This is an unfair characterization of the reasoning in the article. The main point isn't whether or not a service shows you ads. For example, the suggested e-mail provider, Lavabit, shows ads to certain classes of users.
Rather, the point is whether the service sells your information to advertisers, or mines your information-- including the contents of any messages you send via the service. Often this data mining is done in order to target advertisements at you, but the targeted ads, per se, are not the point.
For example, the main objection to Google Search isn't that they use the information to target advertisements (it even admits that doing so may allow Google to serve you more useful ads). Instead, the stated rationale for using DDG is "if you're worried about the amount of data collected on your browsing habits..."
Others on this discussion have pointed out that a lot of people are willing to make that exchange (trade off a little privacy in exchange for a good service). That's fair. But the article is clearly written for those who want to limit the inspection and selling of their personal information, not those who want to avoid advertising.
Oddly, the second part of your post shows that you understand the distinction. But I think any conflation of the two (advertising and privacy) is in your reading of the article, not in the article itself.
That said, in my reading of the article, the point isn't that targeted ads invade privacy more than untargeted ones. Rather, it's the act of mining your personal information that is the invasion of privacy. The fact that your information is read in order to target ads is almost beside the point.
Here's a concrete example from my own life (and now I do bring in my own feelings). Shortly after my wife found out she was pregnant, we started getting flyers in the mail for baby-related products. As nearly as I could figure at the time, the information could only have come from two sources: (1) my credit card (invading my privacy by telling companies I was starting to buy baby-related products); or (2) Facebook (invading my wife's privacy by selling information she posted for her family and friends to read).
After reading this article, I see that a possible third source was Gmail (invading my privacy by scraping the content of messages I wrote to my friends and family).
Now I understand people who say "That's OK, you got a free credit card in exchange for them selling your buying habits to companies." And I understand people who say "That's OK, your wife got a free Facebook account in exchange for them selling her posts to companies." And I understand people who say "That's OK, you get a free e-mail product in exchange for them reading the content of your private messages."
What I don't understand is any notion that it's not an invasion of privacy, because I might want to learn about baby-related products.
To put it in even more concrete terms, here are three forms of advertising for baby-related products:
(1) I see a billboard advertising infant formula.
(2) I buy an issue of "New Parent" magazine and see an ad for infant formula inside.
(3) I write an e-mail to my brother saying "Guess what, we're having a baby!" and five days later a sample pack of infant formula shows up on my doorstep.
Each of these advertisements is progressively more targeted. The first and second do not invade my privacy, while the third does. And the thing that bothers me is not that the advertising is targeted. It's that the advertiser is even aware that I'm a potential target.
I notice you say you bought baby-related products with your credit card. This provides a more likely explanation: statistical analysis of consumer buying habits by retailers (doesn't even have to be the card company) can reveal many sorts of information.
The particular example of pregnancy has been well-covered, see http://www.nytimes.com/2012/02/19/magazine/shopping-habits.h... and http://consumerist.com/2012/02/17/target-figures-out-teen-gi... .
As for your examples: I'd say that while #1 and #2 are targeted advertisements, #3 is not an advertisement, it's straight-up espionage.
> A targeted ad network needs to log/know more about you,
> in order to actually target the ad.
An untargeted ad network needs to know generally the same things, to mitigate fraud and inform advertisers about the size/demographics of viewers.People seem to believe an ad network that selects ads randomly is more trustworthy or less privacy-damaging than one that selects ads based on the current page or current viewer, which is a belief I consider unreasonable.
> Come on HN, you know how this works. If a service doesn't show you ads, and doesn't charge a monthly fee, how do you think they make money? Either they're not, so they'll go out of business when their VC funding runs out, or they're selling whatever they can glean about your habits to third parties.
I've said this before but I'd love to see some people doing not-for-profit versions of major services. Some advertising, but just enough to pay salaries and costs but not return 10x to investors (which you wouldn't get in the first place). No founders cashing out with $1bn+ payoffs after IPO, no IPO, no funding rounds, just a bunch of employees that would be treated fairly and collecting a safe regular salary/benefits with no expectation of a huge payoff.
If only I had made my millions in order to fund such an endeavour...
Come on HN, you know how this works. If a service doesn't show you ads, and doesn't charge a monthly fee, how do you think they make money? Either they're not, so they'll go out of business when their VC funding runs out, or they're selling whatever they can glean about your habits to third parties.
---
Additionally, the article conflates two very different concerns under the banner of "privacy".
If you object to advertising, then many free services also have some sort of paid tier. Send then $5 a month and they stop showing you ads. Problem solved.
If you're concerned about third parties having access to your data, then your only practical options are to either host everything on a computer you own, or encrypt your data before it leaves your desk. Both options are significantly more complex than using free hosting, but can provide stronger guarantees about the safety of your data.
The advantage to encryption is that you can decide on a file-by-file basis whether it's important enough to protect. Maybe you're OK with Dropbox/Google/Box.net/etc being able to see pictures of your cat, but don't want them to see your tax documents or medical records.
Of course, the main problem with taking security seriously is that you are placing yourself into a very small, difficult-to-please market segment. Most companies will find it easier to simply ignore you, so they can make their products work better for people who don't particularly think about data security.