Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Statistical Analysis shows Echos process voice to serve ads

arxiv.org

Post reply on HN
Print view View original

41–50 of 125 posts

Re: Statistical Analysis shows Echos process voice to serve ads

#41

Post by mort96 »

mukesh610 wrote: post #20
mort96 wrote: post #13

Earlier quoted context omitted.

> I don't know why people are so freaked out by living in a surveillance system where every corporation tracks their every move for profit You haven't read enough dystopic fiction.

Fiction isn't exactly a good argument

A lot of fiction isn't just about "look at this whacky world I invented". Especially dystopic fiction usually tries to tell us something about our own world and its potential future.

But the main point with my comment was to re-frame what was advocated, from their fantasy of "one small store owner doing the best for their customers" into "a surveillance system where every corporation tracks your every move for profit".

Re: Statistical Analysis shows Echos process voice to serve ads

#42

Post by dr_dshiv »

mtlmtlmtlmtl wrote: post #16
dr_dshiv wrote: post #9

Earlier quoted context omitted.

Why? If I run a store and I learn about my customer’s interests, why shouldn’t I use that to make better offers? I don’t know why people are so freaked out by data collection. I just can’t believe that for all the data I share, I still get such shitty ads.

This a strawman. We're not talking about your friendly neighbourhood corner store here. We're talking about large scale systems with billions of users, systems that contain and shape a lot of our public discourse. The way these platforms make money inherently affects our society and our democracy. "Helping your customer" is irrelevant here.

I thought this was a topic about Alexa (a voice UI) using your voice (actually, installed skills) to deliver more targeted ads. I don’t see the public discourse connection here?

Re: Statistical Analysis shows Echos process voice to serve ads

#43

Post by LeifCarrotson »

abricq wrote: post #34
wildmanx wrote: post #26

Earlier quoted context omitted.

Given the power demands of such an always-on analysis engine, that would show in battery runtime. At least if it's happening while the device is not plugged in.

On iOS, the word "Siri" can be detected at any time (if set so). Maybe ( just an hypothesis ) there could be a set of other keywords also listened for, that once detected could start another more complex routine. Like that, you could limit the battery impact, and yet be able to listen to the users for advertising purpose.

On my Motorola from 8 years ago, the voice command was "OK, Google" or any custom short phrase you trained for it. But I think it would be really hard to get a useful but concise list of keywords... It would certainly be possible to listen for, say, "new Subaru" and show an ad if that one thing was detected, but the point is that you want to select which one of a million ads to show you'd need far more keywords, which gets computationally expensive quickly. It would probably be more battery efficient to compress audio that sounds like speech aggressively and then send it up to The Cloud for analysis...scary stuff, and part of the reason I don't have any home assistants!

Re: Statistical Analysis shows Echos process voice to serve ads

#44

Post by dr_dshiv »

mellavora wrote: post #29
CaptainZapp wrote: post #15

Earlier quoted context omitted.

Ok, you own a store. A customer enters your store and buys a widget. Since you're smart and want to know everything you record the sale and add it to your database. Only that's not where it stops. After the customer leaves the store one of you minions follows him to every other store he ever enters to record what he buys, or even looks at That's why your analogy falls completely flat. Nobody would complain if it only…

Worse! You report this information to government authorities, who monitor this and may decide at a future time that the things you were shopping for today mark you as a bad or suspicious person. Worse! You also share this information with credit bureaus, meaning the data is available to potential employers. Thus limiting your job options. Looked at a hookah once because you were curious? Sorry, no pot-heads in my com…

I don’t see the connection of this to the topic. Not trying to be obtuse. Are you saying it’s a slippery slope between targeted advertisements and 1984?

In my store analogy, I’m trying to make discourse on this topic less black and white. To recognize that data collection is not some inherent evil. And sure, to point out that a store that tries to peep in your bedroom window should be called to task.

Re: Statistical Analysis shows Echos process voice to serve ads

#45

Post by nairboon »

The full title is "Your Echos are Heard: Tracking, Profiling, and Ad Targeting in the Amazon Smart Speaker Ecosystem"

It's about the interactions with Echos, and not as some other comments imply, about listening to general conversations passively. Although, the experiment setup might be used to study this in a future experiment.

Re: Statistical Analysis shows Echos process voice to serve ads

#46

Post by Kye »

greazy wrote: post #36
osrec wrote: post #28

Earlier quoted context omitted.

I have mentioned this a number of times here on HN, but each time I'm told by other users that it just appears that way because the topic is fresh in my mind... I don't think that's the case.

This is called observational bias. Imo the truth is much scarier: advertises know us better than we known ourselves. There's also the idea that we are so alike, advertises can generalize to great success.

This is especially blatant when you don't fit into any category advertisers understand. Most of them think I'm either a hip urban youth or a well-off exurban housewife. I don't think they know what to do with someone who has associations, friendships, and interests that cross countless geographic and socioeconomic boundaries.

Re: Statistical Analysis shows Echos process voice to serve ads

#47

Post by technothrasher »

greazy wrote: post #36
osrec wrote: post #28

Earlier quoted context omitted.

I have mentioned this a number of times here on HN, but each time I'm told by other users that it just appears that way because the topic is fresh in my mind... I don't think that's the case.

This is called observational bias. Imo the truth is much scarier: advertises know us better than we known ourselves. There's also the idea that we are so alike, advertises can generalize to great success.

> This is called observational bias.

I think you mean confirmation bias. Observer bias is similar, but relates to being non-blinded during a scientific study.

Re: Statistical Analysis shows Echos process voice to serve ads

#48

Post by a3w »

technothrasher wrote: post #47
greazy wrote: post #36

Earlier quoted context omitted.

This is called observational bias. Imo the truth is much scarier: advertises know us better than we known ourselves. There's also the idea that we are so alike, advertises can generalize to great success.

> This is called observational bias. I think you mean confirmation bias. Observer bias is similar, but relates to being non-blinded during a scientific study.

Confirmation bias means that you want to look for examples of proofing something, instead of seeking counterexamples. That's not it, either.

Baader-Meinhof phenomenon or Frequency illusion could be what we were seeking?

Re: Statistical Analysis shows Echos process voice to serve ads

#49

Post by snapcaster »

dr_dshiv wrote: post #44
mellavora wrote: post #29

Earlier quoted context omitted.

Worse! You report this information to government authorities, who monitor this and may decide at a future time that the things you were shopping for today mark you as a bad or suspicious person. Worse! You also share this information with credit bureaus, meaning the data is available to potential employers. Thus limiting your job options. Looked at a hookah once because you were curious? Sorry, no pot-heads in my com…

I don’t see the connection of this to the topic. Not trying to be obtuse. Are you saying it’s a slippery slope between targeted advertisements and 1984? In my store analogy, I’m trying to make discourse on this topic less black and white. To recognize that data collection is not some inherent evil. And sure, to point out that a store that tries to peep in your bedroom window should be called to task.

But your store analogy is stacking the deck in favor of your point. As the people responding to you pointed out, it falls apart when you consider all the larger implications of pervasive data collection. Would you defend the Stasi with the same language? I dont know why you're _trying_ to make this point and gaslight the people who are rightfully very concerned about the concentration of power this data collection results in

Re: Statistical Analysis shows Echos process voice to serve ads

#50

Post by fddhjjj »

yunohn wrote: post #40
doubtfuluser wrote: post #31

I have the feeling, that the paper is flawed and missing an important experiment. All their results seem to rely on skills being used. There is indeed a need then for Amazon to prevent user tracking in skills (like for example Apple does and requires consent by the user). But to come to the conclusion that Amazon shares the data with advertisers I would have expected an experiment with eliminating skills as a reason…

Additionally, they are trying to imply the common trope of “voice assistants are listening to everything we say all day for ads”, whereas their test methodology was to actively use the top skills for those interests and perform actions. While I don’t like the sharing of such data for ads, it’s a far cry from Alexa processing voice in the background with zero interaction.

What part of the paper gives you the impression they imply voice assistants are listening to everything? I don’t get that.

The discussion in the paper is nuanced on that point and does not make that claim as far as I read it. Section 2.2 (page 2):

> The content of users’ speech can reveal sensitive information (e.g., private conversations) and the voice signals can be processed to infer potentially sensitive information about the user (e.g., age, gender, health [82]). Amazon aims to limit some of these privacy issues through its platform design choices [4]. Specifically, to avoid snooping on sensitive conversations, *voice input is only recorded when a user utters the wake word*, e.g., Alexa. Further, only processed transcriptions of voice input (not the audio data) is shared with third party skills, instead of the raw audio [32]. However, despite these design choices, prior research has also shown that smart speakers often misactivate and unintentionally record con- versations [59]. In fact, there have been several real-world instances where smart speakers recorded user conversations, without users ever uttering the wake word [63].

Post reply on HN