I’d love a “phone” that was just a mobile internet-connected device. I very rarely use it as an actual phone, primarily just for text, email, and web browsing.
Voice assistants are money losing products. If they can do something like processing the wakewords on the device before chosing to send to a server they will. These companies are far too stingy to continuously stream audio to their servers
Exactly. If it is practical and money can be made doing it, then continuous, ambient sound parsing will be the norm. Currently it seems like it’s not a valuable business. When it is valuable to them, they will add a checkbox somewhere in your account to disable it, and most people will not be bothered enough to look for it.
Someone can correct me if I’m wrong but home assistant is currently struggling with this and is processing everything on your local box because it can’t do wakewords on the device.
I think they’re choosing to do it that way. Raspberry pi’s easily have that capability to do the wake word recognition on device (i think they are also working on that). Esp’s on the other hand, can only stream audio to the server and not much more. Since esp’s are far cheaper than installing a raspberry in each room, they are focusing to do wake word detection on the server not on device.
My experiences are much MUCH different. The amount of compute waste is through the roof, and we shrug at +$50k/m provisioning. You don’t even need approvals for that, and you can leave it idle and you MIGHT get a ping from gloudgov after a few months.
Back in the day when everything had to be processed server-side sure.
Now we have purpose-built hardware helping work this shit out. The devices are basically capable of handling native language resolution locally. They’re no longer need to farm the data out. I still don’t think they’re doing this we would see it in the open source operating systems, but if they wanted to any late model cell phone would be absolutely fine parsing out your interests from your conversations. Hell, I’m sure the contents of this dictation I’m making now are being reduced and added to my social graph at Google.
If this is a normal Android app, the solution to this is simply to disable the Google app. pretty sure all the voice related shit and google lens and all that crap is tied to the Google app.
There are actually 2 processors in the devices. 1 that constantly listens for a keyword, Al la, Alexa, Hey Google. When it hears it it quickly spins up another “computer” that then sends your voice back and forth to the servers for processing and response. It’s part of the reason that the listen word isn’t easily customized.
It still stores the name triggers, even incorrect matches (last I checked, which was years back).
The recordings can be played back from account history.
The one time I looked at some random, it was mostly snippets of my conversations with friends.
Creepy.
the makers of these things always say that, but I guarantee that long winded explanation is bullshit. Maybe there’s even hardware in there that does those things, but even so, they’re always listening, recording and submitting everything you say to their maker. Primarily for targeted ads and targeted content of other sorts…but also to snitch on you if the cops accuse you of something.
Onboard are >=2 bits of code. At least one of those is a specific system trained to recognize a “wake word”. This specific system (ostensibly) doesn’t send anything to an outside party. Its entire job is to recognize one wake phrase: Alexa, Ok Google, or Siri, and then if that wake phrase is used it responds and tells the second system to listen. As you can imagine, this is a pretty easy job to get right 80% of the time. So that can be put on a chip. So then it does its job, and it’s the second system that sends everything to an internet service for whatever reason.
why is this downvoted? you cant prove its not, if its proprietry. and since the companies listening just happen to profit off data collection (and break/bend the law often), its safe to assume they do this.
I’d love to have this properly audited sometime. I’d slap like to think that we’re generally protected from big companies doing unethical and unjust things to us, by law, … but nah
(That’s not to say I don’t believe this explanation; the second half of my comment was just an addendum.)
Sound’s like it’s just not sending the data back to Daddy Google. The OK Google/Alexa bit is done on a custom chip on the device. Clearly that bit isn’t being turned off, but anything after that isn’t being sent anywhere.
Probably just saves support calls this way from idiots who turn it off and forget.
obviously what you vaguely describe has been around since 1945.
That home assistant devices are constantly listening and feeding back marketing data on every conversation is patent and disproven nonsense.
they have done packet sniffing investigations, they have disassembled the devices, they have run meters on the electrical charges… everything in every way you can imagine.
But even if you just think about it for a second - processing a live audio feed at a rate of 1 second per second indefinitely and correlating that data via voice recognition to your Google profile all to… make your ad personalizations… worse? more inaccurate?
like what the hell is the perceived benefit? That my wife says, “oh my dad found my old barbie house!” while at my neighbors house and my neighbor gets served barbie ads? Why would Google want that?
At some point ever you’re going to realize is that the real things you need to be afraid of are largely caused by the stuff made up by Facebook boomers.
what specifically? vaccines cause autism/monkeypox, the democrats drink baby blood, trump won the last/next election, Putin is good because he’s only killing Nazis in Ukraine, forest fires are caused by Jewish space lasers, LGBTQ+ folks are grooming children and Bill Gates wants to put microchips in your brain?
Like — what are you saying, some misinformation is good?
I’m saying people believe those things. Roughly half of all voters. And those beliefs cause damage, and it will affect you, whether you think it’s stupid or not. You can ignore it and insult it all you want, but it’s not going away. Perhaps you’ve noticed?
Source: I work at Amazon, and have worked on Alexa
They don’t spy on you without your permission. Comments like these devalue actual instances where companies genuinely steal and manipulate data. Take the tin foil hat off…
If you had any remote idea about the tech industry, you’d know what kind of reputation Amazon has. If Amazon were stealing data, you can bet your ass that one of its employees (probably one of the ~6% that gets fired every year) would happily rat them out.
Comments like these amaze me. Even cesspools like Reddit and Twitter wouldn’t be so out of touch and stupid.
They’re not completely wrong, though. If the devices are phoning home when the mic is disabled, then someone would have discovered it by now. There are people who do that shit for fun, and Amazon is a big target.
Source: I work at Amazon, and have worked on Alexa
If you’re high enough level at Amazon to know for sure, you’re also high enough level at Amazon to almost definitely lie to people about it and other things as part of your job.
So your theory as to why you haven’t seen evidence is that there’s a conspiracy of people withholding the evidence. I gotta ask, do you have evidence of that conspiracy?
That doesn’t make any sense. If I were “higher up”, do you think I would be actually doing any IC work? I’d be in management, and probably won’t even know where to look at any of the fucking source code.
Feel free not to take my word for it, but also feel free to ask anyone that has any experience with Alexa, or anyone that has monitored traffic leaving the device.
Is Lemmy just full of conspiracy nuts or something?
That’s not how it works, at all, at ANY tech company. I know, because Amazon has a shared GitFarm, with detailed documentation on how things work, and most importantly the better part of a decade where no one inside or outside of the company has found the device “listening”.
I said it elsewhere, but will repeat since you clearly have no idea about the tech industry. Amazon treats it’s corp employees like shit. If ANYONE was going to leak shit about their employer doing something shitty, it would be an Amazon employee, especially since their URA process is so widely known.
IF Amazon get caught spying, they get everything that they deserve. I’ve never worked in the Ring org, so whatever they do is on them, and if they get caught being shitty with customer data they should be punished severely. What I can say, which (again) is backed by a decade of people not calling out the really-fucking-easily-verified fact that Alexa isn’t phoning home outside of the utterances you say to it. Wakewords don’t leave the device, they’re an offline trigger to get the “actual” content.
I’ll repeat it again, this is an insane take that I haven’t experienced after a decade of posting on Reddit and Twitter. Why is the fediverse full of conspiracy theorists that don’t do basic research before making statements?
lol they are such stereotypical conspiracy theorists too, “of course you’d say it’s not true, that’s exactly what someone who was hiding the truth would say!”
Tell me you’re not a software developer without telling me you’re not a software developer.
If you’re working on the code the only thing that might change is not having access to the release/staging environments (production databases, cloud server, etc.) but you would need access to the code itself (and development database/services), so it wouldn’t be too difficult to check if the code is keeping voice recordings
(italicized is edited in for clarity)
Additionally, the higher up you are, the less code you usually write. With software development being higher up usually means more meetings, team management, planning, and higher level infrastructure talk.
(Obligatory disclaimer that I’m pretty new in software development, this is the experience in the company I work at and seems to be pretty standard among other companies as well)
You should probably edit your comment to clarify that they don’t listen to you.
“Spying” doesn’t really have a clear definition in this context. Amazon employees have been caught spying on customers through their cameras, and giving away clips to authorities without “owners’” consent, consult or notification.
True, that is more accurate. IMO, in those instances, Amazon get all the shit that they deserve…although for many instances these are in their terms of service. There has been no shortage of scandals where Amazon have used utterance data for training ML models, or where they’ve retained voice data for the same reasons, when these have been in the TOS from the beginning.
I see so many attempts to argue with people before the bootlicker insult comes out. At least I know it’s being used correctly and not as some kind of defense against thinking or engaging.
If you have/get a Pixel you don’t even need to be tech-savvy. You literally just plug the phone into your PC, navigate to the Graphene webpage and click “install” right in the browser.
I didn’t say that, I said it’s more comforting. Unfortunately I need my phone for work so I can’t run a de-googled rom so it’s good enough for me. And I never see ads referencing things I talk about.
Honestly if you are thinking about your phone listening to you then you probably should look into running something other than stock. (You are not most people)
It depends. What I would do if you are interested is buy a cheap damaged phone that is still usable and then flash it with Lineage. From there you can break things without causing issues on your main device.
You’re the one that thinks an integrated microphone would just ‘turn off’ because you tapped on a touchscreen so maybe cool it with the schoolyard insults you tossed some elemental chemistry into to make yourself sound smarter.
Yes, but you are the one who started to talk about a software switch. So obviously you were not aware of the fact that there are hardware switches. Or you just wanted to troll.
Pretty sure I was talking about the subject of the post and not, as many have pointed out, all the ways you could possibly wire up a fucking microphone.
You certainly could wire that up and I was aware of that when I made my comment but if you think any of your electronics, especially the big corp home devices like the Google home or Amazon Alexa, are wired like that then you’re out of your mind.
Also, why isn’t there a slide cover to physically cover the camera, and why can’t I turn off the mic and camera separately? So I just use one of those black foam stickers to cover the camera.
So, I don’t trust them to have actually done what I’m going to describe, (and honestly I’ve just accepted that even with everything off, they’re still giving me ads based on stuff I’ve only talked about and never clicked or written anything), but:
The programs that recognize specific phrases(Ok Google), are always separate from normal voice recognition (and muchmuch lighter in terms of processing). So, if they weren’t Google, they might have left the “Ok Google” recognition on, but not process anything else that the mic receives.
I will say that that’s exactly how the google voice api works. Of course it’s all in a black box, but that’s how the documentation describes it and how it functions when making a voice app
Didn’t they just pass a law to make all that illegal spying legal, like that changes anything? Seems obvious if your phone is listening in a device like this will be used no matter what setting you use. I remember Amazon being caught leaving their mics on and also Facebook sending conversations to 3rd parties for transcribing. And this is just a small fraction of the shit we know about.
Not necessarily you or your case, but I’m still convinced that a lot of people just have confirmation bias (only noticing it when it happens and discounting the thousands of otherwise innocent ads). There’s also subconscious ad effects, like you were only talking about it to begin with because your saw it somewhere because it’s been spreading by weird of mouth from people who initially saw an ad
It’s not just that either. Google knows who your family is. They know who lives with you because of location data. So any time those people search for anything regardless of whether they’re on your home network, they likely serve ads to whole families at a time when one person searches for something.
This has been my theory as well, Google presumably knows when I meet up with a friend for lunch (I don’t know if they go to such lengths but they certainly have access to the data to figure it out), if my friend then starts searching for something related to our conversation afterwards, Google could serve me ads about it too, just inferring the topic of our conversation based on that
Doesn’t really explain why I was receiving cat litter ads after only speaking with my husband offhand about maybe getting a cat. We didn’t already have a cat, so hadn’t had any reason to look up any cat care goods ever, and I had never searched for anything even remotely cat-related up to that point. But wouldn’t you know it, about 45 minutes later, I was getting kitty litter ads. Very spooky.
Sorry but I want the true story to be that your husband immediately went off and started googling to find a cat to surprise you for Christmas thus you got cat ads (same network like someone else said).
Lmao, I wish but no, no hallmark movie plots here. This was a few years ago, and we now have said cat :) He definitely forgot immediately after I mentioned it until I showed up with a cat one day lol
As Pole’s computers crawled through the data, he was able to identify about 25 products that, when analyzed together, allowed him to assign each shopper a “pregnancy prediction” score. More important, he could also estimate her due date to within a small window, so Target could send coupons timed to very specific stages of her pregnancy.
One Target employee I spoke to provided a hypothetical example. Take a fictional Target shopper named Jenny Ward, who is 23, lives in Atlanta and in March bought cocoa-butter lotion, a purse large enough to double as a diaper bag, zinc and magnesium supplements and a bright blue rug. There’s, say, an 87 percent chance that she’s pregnant and that her delivery date is sometime in late August.
That’s the gist of how it likely works; the wake word is detected by an “always on” audio DSP, but a software mode prevents the passing of microphone data back up to the SoC. I’m actually quite familiar with Amazon Echo engineering design, and they implement the “mute” feature in a manner that takes privacy seriously: the LED indicator on that button is hardwired to only turn on when the microphone is literally powered off. Thus, an Echo device can’t even manage such a cheeky response, nor can a software bug or hack enable listening while the mute button is lit.
Add comment