Voice assistants are always listening. So why won't they call police if they hear a crime? If you saw someone being assaulted, you'd probably whip out your phone and dial for help. But when one of our newly ubiquitous devices hears a crime, it does nothing. If Alexa or Google Assistant or Siri hears an assault, or a rape, they sit there waiting for their cues to act. But that's a load of malarkey. Listening for a …
What do you suggest they do?
The only thing I can think of is that when they hear the cues that someone needs the emergency services is that they activate but then who is going to be on the other end? You would need it to first send a location then connect and share audio with a person who could determine if help was needed or if it's a mistake. That's a lot of people and you couldn't just use the 999 people because there aren't enough as it is and they need to be dealing with confirmed emergencies. There's also the privacy aspect, where do you draw the line, do you really think the government won't see this as an opportunity to spy on people? Mention drugs, anything terrorist related, something against the government and it gets grabbed.
It's shouldn't happen in my opinion regardless of the perceived benefits.
"I don't think I dare watch another episode of CSI! Heaven forbid I watch a film about terrorism or wife beating."
Quite. This whole article is a load of touchy feely guff by someone who clearly didn't even spend 2 minutes considering the dangers of false positives. But then this sort of quality of "journalism" is where El Reg appears to be heading these days.
100% agree, there is something fishy in the cupboard and isn't an old tin of cat food.
I've just been listening to the news and how all the trains in the East of England are being shut down or limited, with no bus replacements, because there is snow *on the way*!!
The lunatics have taken over the asylum and sanity no longer means what we used to think it did. Being sane in this country now feels dangerous. I think it's time to bug out.
PS It's now a moot point, but the benefits of these types of systems are not for the *consumer*, so why is anyone surprised when they are not geared up to provide things people actually need?* Plus, where is the, very obvious, solution of NOT HAVING THE DAMNED THINGS!!? It isn't like they are providing an essential service, like electricity or water, is it?
*Although as has been mentioned, I'm not entirely sure how a monitoring system would cope with TV shows, or the actors guild rehearsals of Macbeth, for example.
Being sane in this country now feels dangerous. I think it's time to bug out.
Ditto here. For $DEITY'S sake, don't look for sanity in the U.S. (where I live - not that you're likely to look here, but just making sure), Where do you suggest we go to find this sanity you speak of?
Don't bother coming to Australia looking for sanity. It fell down the back of the couch years ago and hasn't been seen since. The drive to back door crypto is on, house prices are mental, business is actively trying to worsen working conditions for wage slaves and the politicians are too busy infighting to even pretend that they're running the place. You'll have to look elsewhere.
Well (on Macbeth) - it would detect the guilt - and if any sane/unburdened by the said self-inflicted emotion - it would probably decide to do what you just described - check out.
Do correct me if wrong - but once the asylum decides its collective guilt must be suffered out - there is not much one can do but wait for it to take place - any help only makes the ones who know (or consider) they have done wrong - just accumulate the need for self-punishment (and they know how they want it too - not too soft, not too hard either - no permanent damage, just cleansing of the soul - I dare say (as a complete non-Briton and so to be booed off the stage for exposing the mechanism - greetings from a place I won't name)).
I have denied the Google App permission to access the microphone on my phone. Saying 'Okay Google' does nothing. I NEVER tap the Google search bar either, I use the Startpage shortcut instead. On my previous phone I managed to disable the Google App completely. It required a hard reset when I tried it on this iteration.
I run Oversight on this laptop, our TV is decidedly non smart and the digibox is not a voice activated one. I'm not interested in that. As a scientist I am adept at forming specific search queries and finding stuff myself.
Yes, it's easy for us techies who have a bit of deserved paranoia. But what about the average "Joe User" who hasn't clue or isn't one of us? They have no idea what's being slurped and most don't care. If they did, FB wouldn't have ever been a "thing". And let's not even get into "smart" devices that send everything back to the mothership.
This post has been deleted by its author
These devices, or the cloud services that power them, can easily understand when someone is angry, or terrified or in pain.
And can they tell when that someone is physically present in the room, and when they are being relayed through the loud, high-quality home cinema setup in the room?
How is Siri going to make the difference between a child being assaulted by a pedophile and one who stubbed his/her toe and is expressing justifiable but innocent pain ?
Siri won't.
So then the argument will become "put a 911 operator on the line to listen and decide", and then we have two issues : violation of a person's property by forcing it to dial 911 without consent (plus all the people in the vicinity - great use of resources there, no possibility of confusion at all) and shoving the responsibility to a harassed phone operator. Sounds like kicking into the sidelines to me.
This whole article is a brilliant demonstration of the stupidity of the "think of the children" mentality. Oh, so phones can do this now ? Well enroll them into the police force and have them tell on everything all the time, because that is the only thing that can protect the children.
No thought of any consequences, or even of the practical issues that can be encountered.
In short, useless and dangerous.
and how it works. These things have the potential to be used buy anyone who can hack the system or gain access via legal means to snoop on you and what you are doing 24/7.
I disable the assistant in my phone and check that it remains disabled after every update (rare these days)
Having been around tech and working on it for more than 40 years NONE of the bits of kit that runs Alexa/Siri/Bixby/or whatever will ever come into my home let alone be plugged in and switched on.
I've done without them for 60 years and can carry on doing without them for a few more decades thank you very much.
IMHO, the sooner these die a slow lingering death the better but people love to show off their blingy toys. 'Alexa, tell Bezos to get F*****d'.
The trend is definitely to move more of the voice recognition into the device itself - at the moment there is a lot of cost tied up in cloud-side systems that isn't covered by the low-margin bug speaker. That could be used to send less data to the cloud or to process and retain more data locally and send it to the cloud on demand. It will all depends on which is more profitable...
Alexa does the same. I thought the linked article might explain that they have secretly been 'listening to everything', but alas not.
The device doesn't send conversations back until it hears it's trigger word. The linked article was just explaining that the developers could get the exact words used to trigger their skill (rather than them having to set the trigger words and just get told that theat pattern had been activated).
So regardless for the fact that there doesn't appear to be any evidence for the fact that it is forever sending every conversation back to the mothership (sure, they could, sure they or the government might want it) the amount of false alarms to a busy service would be ridiculous.
A far better solution would be to be able to record a hardcoded 'code word' into the device that would trigger the emergency service call and tell the operator to listen in with the current location. Just make sure the code word is not something that might be said during 'normal times'?
'I don’t know what Alexa does, but Siri waits for the words “hey Siri”, hard coded in a very low power chip, and doesn’t listen to anything until it gets that signal.'
Which would fail the other way, "HEY SIRI, HELP! I'M BEING MURD...'' gurgle, thump. Hmm, thinks Siri, they are mixing their English and French?
No thanks. It is a dumb an idea as a car that does an emergency stop every time it hears the word "brake" (or was it "break"?). If they have to do something like this (and Alexa at least could do it now), it should be based on a safeword (or phrase), not AI trying to disentangle someone doing DIY while watching a horror movie.
I threw my latest voice assistant in the bin after it blatantly took no emergency action whatsoever upon my extremely audible reaction to the news that the local Krappy Fried Chicken shop was going to be closed for a whole week!, yes, seven whole days! It just bloomin sat there and did absolutely nothing, I was anticipating maybe an automated emergency call for a supply of spicy fried wings to be air-lifted to my address, but no, nothing. I had to put the call in myself, what's the point of these devices. Waste of money!
... you'd surely not want them to call the emergency services. After all the vast majority of calls would be "butt-calls".
Besides the majority of people still believe that they don't send everything to be processed to some manufacturers cloud.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
