Yes, Amazon Echo Is an Ad Machine
gizmodo.com
gizmodo.com
https://jasperproject.github.io/
Lots of related projects could also use your help. You don't need to be an AI genius - every FOSS project can benefit from help with debugging, cleaning up code and writing documentation. Mozilla Open Voice just want you to read a few sentences to help build their training corpus.
By high quality I mean something as good as Tacotron 2 [1]. However, I have no illusions of Google ever releasing code or trained models.
Someone will likely re-implement it from the paper, but it's unlikely to be anywhere near as good as those samples, due to worse tuning and lack of a good training set[2].
This effect can be seen if you compare Baidu's samples [3] for Deep Voice 3 with those from an implementation by a third party [4].
[1]: https://google.github.io/tacotron/publications/tacotron2/ind...
[2]: Common voice, for example, will not work well, since you want lots of data from a single speaker for good results. See for example, how samples get progressively worse when increasing the number of speakers in training data[3].
[3]: http://research.baidu.com/deep-voice-3-2000-speaker-neural-t...
Tacotron 2 was trained on 24 hours of single-speaker transcribed audio, which is comparable to the freely-available LJ Speech Dataset. We know that it's feasible to train using unaligned transcribed speech, which broadens the opportunities to reuse an existing corpus. (Aside: does anyone know the legal status of training a DL model on copyrighted content?) Tacotron 2 was trained on a 32 GPU cluster, which is large but not absurdly so; we are seeing drastic performance increases in low-precision compute, which should hopefully start to trickle down as Volta reaches the market.
Expertise is a bigger challenge at this stage, although that is progressively changing. A huge number of developers are taking a serious interest in deep learning, so hopefully we'll start to see more active contributors to DL FOSS projects. The teams at DeepMind and Baidu Research are clearly highly skilled but relatively small, which suggests that their efforts could be replicated by a small but determined team of FOSS developers.
How would anyone know that it happened?
Seems like it would fit under Fair Use in USA but wouldn't be allowed under UK's Fair Dealing; I can't comment on other jurisdictions.
TVs are the same. And phones. And most web sites. And buses, walls, taxies, public transportation stops, bridges over roads etc etc.
One has to work hard to defend against the ad onslaught these days.
To me it represents pure evil and I don't see how you could defend against it. Except for not buying into this stupid shit.
Cortana or Siri exist without obvious nefarious purpose. Alex and Google are clearly trying to push advertising. I know which I would prefer to set timers and play music for me.
Now imagine the data they're collecting from their suite of amazon devices, where you can't control the data-mining.
That could be look at what $politicalPartyAffiliates in your area consider important enough to change allegiance, so that a party that's marginal can present themselves as good for $socialGrouping in order to sway the vote. It's not necessarily just about purchasing.
It also allows analysis of other aspects of life, like, say, people who search for drug paraphenalia -- or a step further away, music that's associated [by someone] with people who take drugs -- which festivals are they going to this Summer ...
That targetted selling may have no direct effect through the Echo itself, but the potential is there to increase the influence on you in other channels.
There may be other info, we have 3 kids who use our FireTV: I'd imagine they can estimate the child's age & sex and give targetted advertising in emails/web pages to companies selling toys for those ages. Based on the apps (whatever they call them, skills?), they can probably target that much better.
It's not nefarious, but what happens if you ask a question like "open a website where I can buy music" or "find me a powerful new computer", or whatever. Is there no bias in Cortana/Siri towards MS/Apple? How about if you ask "is Siri or Cortana a better voice assistant?" ...
The fact that you interact with it each morning is more powerful than any print, broadcast, or web ad Amazon can buy. They're gotten into your brain. Even if you never see the device (which is now a visual billboard in your home), when you think about it, you're thinking about Amazon. You've already been commoditized and you don't even realize it.
Do you really not see the inherent problems with these devices, or is the remote start just that nice? =D
For example, by far the most common argument against these devices is that they could be spying on you by listening to your conversations 24/7. This is true, they could be. However, it's also true of cell phones, which 99% of the complainers have had near them 24/7 for the past 10 or 20 years. If you think the Echo is a terrible device because it could spy on you but you keep a cell phone in your pocket all day, you're trying to justify a negative emotional response to the Echo, not rationally analyzing the situation.
The comment I replied to above, by reaperducer, attempts to argue that the Echo is implicitly commoditizing me every day because I interact with it. This may be true, but does that also mean I'm being commoditized by Whirlpool and Moen and Philips and Tesla and Trader Joe's? If yes, then what does that even mean, and how am I meant to avoid it?
You then jump in with a vague hypothetical about a future update making the device start trying to sell me socks. I'm pretty sure Amazon isn't going to start putting ads in with my music or weather reports. If they do, I'll stop using the device. It only cost me $30. I've spent far more on devices that now sit on a shelf unused. That's just the nature of consumer electronics.
If you want to convince me that these devices are inherently bad or have massive inherent problems, it's really simple. All you have to do is tell me what problem they cause me that could be eliminated by getting rid of the device. So far, everyone who's attempted to do this has either stated a problem that wouldn't be eliminated (e.g. the spying potential or the daily interaction getting the brand into my brain) or a problem I don't suffer (e.g. getting ads).
I have an Echo and it is helping Amazon get better at selling me stuff. When I try to list the probable or even possible harms from this, I can't come up with a lot.
Compare that to something like driving my car. 100 people every day are killed in their car. Even so, I still drive because the expected utility of getting to point B is much greater than the expected harm from doing so.
I use a car and the worst case scenario is death or permanent disability, right? People die in their cars every day.
So what's the worst case scenario for using an Echo? How many people a day does that happen to?
What's the worst case scenario from having an Echo? How many people does it affect each year? What's the dollar cost of that damage?
IMO, it isn't an imminent threat. It's not going to kill you. It's not going to cause you any physical harm. It's the slow erosion of privacy which is a detriment to everyone.
I'm going to avoid the cliche Orwellian arguments and the normalizing surveillance arguments. Both of these have been addressed ad nauseaum. I'm going to go a different route and see if it resonates with you at all.
It is incredibly easy for humans to be manipulated. This only gets easier as we have treasure troves of data about individuals. Amazon has many of the smartest people in the world working to manipulate you to do things in their best interest. Sometimes their interests lineup with your best interests and things are great. Other times, their interests will be negative to your life.
It is in their best interest for you to not be content with what you have. It probably is your best interest to be content with what you have. It's not wrong to want things, but things will never fulfill you.
I'm not saying capitalism is evil, Amazon is bad, or that we aren't living in one of the best times in human history. I am saying that putting a marketing device (which is also a privacy concern) in my home is not something I am interested in. Especially when I can do the 90% with my phone in 5 thumb presses.
I like the driving analogy because even though the worst case scenario is about as bad as it gets, it's not all that likely. I think people, even here, have a hard time separating possibility and probability.
A connected speaker has real and immediate benefits for some people. People that use it to buy stuff (I do all the time) aren't going to see ads for associated products like website shoppers do. Frankly, if the Echo scares you, https://amazon.com should scare you even more. I bet they collect a lot more data in a 10 minute browsing session than they do with months of Echo use.
This is somewhat correct but misleading. It's always listening for the "wake word" -- alexa or echo -- and then it processes your request. But it's not sending all conversations up to amazon. Just your alexa requests.
However, this is a device that requires internet to function and is managed by them giving you zero control beyond unplugging it.
As a note, for non-rooted users, it's common for Android vendors to push mandatory updates. Both my last Android phone and tablet before I left the platform were pending 'mandatory' updates where I had to kill a process after booting to prevent them from executing.
This person gave a very good explanation here:
https://np.reddit.com/r/Showerthoughts/comments/7m91u9/if_go...
I have a bunch, and you only really use them for about 5 things, with the most useful thing being music.
Once was when an actual person was speaking and another time was when the TV show 'The Sopranos' was playing in the room. In the latter case, I rewound the show several times and it triggered on the same phrase multiple times (and the character was not saying anything that sounded like Alexa, at least to me)
Think of the children!
It should however come with a severe warning about what it is doing in a way that my mother would understand. Otherwise she would just see it as a nice thing to tell her what to do.
So would there be a major difference in consumer behavior if it was generally known that the advice given by the echo is paid advertisement?
If Yes -> then governments should take notice.
If No -> then interference would just inhibit the market.
At one time nobody was forcing us to pay private health insurance conglomerates every month, yet here we are.
All it takes is some clever lobbyists and some cash, both of which Amazon has in spades.
Wow. Do you really believe that it's government job to guarantee ad-free Echo?
I mean you would agree that a bunch of people could organize a boycott of Echo with ads and eventually Amazon would stop selling it and sell Echo without ads and that would be fine right? So why don't we just do that? It's because we lack a coordination mechanism, the incentive of individual boycotters is to cheat. Luckily humans have developed a great coordination mechanism called 'democracy' and we can use it when coordination in the market place has failed.
I could agree that companies be required to have a disclaimer when they provide preference to a certain product set over another.
Sure, that's one possibility. Probably not a very realistic one but it's there.
>I don't see why you would immediately go towards having the government step in and regulate whether Amazon can make the Echo prefer their own products quite yet.
Why should we immediately go towards another company/market-oriented solution? What is it about government that requires we wait around for the market to fail? Why not just go do what we want?
We can. We do. Most people don't care about beeing spied on, so they buy it. So they also would not support your democracy introduced market improvement.
Besides, a market can't regulate itself, if there is allways a government regulation in case something is not nice.
And I do actually believe, people would buy the adfree, non spying device, if they would have a choice
So that's simply a market opportunity.
I find that sort of thinking pretty scary. I'm incredibly hesitant to expand the purview of the government into whether products should succeed or fail. It sounds like what you're saying is that people are too stupid to make decisions for themselves, so we should enact government regulations to think for them. This is something markets are very good at, as others have stated. Expanding the government's role in things should be an absolute last resort, to me.
If people are willing to trade their personal information for cheap, cool gadgets, then who is the government to step in to say "no, you're not allowed to do that"? By the same token, you could flip it around: could the government step in and say "you don't want this gadget that's always listening to you, but you have to have it because some group you don't agree with tells you to"?
I want to be careful not to go too slippery slope, but should the government step in and shut down Google, because they exist to serve ads? What about Facebook? It's the same thing there: people trade their personal information for perceived value in other places (Facebook: keeping in touch with people, Google as access to information). Should the government be the ones to decide who should be allowed to make decisions about their information?
It's scary because once the government gets to decide what people should and shouldn't be allowed to use, or what products should or shouldn't succeed, it can be used as an incredibly powerful tool. To me, a useful way to think of government is as a very sharp knife: it can be extremely useful in the right situations, but you can really hurt yourself if you're not careful.
Third question: should you be allowed to trade my personal information for cheap cool gadgets?
In other words, don't give your number, to people you don't trust.
In the prisoner's dilemma are the prisoners stupid? No, they're actually defined as rational. What makes them come to their collectively poor outcome is that they have no way to act collectively, they require coordination. Those prisoners could be smarter than you and me, they're still trapped by the rules of the system they exist in.
>If people are willing to trade their personal information for cheap, cool gadgets, then who is the government to step in to say "no, you're not allowed to do that"? By the same token, you could flip it around: could the government step in and say "you don't want this gadget that's always listening to you, but you have to have it because some group you don't agree with tells you to"?
Our own government has some problems with this but since we're being a bit high-minded here, let's ask ourselves what a democratic government ideally is. It's just a way for individuals to coordinate with each other.
Prisoner's Dilemma again. By market mechanisms, we're going to jail for two years. We may meet some strange definition of freedom but if we're both rational then we know we will never realize the option of serving lesser sentences. To me though that isn't real freedom so we can see then that a coordination mechanism isn't here to curtail freedom, it's here to grant it (it lets us both choose to remain silent.)
I believe the same things about markets/government. Why should we purposefully handicap ourselves to the individual realm stuck with Echo with ads when we can go to the collective realm to grant ourselves a better outcome.
>I want to be careful not to go too slippery slope, but should the government step in and shut down Google, because they exist to serve ads? What about Facebook?
If the people collectively decide that then yes.
>It's scary because once the government gets to decide what people should and shouldn't be allowed to use, or what products should or shouldn't succeed, it can be used as an incredibly powerful tool.
Well here let me rework this:
It's scary because once the people get to decide what people should and shouldn't be allowed to use, or what products should or shouldn't succeed, it can be used as an incredibly powerful tool.
You're right that is powerful, we should have that power.
>To me, a useful way to think of government is as a very sharp knife: it can be extremely useful in the right situations, but you can really hurt yourself if you're not careful.
Let me rework this also:
To me, a useful way to think of markets is as a very sharp knife: it can be extremely useful in the right situations, but you can really hurt yourself if you're not careful.
> I believe the same things about markets/government. Why should we purposefully handicap ourselves to the individual realm stuck with Echo with ads when we can go to the collective realm to grant ourselves a better outcome.
This is something I think we disagree on: I don't see this as a handicap at all. I also disagree that the better outcome is to grow the government's role in something like this. To me, this is a case where the so-called "invisible hand of the market" is the way to go: ensure that consumers have access to the information they need, and let consumers make the best decisions for themselves.
> If the people collectively decide that then yes.
I think we have to be very careful about how this is used; people collectively decide silly things all the time ("Pink Slime" comes to mind, for some reason), and we have to be very careful about how we encode things into law. Once the government is responsible for something, it's hard to walk back on that. Leaving things to the market leaves things in a more fluid state, in my opinion.
> To me, a useful way to think of markets is as a very sharp knife: it can be extremely useful in the right situations, but you can really hurt yourself if you're not careful.
This may be true to some degree, but it aligns pretty well with people's changing opinions and wants/needs. If we're careful about some things (e.g. monopolies, access to information, etc -- things that allow markets to work efficiently), we can avoid having to bring out the sledgehammer when it's not needed. Some of the early 20th century government involvements led to humanitarian crises (the Communist revolutions in China and Russia); that's not to say that governments necessarily lead to the worst situations, just that sometimes markets are better at things than government involvement. This is a case where I believe this to be especially true.
But that's the point of the prisoner's dilemma. Individuals who are educated, can make informed decisions, have access to information about the results of their decision will always make a decision that isn't the best decision they can make. What the prisoner's dilemma reveals is that the only way to beat the game is through a coordination mechanism. Nothing else will ever help.
>This is something I think we disagree on: I don't see this as a handicap at all.
Hmm so in the prisoner's dilemma you don't see the prisoners without coordination as handicapped? To me it's obvious. Prisoners with coordination realize the best overall outcome, prisoner's without coordination are never able to realize that outcome. Seems straightforwardly handicapped to me.
>To me, this is a case where the so-called "invisible hand of the market" is the way to go: ensure that consumers have access to the information they need, and let consumers make the best decisions for themselves.
I'm providing you with examples where the invisible hand doesn't work the way we like to pretend. The hand is guided towards overall societal benefit by individually selfish actions. What I'm talking about are cases where individually selfish actions lead the invisible hand to overall societal detriment. And that in those instances we need government as a way to coordinate together so that we may reposition the invisible hand where we desire.
>Leaving things to the market leaves things in a more fluid state, in my opinion.
Fluid isn't necessarily good however. We will persist 'fluidly' with Echo with ads until we get the government to act and then we will exist 'statically' with Echo without ads. I suppose I'm fine with that static-ness.
>Some of the early 20th century government involvements led to humanitarian crises (the Communist revolutions in China and Russia);
You're being a bit funny with 'government involvement' here. Maybe there should be some corollary to Godwin's Law where the first one to compare a government program to Mao loses. Anyhow, I'm not talking about seizing the means of production, I'm talking about regulating existing markets.
>that's not to say that governments necessarily lead to the worst situations, just that sometimes markets are better at things than government involvement.
There's also plenty of instances where government involvement is better than markets. That's the case I'm making - essentially that we need government regulation as a way of reigning back the invisible hand when it goes awry aka smart homes that secretly advertise to you.
You keep using the word "coordination", and I guess I just don't see it applying in the way I think that you mean. There are other, perhaps better, ways to coordinate than governmental intervention. You've said that boycotts don't work, but I'd say that if the market doesn't want a product, it will fail; why does a government have to enforce that? A boycott may fail because a loud minority isn't representative of the overall market; isn't that the behavior you want?
You may find the references to some communist revolutions distasteful, but I think it's an appropriate comparison in that folks see government as the solution to a problem that markets solve really well. I specifically didn't say that what you're espousing will lead to 50m deaths, but I do think it's a useful tool for showing where government can take on too big of a role with regards to things that markets are very efficient at doing themselves.
>"I suppose I'm fine with that static-ness."
Isn't Hacker News full of people that explicitly disagree with this? That static laws are a big tool for incumbent companies to build their moat, and entrepreneurs try to break those down in ways that are intended to be better for everyone. Maybe you disagree (which I respect), but in many ways, static laws look good on the surface but have real downsides in the long run.
> aka smart homes that secretly advertise to you.
I don't think that these things should be done secretly. I think companies should have to make this information available to consumers, and I think the government should enforce this to some degree (e.g. if a company is recording my phone calls to send back to their headquarters, absolutely it should be known).
That sounds like the Byzantine Generals Problem. Prisoners can communicate all they like, it just doesn't really matter what they agree at the end of the day.
Even if the prisoners are allowed to communicate, if there is no underlying coordination mechanism, some binding agreement, a punishment for breaking the agreement, etc. then the default state is still defection.
>What it sounds like you're saying is that the government should tell the prisoners what to do. I prefer the former, because I think it keeps the responsibility still on the prisoner ("consumer" in our case).
I think this is the part that most people (and maybe you) get stuck on because the basic idea is that you give up one freedom to gain another. The prisoners are stuck. Until each one tells the other "if you squeal on me, my brother will kill you on the outside." Now, where they were once trapped by the rules of the game a new rule has been added. Do the prisoners gain or lose freedom when they threaten each other? The answer is neither because they never had any to begin with. In both scenarios they are trapped into picking a specific choice by their own rationality. The difference is that pre-threat they were trapped into a suboptimal decision and now they are trapped into an optimal decision. Coordination is the only thing that mattered here and it allowed everyone to benefit.
Now if we move this analogy over to government intervention in markets, which component of the analogy is 'government'? Easy, it's the mutual threat. The government isn't some outside entity and it isn't telling the prisoners what to do, it's the agreement between the prisoners and it's the enforcement mechanism of that agreement. The government here isn't reducing freedom (or increasing it) on an individual level but on a societal level it's granting us the freedom to move to more optimal outcomes.
>You've said that boycotts don't work, but I'd say that if the market doesn't want a product, it will fail
That's just not true. The prisoners want shorter jail terms, yet they'll fail to get them. Where you and I are clashing is the difference between individuals and society.
>why does a government have to enforce that?
Because otherwise it won't happen. There has to be an enforcement mechanism to for coordination to work. Otherwise defecting will dominate.
>You may find the references to some communist revolutions distasteful, but I think it's an appropriate comparison in that folks see government as the solution to a problem that markets solve really well.
Like I said, see Mao's Law. It's absolutely distasteful and not an appropriate comparison at all.
Besides we're specifically talking about problems that markets don't solve.
>Maybe you disagree (which I respect), but in many ways, static laws look good on the surface but have real downsides in the long run.
I do, but if you look at what I said you can see that my 'static' scenario and my 'fluid' scenario are both equally static. In one we're static because we're stuck by government regulation, in the other we're static because we're stuck by market rules (we can't coordinate.)
>I don't think that these things should be done secretly. I think companies should have to make this information available to consumers, and I think the government should enforce this to some degree (e.g. if a company is recording my phone calls to send back to their headquarters, absolutely it should be known).
You know who else liked government enforcement? Hitler. This is an appropriate comparison btw.
Yes I would. I don't see why you need to "organize" it though. I don't need any form of organization to not use product I don't like.
> It's because we lack a coordination mechanism, the incentive of individual boycotters is to cheat. You lost me here. You have internet full of coordination tools if you really feel like boycott of echo should be coordinated, just set up a fb group :).
Boycotts are basically ineffective, especially the social media kind.[1] Why are unions effective at coordinating? Because there's a punishment if you don't coordinate (you have to leave.) Why are elections effective at coordinating? Because everyone reveals their preference all at once anonymously and the collective decision is binding. Why do I and my coworkers coordinate? Because I have a supervisor who will fire me if I screw up.
In this way I see your point is that we should only use coordination mechanisms which are basically ineffective. As soon as we approach a coordination mechanism which is effective (government) you start to condemn it.
Don't get me wrong - I see your point, but in order for this "ad tax" to be legitimate in my eyes it would need to be an actual issue discussed before the elections. If your elected officials declared, that their priority will be to introduce advertisement tax and they won the election with this promise then I'd accept it as a reasonable action. (but I still believe that it might be easier to just ignore all overly ad-ridden products and wait for ad-free alternative)
For TV you can pay for ad-free experience (like HBO), this option is comming to online services as well (youtube red). I feel that this process is more efficient and faster way to eventually get echo like service that won't sell it's users as a product.
My concern with gov't action is the second and third order effects. Governments are implemented primarily by bureaucrats, whose primary incentive is to maintain their job. Any policy that gets implemented poorly (inefficiently), tends to continue because of the employees personal incentives.
This suggests, to me, that governments (as they are in America today) should not be experimenting. Culling the losers is super important for exploration, and probably the thing our local/state/federal gov't is worst at.
(If a company makes repeatedly poor decisions, they will tend towards insolvency/change in management.)
So, yes, we as people, as citizens, and as a society, through the collective apparatus we call "government", can and should step on advertising quite strongly to help us as people, as citizens, and as a society recover from the onslaught of the last few decades.
This is but one way.
I was not questioning whether the advertising industry is good or not, but rather OP's notion that he is entitled to ad-free echo service.
I agree that advertisment should be more heavily controlled in instances where people don't have a choice whether to see the advertisment or not. But nobody is forcing anyone to buy a device that is known to serve you ads. If enough people will make this choice, the business incentive for the producer of the device will disappear and so will the device.
Government exists to curb abuses.
The internet and smartphones are, but even those are not snooping in on your conversations (or so we believe, hence the hue and cry about Facebook listening in). Alexa is different in that it’s listening to all your conversations, including presumably extremely private ones.
My thesis in case someone is interested: https://ir.canterbury.ac.nz/handle/10092/13598
Of course, this assumes people still 'own' their devices, at that..
For amazon to say, "Use _____ to remove that stain," and allow my company to put in a $0.35 per request bid to fill it in with "Jeremy's Awesome Cleaning Sauce," seems inherently more wrong.
It could be useful to allow people bid on specific kinds of general queries. "You can use our cleaning product for that" is a better answer than "Sorry, I don't know."
There's also some signalling here. It shows that someone is willing to bet money that they have what I want. This is similar to search engine ads.
Like with search engine ads, transparency & privacy are really important. I want to know when I'm receiving sponsored content. I want to know what information was used to decide to show me that content. Duck Duck Go gets this right. Podcasts that play special background music during ad breaks get this right.
Maybe Alexa needs a separate voice or tone for responses with sponsored content.
When I ask for, what I expect to be, a data-driven opinion about toothpaste from a device that is effectively my oven timer, I expect it to be honest!
The real problem with Echo is that it's an always-on microphone connected to a networked computer that runs proprietary software (see https://www.gnu.org/proprietary/ for how proprietary software is often malware). Users have no software freedom to inspect what the software on the Echo does, change the Echo's software to make the Echo do something different, or share the Echo software (whether modified or not) to help their community. Even if Echo's software comes with some form of user controls, those controls are implemented with proprietary software, software users cannot trust. No matter how technical and willing a user is to improve their Echo device to meet their needs, the permission to do so is purposefully missing. Amazon does not respect a user's software freedom. Therefore there's no good reason to trust the Echo. For all we know Echo is now or at some later time will relay whatever is within mic range to someplace else, thus presenting a constant threat to one's privacy.
Nobody needs Echo; it does nothing users weren't perfectly fine to do another way by visiting a website to place orders. But one should reconsider doing business with Amazon at all (see https://stallman.org/amazon.html for details).
Advertising is a very visible effect of both privacy-invasive systems and proprietary software. People care about how ads affect them because they can see it.
Also, the case of what the "average consumer cares about" (almost invariably an evidenceless case which serves to benefit a business) is often pitched but always shown to be wrong. As Glenn Greenwald points out very clearly in https://www.youtube.com/watch?v=IOksJKfapVM (around 28m30s) he has challenged the notion that privacy is vague or somehow unimportant, or that anyone doesn't care. In fact, he demonstrates how people who say something like what you assert, "They don't actually mean it. At all. And the proof of it is they do all sorts of things to safeguard their own privacy; I mean we all have things to hide. There are things that everyone in this room would be willing to have their spouse or their best friend or their physician or their lawyer or their psychiatrist know but would be mortified to have anyone else know. The people who say they don't value privacy, they put and use locks on their bedroom and bathroom door, they put passwords on their email and social media accounts, they do all sorts of things there's a place they can go in the world to think and reason and explore without the judgmental eyes of other people being cast upon them. This is really critical to human freedom.".
Greenwald challenges people who pitch a line quite comparable to what you said to take him up on his offer to divulge all of one's credentials to everything they have (no exceptions, all accounts should be included) by emailing them to an account he maintains. He asks them for this info and tells them explicitly, "I just want to be able to troll through what you're doing online and publish under your name...". A perfectly reasonable request for anyone who wants to insist people don't care about privacy. The results of his request: "To this day, not a single person has taken me up on this offer.". That whole talk is worth hearing and reflecting on.
People do value their own privacy even if they hold immature evaluations of the importance of privacy or don't know enough about how computers work to think through the consequences of their tech-related choices. People need to learn about the cost of "things being free" and hold mature discussions of important issues of our day which certainly include how everyday technology affects one's privacy (and the privacy of anyone who goes to their house within range of that Echo always-on mic; it's not just about your choice for yourself in your home). People need to learn about the inherent value of privacy as well as the notion that some degree of privacy is not easily regained when lost, so one should not trade privacy away for convenience (not that I agree with you that Amazon Echo is convenient; even on its own merits it has already failed to distinguish between its owner and people on TV demonstrating the device causing lots of unwanted orders. This clear exposure of a security problem is hardly convenient). We do that work, in part, with discussions like these, not by downplaying the importance of privacy but by teaching people the very practical outcomes you inchoately chastise Stallman for presenting.
[1] https://blog.mozilla.org/blog/2017/11/29/announcing-the-init...
edit: follow up clarification, the echo is NOT doing voice processing on the device itself, it ships it up to the cloud to do so. You could of course set up something similar to that using a raspberry pi and shipping audio to your desktop to be processed.
I personally strongly recommend people look to put the "personal" back in personal assistant. Build something that does what you need it to, you'd be surprised how easy it can be, and how educational doing it is. I've learned a ton of stuff I didn't know just implementing basic features.
It kind of grew organically. I don't think when I started if I planned to build what I've put together so far, I would have ever started, because it sounds daunting, especially given my level of experience when I started.
Do one thing. Then do something else. :)
As a sidebar, if you want the general knowledge type queries, Wolfram Alpha's API is incredible for this. They have a "spoken" API which presents the result of your query as a single text-string, ready to be output to a text-to-speech client. I think adding the ability to use Wolfram Alpha was all of a couple lines of code tops to my existing project. And of course, it's free for way more queries a month than you'd ever run on a personal use project.
Previously their business model was simply to create a fast-lane into Amazon ordering (e.g. "Alexa order me some toilet paper"), but if what the article says is true they're going to ruin normal searching by only giving you product results (Q: "alexa how do I remove a wine stain?" A: "buy Dr. Bubbles Super cleaner"). If Echo is going to give me bad results for questions I'd just stop using it for that.
I'll use whatever personal assistant gives me the highest quality results. If your business model inherently degrades the quality, then your business model is self-destructive.
Ultimately this is the problem. They don't have to be good, they just have to be less shitty than their competition and the major players are making their money on ads.
most likely all these assistants will give you the "sponsored" results first. The sponsored products always appear before highest rating ones whenever I do a search on Amazon, Google, etc.
"Oil stains are better cleaned by treating the stain with vinegar and, after a few minutes, using an specialized local cleaner. Would you like to by one? I will arrive in 1 hour."
"Ok alexa"
"Would you like Dr Bubbles Super Cleaner, for 5$?"
"Sure"
Anyway, I'm sure they can include shitty advertising without diminishing the quality of the results in a way that normal, non-privacy-aware people will care about.
Obviously the simplest solution is credible open-source alternatives; or there is always that pesky thing, the legislative process.
"Proctor & Gamble as well as Clorox are reportedly in talks for major advertising deals that would allow Alexa to suggest products for you to buy."
The thing with the current voice assistants is that they store every. single. piece. of voice snippets they hear. Those recordings sit on their servers. At the end of the day, these corporations have a purpose other than creating a voice technology--they have very specific ways to pay the bills. Amazon is an e-commerce platform. Google lives off AdWords. To think they won't utilize the data they're collecting.. doesn't sound like a strategic decision they'd make.
For those looking for a voice assistant that represents them--and not the corporation behind it, is Mycroft. A way to democratize voice and AI.
Nonetheless, we're in the process of making our second device and if you're interested, you can get updates on it in the link below! We'd love for you to join our community and help us create an open voice assistant. Head to https://mycroft.ai :)
https://mycroft.ai/kickstarter-mark-2/?utm_source=ActiveCamp...
IMO Amazon can't afford to squander their voice lead by tarnishing the Alexa brand with poorly implemented ads.
When the technology (NOT any particular product) eventually reaches "general public use"[1], the test established in Kyllo v United States[2] means that the government does not need a warrant to use their own devices that use the same technology to see "details of a private home that would previously have been unknowable without physical intrusion"[1].
Corporations pushing their surveillance devices will change where privacy is expected. The market cannot fix this because Amazon et al can still damage my future 4th Amendment protections simply by selling their surveillance devices to other people.
In Kyllo, the question was whether the general public used thermal imaging to look at people. Even if every household had an Alexa, that doesn't mean the general public uses Alexa to listen to people (just that Amazon does).
By your logic, the Kyllo decision should have been the opposite since the general public does emit unshielded heat signatures (slash use voice assistants) but doesn't observe those signatures (slash/process receive voice assistant requests).
Two lawyers I've talked to about agree with my interpretation. Dan Geer has also been warning[1] about the consequences of the Kyllo decision.
> that doesn't mean the general public uses Alexa to listen to people
The Kyllo decision doesn't dep[end on the use, only if the public is generally familiar with the technology. I linked to the court's opinion in my previous [1]; the complete opinion is at the end of section (b) on the first page:
>> obtaining by sense-enhancing technology any information regarding the home's interior that could not otherwise have been obtained without physical "intrusion into a constitutionally protected area," Silverman v. United States, 365 U. S. 505, 512, constitutes a search--at least where (as here) the technology in question is not in general public use.
The court did not specify how the general public was using the technology or the purpose for using it. Justice Stevens dissent specifically addresses this problem
> Despite the Court's attempt to draw a line that is "not only firm but also bright," ante, at 12, the contours of its new rule are uncertain because its protection apparently dissipates as soon as the relevant technology is "in general public use," ante, at 6-7. [...] it seems likely that the threat to privacy will grow, rather than recede, as the use of intrusive equipment becomes more readily available.
> the general public does emit unshielded heat signatures
This is irrelevant. Of course they do. However, the court observed that the general public \ was not using thermal imaging cameras to receive those emissions, and thus the had an expectation of privacy about their heat signatures. Note that this expectation co9ud change if most cellphones start including the newer "multispectral imaging" devices.
> but doesn't observe those signatures
People using an Alexa do expect that some of their speech will be recorded and transmitted to a large business over the internet.
I'm torn in the thought. At face value, I can't but agree. However, I also enable crash reporting in many applications. Have for years. Firefox gets usage stats. At some point, I have to wonder why I trust one group of people and not another. More, have the groups I trust changed out from under me?
I also have a robot vacuum cleaner, but I don't connect it to my network. Too much potential information can be gleaned from a floor plan and the obstacles around the house. You'll never get me to install one of those telemetry devices in the OBD port on my car for any company. I don't care how much money it will save me on my phone or insurance plan.
Is it all rational? Of course not. My personal policy concerning my personal data is a mish-mash of contradictions, because it's fatiguing to be on the offensive 100% of the time. For example, getting a Nest thermostat is something that I want, but I don't want it phoning home to cloud servers. It's heating/cooling data about my house. That's probably some of the most innocuous data you could have about someone, yet if it isn't self-hosted, I don't want it. Yet Maps has tons of location data on me, Gmail holds my emails, and I have plenty of conversations over Hangouts.
Contradictions all the way down.
My FB usage is now scaled down to basically just an event RSVP service with a few token page subscriptions for bands and venues. I've removed all photos and videos, cleared out all profile information, set visibility on everything to "just me" or "friends only", and I've untagged myself from everything I was ever tagged in. Yes, I know FB is probably keeping a lot (all?) of that in their archives for an indeterminate period of time, maybe the EU "right to be forgotten" can help with that.
I've replaced Google search with Startpage and I've migrated my email, calendar and contacts from GMail/GCalendar to a paid hosting/email provider, with my own personal domain name.
My phone is still running stock Android 7, but I've significantly cut down the amounts of apps installed, I much prefer to use web versions in Firefox with a decent lineup of privacy-protecting and tracking-inhibiting extensions. Some mobile web sites are really shitty about forcing their apps on you, but I'd rather just do without them. Obviously on the desktop, Firefox with the appropriate extensions is the browser of choice.
I refuse to get any "smart" devices, they're not welcome in my home. I'm working on replacing my Chromecasts, if nothing else I'll just use my laptop and an HDMI cable. For music in my kitchen/bathroom/basement, I've got a decent BT speaker instead.
Kicking the Google/FB/whatever dependency is hard.
I guess part of the reason is I started using GMail, Google Maps and Facebook long before I developed my cynicism towards our market reality[0]. It's way too much effort to drop GMail now, Google Maps are way too convenient, and most of my real-life acquaintances are on Facebook.
But I strongly avoid introducing any new SaaS and other spyware software in my life now.
--
[0] - I should say "cynicism towards tech industry", but it's not really about technology per se - it's about regular market incentives that fuck up everything, regardless of the technology involved.
i.e. you are very protective of information about the inside of your house: floor plan, temperatures. Makes sense considering it's your home).
Seems like the same thing with the car, though. It seems like you think of the inside of the car or the workings of the car as a more private space than where the car has gone?
Emails and Hangouts conversations are harder to pin down. Could be that they are more abstract and harder to maintain a concept of private vs public spaces on the internet? Could also be that there aren't really any alternatives that provide an obvious privacy benefit for their cost/effort.
Less than $15/month for full-featured IMAP email and groupware for an individual. Is it super cheap? No. But you can't argue with their privacy track record.
https://en.wikipedia.org/wiki/Kolab_Now
(I'm not affiliated, in fact I migrated from them to a different provider that also does web hosting. But I was very satisfied while I was a customer)
My email for my custom domain is on FastMail, but I have a couple of gmail addresses that would take a long time to get fully migrated over.
The company I work for also uses GSuite, so I can't really escape it, anyway...
Totally agree. I don't feel like it would be 'saving' money either.
I think the thing to realize is that trust is inherently faulty. Trust requires some level of unproven belief. This is why it has to be earned and why it is so easily broken.
I think the better way to put it is that you shouldn't really "trust" any company (or anybody that has motivation to exploit you) 100%.
You can establish different qualifiers to trust, as well. For example, I (currently) trust using my Echo for small calculations and timers. If Amazon were to sell that data to 3rd parties, I would trade that bit of privacy (about my timer usage) for the convenience. The moment I see anything about Amazon doing an "always listening/recording even without the wake word", that thing is getting unplugged or thrown away. (if this is already a KNOWN thing, somebody please tell me)
tl;dr trust isn't an all-or-nothing proposition.
I do the best I reasonably can, in regards to privacy and trust. When I identify an issue, such as a privacy hole I wasn't aware of, or a company that proves itself undeserving of trust, I update my approach. Maybe I'll have to migrate some services or buy a more privacy-focused alternative, maybe revise which information I provide in user profiles, but I think that's worth it.
You gotta start somewhere.
I was actually feeling quite irritated this morning because I have to sign in to use Visual Studio - in order to have my settings. I don't know what stuff they collect, but it's not nothing because they can for sure track my sign ins.
You raise a great point about who we trust. I feel okay trusting Microsoft, but there's no way in hell I'd trust Google or Amazon.
Recently I got a Razer Orbweaver, heavily discounted, and installing the drivers for that required me to create an account with Razer and install "Razer Synapse". I went along with it thinking that it would at least give me one package to keep drivers and options for all three of my Razer devices in one place. But it doesn't, Synapse only recognizes the Orbweaver, the other two Razer peripherals still need their own dedicated driver and software packages installed.
Synapse phones home all the time when booting up it tells me how it's connecting to the "Synapse cloud". Sounds really fancy and all but what use is that if it can't even get drivers for other Razer devices from the "cloud"?
This feels like getting all the disadvantages but none of the advantages.
I couldn't have said it better. It's funny because I'm totally cool with having an account provided it does things like gives me access to software, drivers, tools, backups, whatever. But yeah, I'm not going for it if it's just an excuse to vacuum up data. It's gotten to the point where I actively avoid companies/products that just don't get the message.
To this point, I switched from Google to Microsoft a few years ago because I really got sick of feeling like my personal data was just making them richer in exchange for nothing. Email and web apps aren't worth the loss of privacy.
At first Bing search was a bit weird, but then it turned out that it worked just as well as Google for most things, and better for some.
But in regards to private data, the parent does have a point. Google's main business is aggregating data for ad purposes, selling their users privacy is how they make most of their money, you wouldn't expect anything else from them.
MS does some of that too, but they make most of their money through other means, so they have less incentive of disrespecting their user's privacy. Even more so considering the PR blowback if it should ever turn out that Windows spies on its users so MS can sell that data.
I feel curious as to why that is.
Is it because you know Microsoft is blatantly in it for the money, meaning you are the direct customer. As opposed to Google/Amazon/Facebook, where you're the product?
I definitely don't trust Google, Amazon, or Facebook at all. I'm only on Facebook a few times a month, and only from my computer. Never from my phone. No way in hell I'd ever install the FB surveillance app. FB is on probation with the FTC until when? 2030? They're an evil bunch of jackasses.
IMHO, Amazon is the most sociopathic of the big four/five. They don't invade my privacy (because I don't let them), but I think they're pretty evil. I mean, Google is sort of soft evil whereas Amazon is WalMart level evil.
If you consume your content through audio then you're forced into listening to the ad / sponsored answer.
"I'm not affected by ads" is just a lie we tell ourselves. Why aren't you just blocking them? Around these parts it's basically digital hygiene.
For some unknown reason I thought adwords ads came through and I just happened to subconsciously ignore them, but having just Googled for something now, I see they are not even there.
Funny how the brain works.