Hacker News Reader

They don’t want to know us, but they appear to have very few limits on what they’re willing to do to sell ads. So far we’re basically counting on our interests and theirs being coincidentally similar, I would not bet on that in the long run. Better to handcuff them before they decide that doing something incredibly unseemly is necessary for ad sales.

> Physical storefronts ((...)) with time they added cameras which helped understand customer behaviours.

R. Doisneau, a French photograph, may have in a way be a precursor https://www.christies.com/lot/lot-4572128/?intobjectid=45721...

> you can infer a lot about an identity from seemingly unimportant things like browser resolution

Oblink: EFF's "Cover tour tracks" https://coveryourtracks.eff.org/

> these companies don't really want to know you, they just want to ensure they are able to serve relevant ads to someone like you.

"Relevant" is their PR-speak, but really it's just whether you're in the desired target audience. If I target an ad that discourages people from voting to vegetarians or people who like Fox News, that ad is not necessarily more relevant to them.

One thing I've wrestled with with the rise of online news and its effects on physical newspapers is how much I miss certain things about the physical newspapers. I don't miss the physical format, but I do think the old-school paper newspapers were much more enjoyable to read than most online equivalents.

At some point I realized that one major issue is that advertising in many of the paper copies was based around content area: if I went to the performing arts section, for example, it would be filled with ads for performing arts events. I loved this as it was actually useful and informative to me. I went to that section looking for performing arts, and that's what I got.

In online news, though, if I go to a performing arts, I don't get informative, unintrusive ads for performing arts events in my area, I get bombarded with random ads for things unrelated to what I'm looking at. Even if, say, earlier in the day I was looking for shoes, I don't want to see ads for shoes if I'm browsing performing arts, I'm interested in performing arts.

What you're talking about is a broader observation about identification of individuals per se versus patterns of interests and behaviors. However, I'd argue that a major failure of online advertising (with very important exceptions, including Google, DuckDuckGo, and many other places) is the recognition that what matters for ads is interest at any given moment, and not interests at any other time. I suppose someone might say "but a good ad is something that gives you what you are interested in even if you might not recognize it" but this is really difficult to get right, especially given that my interests in a given moment can shift from minute to minute.

If I'm moving from, say, shoe shopping to, say, performing arts, I'm deliberately moving my attention away from the former to the latter. Showing me ads for shoes is something that's specifically going against my current attentional goals. It's like saying "hey Honey, I'm done in the kitchen and am going to go into the garage to work on something" and then having some random stranger show up and pull you back in the kitchen.

This seems to be a fundamental screwup with a lot of online advertising: the failure to recognize that I'm functionally a different person from moment to moment, and when I move from one page to another there's a reason for that.

Email surveillance is maybe going even further in a worse direction, in that it's even more decontextualized and time-independent. Part of the brilliance of Google search ads, and things like DuckDuckGo, is that they catch you exactly in that moment when you're looking for something on a specific topic. Newspapers and everywhere else needs to take better advantage of that paradigm. Show me what I'm looking for now, don't take a shotgun guess at what I might want based on what I was doing in the past.

> Email surveillance is maybe going even further in a worse direction, in that it's even more decontextualized and time-independent.

Are any major email providers still selling ads targeted by the content of messages?

It's doubtful that any of us are in a position to know if they are or not.

Let's assume, for the sake of argument, that 100% of major email providers have stated they do not sell ads based on email content.

Next we have to either: take their word for it or have the means to verify their claims.

Taking their word for it is difficult because many major email providers have a spotty relationship with honesty. This issue of honesty is not necessarily very different from other large corporations and in truth might be a factor in what made them a large corporation in the first place.

(As First Baron Thurlow is claimed to have said: "Did you ever expect a corporation to have a conscience, when it has no soul to be damned, and no body to be kicked?")

And so we would instead need the means to verify the claims of these major email providers. I'm unsure of how to reasonably do that.

►Perhaps allow Qui Tam claims for privacy issues combined with a statutorily defined "cost" for each false claim instance?

Qui tam allows, for example, private citizens to file suit against bad-actor govt contractors in the name of the govt. The "whistleblower" then receives a share of recovered proceeds.

Here, if a statutory "cost" was defined for every false claim related to using the content of email messages (say $1 per message) then this might provide a way to help verify that the major email providers are being truthful in what they claim regarding their use of content in messages.

Email providers would know their employees are on the lookout for a big payday and might honor their public promises. And if they don't, a few large qui tam lawsuits would quickly get their attention (or drive them into bankruptcy).