That's assuming they're listening/watching all the time. Google Assistant (and all the other assistants) wake up on a certain magic word; who's to say that other apps/companies aren't doing the same? Well, probably your battery life might be terrible and your phone running unusually hot, but Assistant and co get away with minimal battery usage...

Quite often while talking about a restaurant, opening google maps will have it highlighted. I think this shows that not only does google not require a magic word to start recording, it actively processes the input it gets. I'm highly considering switching to lineage os.

Try making a controlled experiment. Humans are notoriously good at seeing patterns that aren't there.

Someone conducted a simple experiment to see if the facebook app was listening to their conversations. Spoiler: it's listening to your every word.

https://www.youtube.com/watch?v=U0SOxb_Lfps

Yeah, that's not a controlled experiment. I can think of at least three benign explanations that is consistent with not listening.

The experiment, the real, controlled (replicable!) one, isn't hard to do. You just need a couple of spare phones.

The patterns are there. But often the connection is meaningful, not causal.

https://en.m.wikipedia.org/wiki/Synchronicity

The pattern in this case is the connection between surreptitious listening and the suggestion.

The brain is still seeing a pattern though it is one of correlated meanings/abstractions, not concrete causes.

Dont be silly. Companies and institutions can just pay or coerce the data operator for transmission, o do it when your on wifi.

In Sao Paulo Brazil, the mexican-based operator Claro offers free bandwidth for facebook and whatsapp access.

I imagine it would be possible to process media locally into actionable and highly compressed data. For instance use the google text-to-speech toolset on media while the phone is on sleep and even run it though a sentiment analyzer and wordlist. eye tracking and activity correlates are even more compressible.

Earlier in the year the police were monitoring activity on public wifi hotspots and using the info to target people labelled as undesirable in the quickly gentrifying city center.

Since I am a a)foreigner that b)transverses a diverse array of classes, groups, and political affiliations on a daily basis, i go ahead and always assume that i'm being constantly monitored.

Phones are so full of back-doors at every level that you wouln't notice if it's sending information. This is why we can't have open phone firmware. I don't think Google is secretly transmitting data, but it wouldn't surprise me if they do. Only way to stop it is to make it illegal to collect any kind of data without user consent. If you read the user agreement it probably already say what information is collected.

They don’t have to send back all of the raw data, just analyses like ‘person X’s face detected, is in a noisy place’ or ‘conversation included these 5 keywords’. The processing and data transfer could also be deferred until the phone is idle or on WiFi.

It's just as farfetched to think they're doing constant image/voice recognition with your phone's processor without you knowing. Those models are big and take a lot of computing power

It doesn’t need to be constant or continual. I don’t talk continually near my phone, for one.

Siri and other voice rec is able to get words from my audio without appearing to affect my phone function or slow it down. There are plenty of times my phone is idle that it could be doing that.

Same with photos. When you give many apps access to your photos, as we know, they will analyze and/or transit the metadata of all photos on the device. It’s not hard to imagine going farther than that, and as noted, processing or data transfer could be deferred.

Siri listens for a hotword, which is tremendously easier than running general speech detection models all the time.

Seems like you have to do at least minimal analysis of all sounds to be listening for a hot word.

Exactly, that's why "OK Google" isn't a thing.

There's an enormous difference in model complexity between key phrase detection and (general) automatic speech recognition + natural language processing. There's a reason nearly everything Google Assistant does except "OK Google" is done on Google's hardware.

> They don’t have to send back all of the raw data, just analyses like ‘person X’s face detected, is in a noisy place’ or ‘conversation included these 5 keywords’.

Not without making a noticeable drain on the user's battery.

When you fire up an app like Instagram, it does many things that create various drains on the battery. It opens the camera repeatedly when I don’t intend to use it, for instance. When you click stories, it opens the camera, often facing you. Who’s to say it doesn’t send that pic or some analysis of it to FB? I don’t think that would use any significantly higher battery or network anything especially since the processing and data transfer could be deferred.

> the data costs associated with the spying would be so great

I've just checked my monthly Discord traffic and it's mere 251MiB in/73MiB out. For roughly an 1-2 hours of gaming every other evening, and some longer weekend sessions.

Power usage, on the other hand, is a different story. When I've experimented with an app that'd continuously record everything around me, battery ran out in just a few hours.