No. The only way to be really safe is to go live in a cave, not to have any friends and never go anywhere near any tech ever again. I'm sure you can tell that I don't consider that to be much of an option.
It's unhelpful to label 'data collection' as the source of the problem as 'collection' happens everywhere in our interactions with the world. There are valid arguments and discussions to be had about who really owns that data. It may well be about me, but that doesn't necessarily mean that it's mine. I'd say the majority of people are only just beginning to understand that a discussion needs to take place even though technical folks have known this for a long time.
One solution to the 'ownership' problem is to make it possible for everyone to run their own infrastructure. i.e. have their own 'backend' that apps/services can be installed into, which their end-devices can then connect to. This is already possible for the technically savvy but a lot of work is needed before I feel we can trust such systems (I'm working on approaches to this [1,2], based on unikernels).
In the meantime, I applaud business models that are not built around profiling (which is the crux of the issue cf. advertising). They're the only ones where the incentives have any hope of aligning.
+ "Run your own infrastructure" is currently a pipe dream for what is 99.9% of people.
+ Companies selling data when they go bankrupt or M&As changing the nature of what they do with the collection happens regularly.
+ Discussing ownership is fairly simple as well, "Does the company already have your data?"
+ Any business not based around profiling has a additional options to make more money by selling their (your) data.
Why isnt excess data collection a problem? Is it because its very very hard to solve and if we pass laws bad actors will continue to flaunt them? Is it because we are past the point of no return?
Or not even a dream, because those 99.9% prefer not to run their own infrastructure even if it's easy and cheap.
However, "local apps" (ones which don't require infrastructure) could be appealing if they could achieve parity with the enormous-backend apps.
I tried DKIM, but the e-mails were still sent to Gmail's spam folders. Eventually I gave up on self-hosted e-mail.
If you look at the computer history, we already had several cycles from single computer to timesharing to mainframes/supercomputers to PC to thin-clients/mobile to (?). There has always been a time when single powerful computers were in the majority and then there has been times when thin-clients with big server structures were the majority. The Internet supports also the model of more decentral computers and everyone could host their data on their own devices. It's currently just not the right time and there is no business model behind it, so no one push for that solution at the moment.
I think the biggest difficulty with balancing privacy concerns against other factors -- or even encouraging debate about and awareness of the issues among non-technical friends and family -- is that an item of data is itself neutral. It is how that data is used or combined with other data that may or may not be in any given party's interests.
Sometimes the exact same technology or data could be used for very good purposes, useful and generally harmless purposes, or hostile purposes, depending on the context. For example, consider automated number plate tracking of motor vehicles. If your child has been kidnapped and a witness caught the plate of the kidnapper's vehicle, you're going to appreciate the police being able to find and intercept that vehicle as quickly as possible. If you're an urban planner responsible for keeping transport infrastructure as efficient as possible in the face of a rising population, an aggregated data set showing how real travellers want to move around your city could be very useful, helping you make decisions that improve the system for everyone. If you're that same urban planner but on the side you're working with a load of jewel thieves and abusing your access to historical movement records to figure out when specific wealthy residents are usually away from their homes so the thieves can break in and then abusing your access to real time tracking to confirm that the residents really are out or warn if they come home early, that's not such a happy ending for the data subject. Analogous issues arise with many kinds of personal data, including more sensitive areas like financial or health data.
In the modern world, with vast databases and powerful data analysis tools and effectively instant communications and effectively unlimited storage, sometimes seemingly innocuous data can also give away a lot more about you than you might want or need. Things like what you bought at a store, or who you were tagged with in photos on social media, or a recording of you walking across a street on CCTV, can be used to determine many apparently unrelated things about you with relatively high (but, significantly, not complete) reliability, again sometimes quite sensitive ones that you might very much prefer to keep to yourself. Consider the store loyalty programme that determines a girl is pregnant from her purchasing patterns long before her partner or parents know. What about the person outed as gay because they were tagged with a whole group of openly gay people on holiday? Oh, by the way, unlike their friends in the photos, the first person lives in a country where homosexuality is still frowned upon and being open about it has real consequences. Also, gait analysis said you looked nervous as you walked into the airport, but it couldn't tell whether you just read a news story about a plane going down or you have inside knowledge of a terrorist threat, so unfortunately you won't be flying today. Just wait until automated text analysis software reaches the point that it can effectively de-anonymise posts like this one, and suddenly a lot of people who thought posting under a pseudonym was going to hide their criticism/whistle-blowing/advocacy of some controversial subject realise they weren't as safe as they thought and those comments are now permanently recorded on some public web site.
We therefore need to move beyond black and white assessments like "data collection is dangerous" or "social media sharing is fun". What matters is not just what data is collected, but also who has access to that data, what they are allowed to use it for (including for how long it's stored, what else it can be combined with, and the like), and crucially, how these things can effectively be controlled or regulated to ensure that everyone is playing by the rules when data is data and once someone has it for one purpose it can readily be used for another.
I think at the core however the issue is not how the data is used but one of ownership and privacy. It is "my" business where I drive. It is "my" business whom I call. It is data about me and my behavior. So to have that information about me used, even if it could be assessed publicly, without any need for consent from me becomes the issue.
License plate readers are tantamount to placing a tail on me. The same with tracking my movements via several CCTV cameras, etc.
Retention just makes the situation worse. Yes, you could approach it with laws of appropriate use of this data but the problem is what about when the laws change. And the data is still there. Similar to how people gave data to Radio Shack years ago to find it now being sold to 3rd parties. Privacy policies are constantly shifting for the companies' best interests. Who can keep track?
The problem is that once the data is collected it is out of your hands. And why the regulations need be at the collection level with strict opt-in requirements as well as the usage and retention level.
True enough. However, it is also the business of the people who are responsible for building and maintaining the roads to know how those roads are being used.
It is "my" business whom I call.
True again. However, it is also the business of the phone company who must provide the service you ordered and to which you will owe money based on the calls you make and, in some cases, who you are calling.
I do understand your concern about data being out there at all. However, the reality is that we interact with the world and the people around us all the time. Sometimes that will result in data that refers to us but also affects other people for legitimate and often unavoidable reasons. So I don't think an extremist position that no-one should ever be able to collect data about you without your explicit consent is ever going to work. As others have noted, that would mean you couldn't interact with almost anyone or anything in the modern world, and unless you're planning to live 100% off-grid as a hermit that's just not a viable possibility. Instead, we should consider issues like the retention and repurposing of data.
Licence plate readers that are automatically tracking cars through a congested area that has variable speed limits are potentially in everyone's interests: smoothing out the traffic flow makes everyone's journey faster and safer, and has basically no downside. However, once a vehicle has left that area, it is no longer necessary to keep any specific details about it for that purpose. The data can be discarded, or completely anonymised simply by turning each plate into a unique but otherwise meaningless number before it's recorded if it's useful to store aggregated data for more general traffic planning purposes. Similarly, if plate recognition is being used for enforcement of that speed limit, there is no need to record the details of anyone except those the system has determined to be exceeding the speed limit, where the evidence will be used for a subsequent prosecution. Once any resulting legal processes have run their course, the data can also then be completely discarded if no conviction resulted.
The risk in either case is not the scanning itself, it is the retention of the data and potentially use for other purposes and correlation with other data sets later. Given robust rules about keeping personal data no longer than necessary for its stated purpose, and probably about declaring its stated purpose in a meaningful and usefully specific way, this is not so much having a tail as having a driver in a car behind who happens to be following you for a while on the same road but then forgets they ever saw you within moments of going your separate ways. I doubt even the most privacy-conscious person would consider that an unreasonable risk in other contexts or expect to be able to prevent it.
Apple improves their products for several reasons, but as a for-profit business the main reason they would use some of their limited capital to "improve their product" would be to make money.
The details about how a business exploits user data is not that interesting. Arguments about this kind of minutia are a distraction and a waste of time. The problem is, as you say, the fact that they collect data at all.
> That means gimped Siri
As Dan Geer said regarding the inevitable software-industry complains about liability legislation
"Yes, please! That was exactly the idea."
In his recent (highly recommended) talk[1], Aral Balkan suggested a very interesting model, because there is another product that is well-known to adversely impact the people "near" the user as a secondary effect: tobacco. Siri, Amazon's Echo, and most of Google's services are like software "second-hand smoke". Like smoking, I don't care what you do in your own home. What we need - at a minimum - is new social conventions/etiquette where you ask permission or step outside before enabling any of these recording devices.
[1] https://projectbullrun.org/surveillance/2015/video-2015.html...
Not on a social networking site? Don't worry, they probably know a significant amount about you anyway, just from friends who have given the social network the entire address book from their phone and a few photos with you in them.
Decided not to use Google because you're not sure about how much data they collect these days? Don't worry, they probably have a significant proportion of all the e-mails you ever send from one of the other parties anyway, and you have no reliable way to determine when this is happening or to avoid participating.
And of course there are the numerous third party widgets that not only make the modern web a slower, sometimes malware-ridden, often ad-ridden mess, they also follow those not technically skilled enough to protect themselves around. Don't worry, they promise they won't tell your health insurer about the medical conditions you were looking up. No, wait, they didn't.
Now we have the kinds of technologies mentioned in the parent post, with smartphones and headsets and home automation systems that have sensors and network connections being placed just about everywhere, and we really are starting to live in a society with permanent surveillance.
In many ways, these kinds of issues are much harder to address than the already difficult questions of how to control first party data collection. Most of them by their nature operate in the background where a lot of people will never even know they are there, and even those who do may or may not be able to do much about it.
Social conventions would not/do not work as much as we would like.
Siri (as well as several other similar products) is based on Nuance technologies (speech recognition, text-to-speech, etc). It was recently in the news that Nuance outsourced their voice technology quality control (US based human workers listen to your voice data and checks the speech recognition quality, some workers mentioned they recognized single persons from reoccurring voice data). So your voice data is being processed by several parties.
The two big mobile OS vendors Apple and Google and also smaller players like Microsoft make it incredible hard to not-use their cloud storage&sync offering. Storage space is synthetically limited to low GB, with an high up price for additional storage. An SD-Card slot (= cheap extra storage, as a card is very cheap) is only available in Android devices (third party vendors hardware manufacturers like Samsung, Sony, etc.) but not from Google itself. All the inbuilt sync functionality only works with their own cloud storage, no third party cloud nor a simple sync over Bluetooth, Wifi or USB-cable. They could do better.
When Google(+) Photos auto-tags my photos, still my data, although I'd admit the IA training requires a big amount of photos.
Looking at https://www.google.com/landing/now/#cards ; it's obvious a lot of those need online information. It's not obvious that the data source for requiring those online informations has to be in the cloud, or combined from multiple people's data.
It is, but that's pretty useless unless Google also knows so that it can send you the relevant updates.
>When it knows where I work and live to ask Google servers the weather or how long it would take to go somewhere, still my data used as the request base.
The weather, possibly (although again, it's not much use unless Google know enough to target the local weather at you). The how long to get there, not so much. One of the key things they use is info on where everyone else is. This is a classic case of pooling everyone's data in order to add value.
It's shortsighted to say that just because (example implementation) my device polls their server for $team, that it is the same as "google knowing" I like $team. The interface does not need to be authenticated, or keep logs. There is just a slippery slope between monitoring for diagnostics and abuse that incremental business value very quickly slides into tying team-specific view counters and a favorite team to my user identifier.
And Google doesn't necessarily know that you like $team. It knows that the email address you used to sign up to the alerts has requested updates on it. Is that really particularly scary?
Even with a locally installed AI, it will need to be trained on real and up to date data, it needs feedback on what was well recognized and on which data it failed, etc.
the training and update cycle can be done on an individual basis, but having a distributed system sharing a huge set of varied data would accelerate the process exponentially.
Ideally training data shouldn't be stored once it is ingested by yhe AI, but we'd need at least interconnected AIs dharing data from different users
Viewed differently, the value of N specifies how uniquely identifiable the user is.
I gateway all my devices through one computer running a stripped down kernel. Through this host I monitor all attempted network connections. I block ads and of course this means I block Apple.
Apple devices are like beacons, attempting to contact a variety of Apple servers from the moment you turn them on and continuing to "phone home" incessantly 24/7 until you turn them off. That is only one example, but to me it is the most egregious.
Apple's competititors are no better. Do users really believe that one of these companies is going to protect them?
When the owner of a Chromecast plugs it in for the first time it tries to connect to Google to patch itself and prevent the owner from jailbreaking.
Perhaps my favourite was this patent application from Apple, with the CEO as a listed inventor.
http://www.google.com/patents/US20090265214
As for "providing advertisers with guided access to your eyeball" I believe that is what the App Store system does; it lets the app developers do the providing by embedding advertising into applications (cf. operating system, as in the above application).
Do you do this always via a vpn or something or was it just to try it out?
Yes, I do this always.
For me, the ad blocking is most easily done via DNS. djbdns features make it especially easy. I have been using DNS to block ads especially in app ads for several years now. Never expected it to work so well (such heavy reliance on DNS by developers), but it does.
There are of course other ways besides DNS to block unwanted connections including ones to Apple and third party ad servers.
As for Apple's VPN capabilities, I have experimented with xl2tpd but not as a way to block ads.
If you use Google's products(and who doesn't use Google Search ? ) , Google already knows plenty about you. And even if you don't use Google products, Google knows plenty about you, since it's tracking software is installed in large percentage of sites, etc.
And if you add machine learning on top of that , the amount that Google knows about you is great - even if you use Apple's products.
In reality, by using Apple's products, you don't gain privacy , you lose privacy - now two companies know a lot about you, instead of one.
With the car industry about do become part of the ecosystem it makes sense. Get in a Google Taxi and you'll have to watch targeted ads the whole way there. Get in an Apple Taxi and listen to classical music.
Essentially Tim Cook said: "Apple is a good girl (we do the same things that bad girls do but better)"
The fact that Google abuses the data and Apple doesn't is already a major difference. Just because government can use its monopoly on violence and imprisonment to go one step beyond doesn't make what businesses do harmless.
Yes, the collection is already inherently dangerous, and what Apple does is certainly not harmless. And yes, they will use that data to their advantage.
But the unscrupulous abuse of that data, the blatant disregard for privacy, a business model that depends on abuse of that data, and the obvious hunger for social and economic power by Google and co is not a minor difference.
The difference that Tim Cook wants you to believe in is that Apple doesn't directly make money from your data, they just use it to improve their product; alternatively, Google makes money from your data by providing advertises with guided access to your eyeball.
But the danger isn't in the fact that Google lulls you into complacency with free services. The danger isn't in the fact that Google sells guided access to your attention. The danger is just in the collection of data, and the fact that one day the government or somebody could find a way.
There's only one way to be safe. And that's to collect only minimal amounts of data for minimal apps. That means gimped Siri. No Apple competitor to Google Photos. And I don't think Apple will do that. I think they'll continue to amass all the data they use to improve customer experience. The fact that Apple makes money differently off the data doesn't change the fact that it's collection that's inherently dangerous.