>Which seems to be a different argument entirely, namely one of weighing the positive externalities against the negative externalities?
Not just that, it's also about weighing the restrictions on personal freedoms against the extent of negative externalities. In some cases the protection of personal freedoms is absolute. The ban on torture is a precedent for that.
>Now, there is actually precedent for banning consenting adults from sharing private information in order to ensure the survival of the democracy [...]
It's a precedent only if it's comparable, which it isn't.
You are overstating your case. There is no reason to believe that letting an algorithm parse some of your data for clues as to what you might want to buy makes democracy fall apart.
What is likely to have an effect on democracy is when governments analyze data for political reasons. Where that happens, people will simply stop volunteering so much private information. They don't need to be banned from doing that.
> What is likely to have an effect on democracy is when governments analyze data for political reasons.
What if companies analyze data for political reasons?
> Where that happens, people will simply stop volunteering so much private information. They don't need to be banned from doing that.
Except that they don't. Heard of a certain Edward Snowden, maybe?
Except that they can't because the data is already there and isn't under the control of the affected people?
Except that the nature of data is that people don't know how it's being processed?
Except that there are network effects that limit people's actual freedom to decline using a quasi-monopolist's services?
Your argument could pretty much be applied to the Stasi in the GDR: Do you have an idea what the consequences were when individuals simply stopped providing more data?
I never disputed that there are dangers and negative externalities to lack of privacy. But democracy falling apart isn't one of them.
Your logic is flawed. You are reversing cause and effect. Where there is a lack of democracy, privacy is often taken away in order to make it easier to oppress political opponents. But giving up some privacy in exchange for services doesn't make that happen. Yes the Stasi would love Facebook's database, but using Facebook doesn't cause the Stasi.
Snowden isn't proof of that at either. The NSA hasn't hunted down and imprisoned supporters of Bernie Sanders after all, has it? American democracy hasn't fallen apart and where it is lacking it has little to do with Facebook.
But please don't get me wrong. I find it very problematic when governments scan everybody's data. It could have an effect on how people behave even if it doesn't make democracy fall apart. It may deter people from doing perferctly legal things because they could be misinterpreted. None of that is voluntary though. I haven't given the NSA permission to scan my data at all.
I don't use Facbook and I am very careful with information I make available publicly. I just don't feel very threatened by any negative externalities created by those who choose to use Facebook.
My problem with your line of thinking is that you are extrapolating every single danger to the extreme until you arrive at the conclusion that people must be banned from ever making these choices. As I said, you can ban everything based on that sort of thinking.
> Your logic is flawed. You are reversing cause and effect. Where there is a lack of democracy, privacy is often taken away in order to make it easier to oppress political opponents. But giving up some privacy in exchange for services doesn't make that happen.
Your logic is flawed. You are reversing cause and effect.
There simply is no unidirectional causation between lack of privacy and lack of democracy. Obviously, the Stasi took away privacy in order to suppress democracy. And at the same time, the lack of democracy helped with establishing and maintaining the Stasi. It's a positive feedback loop.
> Yes the Stasi would love Facebook's database, but using Facebook doesn't cause the Stasi.
Does using Google cause the NSA? No. Did the NSA take advantage of the data collected by Google (by eavesdropping on their internal communication), thus increasing their power? Yes.
Things aren't simply a linear chain of causes and effects. There are lots of factors contributing to certain historical developments, there are feedback loops, it's complicated. In particular, the development of power structures doesn't care about intentions behind a source of power. When the Netherlands collected information about the religion of their citizens prior to WW2, they didn't have an intention of killing all jews. But the database confers the power to find them very efficiently. Which is what the Nazis did when they invaded.
As you yourself say, the Stasi would have loved Facebook's database. Does that mean that Facebook is the Stasi? No. Does that mean that it cannot possibly attract interest from people who want to use it like the Stasi, who might end up getting access to it, thus contributing significantly to the establishment of an authoritarian dictatorship?
> Snowden isn't proof of that at either. The NSA hasn't hunted down and imprisoned supporters of Bernie Sanders after all, has it? American democracy hasn't fallen apart and where it is lacking it has little to do with Facebook.
1. So, even if that abuse by the state did not happen, what about all the abuse by the state that indeed does happen? What do we know about which role the NSA plays in those cases?
2. More importantly: What kind of evidence would you accept to consider something a risk to democracy? So far, it seems like democracy needs to have been destroyed before you accept that whatever has caused it to fall apart could indeed cause a democracy to fall apart.
> But please don't get me wrong. I find it very problematic when governments scan everybody's data. It could have an effect on how people behave even if it doesn't make democracy fall apart.
Well, except that is a major way in which it does? Surveillance changes how people discuss their opinions. People changing how they discuss their opinions is how people end up making different choices in elections. An idea that's not being discussed is less likely to gain support in an election.
Imagine a world where everything was being monitored by a Stasi-like entity, but with the exception of actually fair, secret, democratic elections, where citizens actually knew and could verify that the elections were fair and secret. Would you consider that a functioning democracy?
> It may deter people from doing perferctly legal things because they could be misinterpreted. None of that is voluntary though. I haven't given the NSA permission to scan my data at all.
1. The externalities of people using facebook aren't voluntary either.
2. The point wasn't whether you have given them permission, but whether the majority of people have given them permission. Your claim was that people would stop volunteering their data. If you go by the general reaction to Snowden's relevations, people don't care. It's probably not informed consent, but neither is it with Facebook.
> My problem with your line of thinking is that you are extrapolating every single danger to the extreme until you arrive at the conclusion that people must be banned from ever making these choices. As I said, you can ban everything based on that sort of thinking.
Except I don't, that's just your interpretation. Just because I don't write pages upon pages for a short HN comment, doesn't mean that there isn't any more reasoning behind it.
I think we can agree at least on one thing. Lack of privacy makes undemocratic regimes more effective (at oppressing people) and possibly longer lasting provided they can actually access any data that people give up to advertisers.
I'm not convinced that this is reason enough to ban advertising based business models in democratic countries. We can't live our lives preparing for the day democracy ceases to exist. Any negative effect the lack of privacy has on stable democracies seems small and rather speculative in the first place. I don't believe it's a real threat. You do believe that. So we're going to have to agree to disagree on that one.
Also, I'm not convinced that globally the effect of Facebook, Google or Twitter on democracy is a negative one. On the contrary. Free services provided by overseas companies are often the only way for people in undemocratic countries to get around their local police state. Pay services are even more at odds with privacy than ad based business models.
> I think we can agree at least on one thing. Lack of privacy makes undemocratic regimes more effective (at oppressing people) and possibly longer lasting provided they can actually access any data that people give up to advertisers.
Well, yes, but I think it's a mistake to divide the world into "undemocratic regimes" and "democratic paradise". It's a sliding scale and not a binary distinction, and every democracy on earth has some rather undemocratic stuff going on.
> I'm not convinced that this is reason enough to ban advertising based business models in democratic countries.
Not advertising based business models, but surveillance based business models.
> We can't live our lives preparing for the day democracy ceases to exist.
The point isn't to prepare for the day democracy ceases to exist, but to prevent it from ceasing to exist. Why can't we do that?
> Any negative effect the lack of privacy has on stable democracies seems small and rather speculative in the first place.
How do you know that?
> I don't believe it's a real threat. You do believe that. So we're going to have to agree to disagree on that one.
Nope, we don't have to "agree to disagree". I don't just believe that because I believe that, I believe that for what I think are good reasons, which can be examined for their logical consistency and for consistency with reality/evidence. Now, possibly you aren't interested in doing that, which is your decision to make, but that's not some kind of unavoidable fate, it's simply your disinterest in the matter.
> Also, I'm not convinced that globally the effect of Facebook, Google or Twitter on democracy is a negative one. On the contrary. Free services provided by overseas companies are often the only way for people in undemocratic countries to get around their local police state. Pay services are even more at odds with privacy than ad based business models.
Possibly. Though it should be noted that free services do not necessarily have to be surveillance-based services. There are lots of free to use IRC networks, for example. Or Tor is also free to use. But those are not surveillance-based offerings. And not even advertisement has to be surveillance-based.
>> Any negative effect the lack of privacy has on stable democracies seems small and rather speculative in the first place.
>How do you know that?
I don't claim to know anything with any certainty. That's why I used the word "seems". You are the one who is making extraordinary claims based on extremely thin evidence. You are claiming with great certainty that giving up any private data to advertisers makes democracy fall apart, and yet you are unable to point to a single stable democracy that has fallen apart for that reason. Your sample size is zero.
>I believe that for what I think are good reasons, which can be examined for their logical consistency and for consistency with reality/evidence.
I have examined your reasons, and even though I can follow you on many individual details, I find your main conclusion contrived, exaggerated and inconsistent with the thin evidence we have. You are overstating your case.
This is a very common way in which predictions of doom tend to go wrong. Take one aspect, follow it to extreme conclusions as if there were no other variables or counter forces or reactions to what is happening. This isn't logic or consistency.
And this lack of a good model starts where we haven't even discussed what privacy or private data is in the first place. It's definitely not one thing that you can switch either on or off. We always have to trust someone with some of our data. Much of our "private" data isn't ours exclusively as it may be about interactions with others. The data we give up to advertisers isn't used exclusivly for advertising either. It's also used to provide some of the services we want (like search or spam filtering).
"Ban it all!" is neither proportionate nor even possible.
And even if I did follow your argument that giving up any privacy at all makes democracy fall apart, I see no alternatives to advertising that are better for privacy.
Just look at the numbers involved and you will quickly find that funding everything like Tor or a few IRC channels is impossible. Donations, self-funding and sponsorship involve payment and are subject to extremely detailed surveillance. There is nothing more heavily regulated and traced than payment and you can be absolutely sure that governments will never allow large anonymous money flows because everything that could fund something like a Google data center could also fund organized crime and terrorism.
That said, I also see a lot of negatives in the excessive and brazen behavior of the ad industry, and I also think there is a need for regulation. A better regulated ad model is our best shot at more privacy and fewer negative externalities.
> I don't claim to know anything with any certainty. That's why I used the word "seems".
Sure, I don't expect certainty, just some justification. Are you saying it's just wild speculation on your part?
> You are the one who is making extraordinary claims based on extremely thin evidence. You are claiming with great certainty that giving up any private data to advertisers makes democracy fall apart,
First of all, I don't actually make that claim. I can see how you maybe can get the impression from the short version (as I mentioned, it's kinda hard to condense a complex topic into an HN comment), but really, it's a matter of the severity of the possible consequences combined with non-trivial probability. Which is one reason why the comparison to laws regulating how you cannot prove to other people how you voted in an election is an insightful one: Those restrictions don't exist because the probability is particularly high that any given election would be corrupted, they exist because the downside is massive if it happens even once under the wrong circumstances. We make every single election more complicated than usually necessary, in order to help protect against the few cases where the consequences would be catastrophic, even though there is not even remotely a guarantee that that would ever actually happen.
An important question to ask is: Does a stable democracy protect against those risks, or is preventing those risk by other means what keeps the democracy stable?
> and yet you are unable to point to a single stable democracy that has fallen apart for that reason. Your sample size is zero.
May I suggest that your categorization is counterproductive?
Suppose I build a bridge from papier-mâché made from cardboard that was used to package the iphone 7. You might argue that that's not a reliable bridge and it shouldn't be opened to traffic because of the risk that it might collapse, given the track record of papier-mâché and our knowledge of physics. Now, I would reject all your arguments by pointing out that no bridge built from cardboard that was used to package the iphone 7 had ever collapsed and that your sample size is zero, and that we therefore have no clue whether that bridge is a risk to public safety. Would you agree?
It's hardly surprising that that specifically hasn't happened yet, given that the technical developments are relatively new, and some of it is yet to happen. That doesn't mean that we can't draw any conclusions from similarities to other power structures in history, or that we have no clue how current or expected future technology could be abused.
> This is a very common way in which predictions of doom tend to go wrong. Take one aspect, follow it to extreme conclusions as if there were no other variables or counter forces or reactions to what is happening. This isn't logic or consistency.
It seems like you are just filling in the holes in what I wrote with the most idiotic ideas that I could possibly have, and then you pretend that strawman is what I actually wrote.
BTW, as far as I can tell you are making that exact mistake in the opposite direction: You find some aspect that isn't necessarily quite as bad as my simplified model desribes it, as there is some other influencing factor or some counteracting force that could decrease the impact, and from that you conclude that therefore a catastrophe is completely unrealistic.
> And this lack of a good model starts where we haven't even discussed what privacy or private data is in the first place. It's definitely not one thing that you can switch either on or off. We always have to trust someone with some of our data. Much of our "private" data isn't ours exclusively as it may be about interactions with others.
As you say, we haven't discussed it. From that it doesn't follow that my concepts are idiotic.
And yes, I agree that privacy is not a binary distinction, nor is privacy an end in itself, nor is some sort of "total privacy" a sensible goal, let alone achievable.
The root problem is that personally identifiable information confers power. Power is not inherently bad either, nor something that could realistically be avoided completely. But if there is one lesson we as humanity should have learned from history, very much the hard way, it is that concentrations of power are dangerous. There is no clear line between how much concentration of power is harmless and where the danger begins, but it still is quite clear that too much of it in one place/in one person's hands is a terrible idea.
If you think about it, the whole point of democracy (and the rule of law and checks and balances) is to prevent concentrations of power. And we accept a lot of inefficiency in the government, compared to dictatorships, in order to maintain that distribution of power. The laws governing the election process are only a small part of how we build democratic societies with lots and lots of expensive safeguards to prevent the concentration of power.
> The data we give up to advertisers isn't used exclusivly for advertising either. It's also used to provide some of the services we want (like search or spam filtering).
I think most of the data shouldn't really be necessary to achieve a sufficiently similar result.
> "Ban it all!" is neither proportionate nor even possible.
Depends on what exactly you mean by "it all".
> And even if I did follow your argument that giving up any privacy at all makes democracy fall apart, I see no alternatives to advertising that are better for privacy.
Is it intentional that you are equating advertisement and surveillance?
> Just look at the numbers involved and you will quickly find that funding everything like Tor or a few IRC channels is impossible. Donations, self-funding and sponsorship involve payment and are subject to extremely detailed surveillance. There is nothing more heavily regulated and traced than payment and you can be absolutely sure that governments will never allow large anonymous money flows because everything that could fund something like a Google data center could also fund organized crime and terrorism.
Why is it necessarily a big problem that payment is subject to surveillance in this context? How is surveillance of the fact that the average human in the industrialized world pays a certain amount of money per month for telecommunication and data storage services equivalent to surveillance of the contents of their interactions with each other?
Also, who says that we need something like a Google data center?
> That said, I also see a lot of negatives in the excessive and brazen behavior of the ad industry, and I also think there is a need for regulation. A better regulated ad model is our best shot at more privacy and fewer negative externalities.
And I don't think there is one that's going to improve much other than to forbid the collection of personally identifiable information, possibly unless it's completely optional (not in exchange for anything) with requirement for informed consent before any data is collected and with the option to revoke consent at any point, in which case all collected data about a person would have to be deleted immediately.
Just one last thing from my end on the dangers to democracy. I never said or thought that your bringing this up was idiotic. It is a valid concern that is worth thinking about. I'm just not reaching the same conclusions as you at this point in time.
>Why is it necessarily a big problem that payment is subject to surveillance in this context? How is surveillance of the fact that the average human in the industrialized world pays a certain amount of money per month for telecommunication and data storage services equivalent to surveillance of the contents of their interactions with each other?
If payment wasn't for specific websites or content then it wouldn't be a privacy problem at all. But how does that money get allocated fairly to individual websites without putting governments (or someone else?) in a position to say yay or nay to funding pornhub.com or alt.com?
> If payment wasn't for specific websites or content then it wouldn't be a privacy problem at all. But how does that money get allocated fairly to individual websites without putting governments (or someone else?) in a position to say yay or nay to funding pornhub.com or alt.com?
... which is a completely different problem than funding services such as those provided by facebook and the benefits in terms of communication possibilities those services provide.
One problem is the funding of infrastructure, another problem is the funding of the production of "content". Those happen to be intertwined a lot at the moment, but that's not a technical necessity, but rather a result of the current business models that build on surveillance (and the creation of de-facto monopolies and the subsequent potential for vendor lock-in).
It would be relatively unproblematic to fund development of infrastructure with donations or even tax money--think development of open standards/protocols and possibly reference implementations of those protocols. Then, it's unproblematic for most people to pay directly for the operational costs of processing and storing their own data (that is, computers running implementations of those protocols, storage devices, bandwidth). None of that requires anyone to have any access to or control over the actual content that people communicate. For that aspect, there is no need to have "websites". The central entity is completely unnecessary, you simply pay directly for the actual costs of the technology without the need to reveal anything about what you communicate or process or store.
As for the funding of content production: I think we have to have some possibility to pay for that anonymously if we want to keep democracy stable. We can buy newspapers with cash without leaving much of a trace in anyone's database as to which paragraphs of which articles we read. Letting terrorism paranoia destroy that would be a terrible idea. Also, as I have repeatedly said, I don't understand why you seem to be constantly equating advertisement and surveillance. Just because you can use surveillance to make advertisement more efficient, doesn't mean that advertisement without surveillance is useless.
>which is a completely different problem than funding services such as those provided by facebook
It's not a different problem at all. Facebook is not infrastructure. They provide content and communication services that many find useful but not everyone does. It's exactly the same with porn.
> As for the funding of content production: I think we have to have some possibility to pay for that anonymously if we want to keep democracy stable.
Good luck with that. So far it hasn't worked on digital platforms and I fear it never will. There is very little governments pursue as vigorously as cracking down on unexplained money flows. They will stamp out anything that reaches mass appeal with an iron fist.
Also, making people pay for everything puts poor people at a great disadvantage.
>Also, as I have repeatedly said, I don't understand why you seem to be constantly equating advertisement and surveillance.
That's right, I have been doing that. Mainly because the link between advertising and wanting to know as much as possible about potential buyers is very strong indeed. Advertisers will always be highly motivated to find ways around any ban. It's unenforceable.
You may ask why I am so confident that governments are able to enforce a ban on anonymous payments and so pessimistic about them enforcing a ban on ad targeting. It's because there aren't that many payment networks and they are highly regulated in the first place.
Financial sector regulation is extremely deep. You can't regulate millions of service/content providers worldwide in the same way. It would take enormous resources and international treaties that take decades to negotiate. In the end you would have something that doesn't just enforce an ad targeting ban but also provides ideal opportunities for censorship.
If you enforce a ban only for the big US based internet companies, then others will will quickly take over their ad business. If Baidu and Yandex can show us personalized ads but Google and Facebook can't, then ad money is going to Baidu and Yandex and they are going to subsidise mobile phones and other free offerings. It's not going to be long until we're back to square one in terms of privacy.
But I do think it is possible to force the biggest internet companies to not use some particularly privacy hostile practices like real name policies, browser fingerprinting and a couple of others. The goal should be to let them target ads all they want, but allow people to unlink their real identity from that ad targeting profile.
> It's not a different problem at all. Facebook is not infrastructure. They provide content and communication services that many find useful but not everyone does. It's exactly the same with porn.
I am not sure what you are trying to tell me here?!
Also, Facebook provides content? I don't use them, so what do I know, but from what I understand, Facebook only provides the platform for others to publish content and to communicate through, don't they?
> Good luck with that. So far it hasn't worked on digital platforms and I fear it never will. There is very little governments pursue as vigorously as cracking down on unexplained money flows. They will stamp out anything that reaches mass appeal with an iron fist.
I am not sure what your point is here?! Government will try to expand surveillance as much as possible, therefore, we should accept surveillance? I mean, I agree that it might not be easy, but I didn't say that it would be easy, but that there will be bad consequences if we don't, whether easy or not.
Also, perfectly legal possibilities to pay anonymously online for small amounts do exist, such as paysafecard.
> Also, making people pay for everything puts poor people at a great disadvantage.
Making people pay for everything and surveillance are not the only options. You can have donation-funded stuff, you can have publicly funded stuff, you can have advertisement funded stuff, and also, you should have a social security system that provides for people's ability to participate in culture. None of that requires surveillance.
> But I do think it is possible to force the biggest internet companies to not use some particularly privacy hostile practices like real name policies, browser fingerprinting and a couple of others. The goal should be to let them target ads all they want, but allow people to unlink their real identity from that ad targeting profile.
What do you mean by "real identity"? The name on the birth certificate? How does that limit any of the potential abuses of the data that's being collected? Or what else do you mean by that? Which data specifically should they not be allowed to collect?
As for the ban being unenforcable: Well, I am not that sure. Sure, the US cannot regulate what tracking Yandex does. But you sure can regulate what business a US company can do with Yandex, and Yandex won't advertise for US businesses for free.
Also, there is no need to be 100% effective. The problem is mass-surveillance, not that some company might know everything about a bunch of people. As far as the abuse potential is concerned, regulating 80% of the market probably is more than good enough. The problem, after all, is a concentration of power. A few million companies that make up the remaining 20% of the market are most likely not a concentration of power.
As for the supposed connection to censorship: How is that supposed to work?!
Not just that, it's also about weighing the restrictions on personal freedoms against the extent of negative externalities. In some cases the protection of personal freedoms is absolute. The ban on torture is a precedent for that.
>Now, there is actually precedent for banning consenting adults from sharing private information in order to ensure the survival of the democracy [...]
It's a precedent only if it's comparable, which it isn't.
You are overstating your case. There is no reason to believe that letting an algorithm parse some of your data for clues as to what you might want to buy makes democracy fall apart.
What is likely to have an effect on democracy is when governments analyze data for political reasons. Where that happens, people will simply stop volunteering so much private information. They don't need to be banned from doing that.