Not in the case of the device being someone else's. I don't have any Google or Amazon smart listeners, so I never accepted them. Yet if my voice is in any of these recordings, well...