[1]: https://support.apple.com/en-us/HT205223

This shouldn't exist in your operating system at all.

But yes, looking at ads on the Google Play Store recently tells me I made the right decision to switch to an iPhone.

It's pretty obvious Apple doesn't have financially valuable properties on the web but they have plenty in their application ecosystem.

In Apple's view Apple News is how people consume news, not go to the web where the user experience isn't custom tailored by Apple. Apple News is barely different than Google AMP that everyone rages against, try to get someone to share a link to a piece of news from Apple News and it will come with the ask to subscribe or download it.

So in Apple's ideal plan almost every place where today ads are displayed will be inside Apple News.

And hey, I wasn't aware there exist another App Store where people can buy applications or in-app purchases on iOS.

And yeah you can turn FLoC off too both today and when it will be in production.

Lastly, when did the argument move from "advertising bad, tracking data always bad, targeted advertising terrible" to "if Apple does it then I'm good with it"?

when did the argument move from "advertising bad, tracking data always bad, targeted advertising terrible" to "if Apple does it then I'm good with it"?

It'd be a lot cooler if you could respond to my actual argument without namecalling and putting words in my mouth. For one thing, it's rude; for another, it doesn't contain much of a counter-argument.

You said the distinction between what Apple does and Google does is 'thin'. The distinction is Apple does this in two specific services of their own, one of which is completely optional and in both cases, the cohort bucketing and tracking can be turned off. Google is putting it in the browser people use to do just about everything on the internet. A roughly similar thing would be Apple adding cohort bucketing to Safari. They haven't. These seem, to me at least, some significant distinctions. Why do you think they aren't?

So your argument is maybe valid (I don't think so, tracking is tracking and this isn't just contextual targeting) only now, as time goes by and Apple accumulates more power your argument is back to being invalid again.

I have an Android phone: I use a non Google browser, non Google app store and non Google messaging app.

Nevertheless it seems a formidable addition to a browser user fingerprint.

Which reality is better is up for debate but they won’t be doing both for the same client.

Fingerprinting already slows down websites, and websites still use it extensively. The cost is paid by the users and nobody cares enough to fix it.

I really have enough projects, TODO lists, interesting things to buy to last for lifetime of 250 years.

Why I would want ads likely to influence me? There is basically no chance that ad will show me useful, worth using object or service to buy that I would not discover anyway.

It would be just another scam more effectively targeting me.

Advertising is psychological manipulation at its worst, and we'd all be much better off without it.

Hyperbole is unbecoming.

Off the top of my head I'd say grooming someone for sexual abuse is a significantly worse form of psychological manipulation. I speak from having experienced both, and only one has had a catastrophic effect on my life.

I've never had this experience, between Google, Facebook, and Amazon. Every "targeted" ad I've been exposed to has either been for something I already have, or for something that's insufficient to my needs.

Because of this, and because I dislike the invasiveness and anti-privacy of modern advertising, I started keeping a log of all of the ads I can remember. Then before I make a purchase, I lookup the product on that list and pick a different company to buy from. If you, as a company, use invasive advertising, I will not be your customer.

What do you mean irrelevant data? I am much better at choosing a product that suits my needs than Google and friends are. The only irrelevant data is the one provided by the ad companies.

> the company that paid for the advertisement likely won’t even miss your sale.

If that's true, then why do you think they advertised to me? Why would they possibly choose to advertise to people who they weren't interested in selling to?

This would all be ok if the tradeoff was a strong signal to the seller that advertising has a negative impact on sales, but I’m positing that the impact of that one sale is negligible and well within the noise threshold of an extremely noisy process. For all you know, if the seller notices at all, they might conclude that they didn’t advertise to you enough.

However I’m totally ignoring any utility you might gain from a principled approach and if it works for you, then more power to you.

Very true, I've considered this many times. What it comes down to is that I value companies that prioritize investing in their product rather than their advertising, so on the whole I'm more likely to get what I want if I stay away from products advertised to me.

> This would all be ok if the tradeoff was a strong signal to the seller that advertising has a negative impact on sales, but I’m positing that the impact of that one sale is negligible and well within the noise threshold of an extremely noisy process. For all you know, if the seller notices at all, they might conclude that they didn’t advertise to you enough.

This argument kind of ignores the whole underpinning of economics. Let's consider a different situation and apply your argument. Let's say a company sells a toothbrush that doesn't clear plaque away very well. Over time, as I use this toothbrush, I'll find that I'm getting cavities on my back teeth. So I stop buying that brand of toothbrush. Under your argument, my doing that would be a weak signal, because it's just one sale lost for X dollars spent on advertising. The seller doesn't notice, and keeps on pumping out toothbrushes, and lots of people get lots of cavities. I know this is not analogous to the situation you're talking about, but it's still a plausible situation, and if your argument is correct, it should be correct across most situations it's applicable to.

And thanks, I do enjoy your confirmation of my power.

But overall risk of getting persuaded to buy worthless products or scammy "investments" seems to not be worth discovering real gems.

Overall I would prefer both scams and genuine offers to be as far as away from my interests as possible.

Obviously, I would prefer actually useful offers to be close to my interests and to have no scams/worthless trash/malware/porno in my ads, but it appears to not be an available option.

But in reality, ads are always "relevant to seller/advertiser". Seller is almost never interested in selling you anything but what makes most profit from you. Eg. A decathlon might be better cost/benefit ratio wise, but you will always get an ad for Nike, because Nike is paying for the ad, not decathlon.

Such is the lie that was fed to us, and most of us accepted it. Tragedy of the commons indeed.

One day, you may have a need for a product, and someone who previously solved your need can’t afford to make the product anymore because only a small percentage of people realized it was a solved problem.

If blog or similar can finance part of their cost with Ad's, why not.

Sure on slow internet it's a somewhat different matter.

Anyway Ad's okay, but tracking for me is spying and should be treated like that, i.e. it should be illegal.

There are ways to have personalized ads without it, like local learning which then selects to get an add for a specific topic, with that a side could still track which ads you got, but it would be much less useful as FLoC, and then you add additional steps to even further decrees any chance of tracking.

Companies still get implicit feedback by what adds get selected more then others.

People can explicitly blacklist annoying (or offensive) adds making sure they don't see them ever again and if enough do so in turn making that app not seen much more at all etc. etc.

But the FLoC cohorts are WAY to small/identifiying to not allow you to be tracked fairly easy with it. Just combine it with other identifying aspects of browsers (there are a lot) and I wouldn't be surprised if it's often allows a 100% unique identification.

And given that you likely wont be able to "ad-block" FLoC this makes it way worse then the status quo.

It's fair to believe that ads are 'ok' - it's also possible that people in general don't want them. Hosting low-traffic websites on a smartphone with cellular data is just-about-feasible now.

Rather than inventing very technically clever ways to continue on our existing path, as Apple and Google are doing, I think it'd be pleasant to simply move beyond the advertising model of content on the internet.

If asked, I don't believe people would want to be tracked or have ads mixed into their content.

Could we make it cheap enough to host a personal blog and/or wiki on a smartphone for, say, $10 a year?

The economics would be very different as sites scale up their traffic (compare with, for example, The New York Times).

I think the (relatively) small number of high-traffic sites would be able to find revenue streams to support their hosting - perhaps those would be where subscriptions would be effective.

This may of may not be good, but I’m a bit concerned that destroying the adtech industry may have ripple effects that end up being worse than some middle way like FLoC

Yeah, in that world, ads would be ok with me. Sure they're pointless because I won't click on them - but not really worse than having a blank spot in the page because of an ad blocker.

While I have no problem with the concept itself of advertising, I do have a problem with modern implementations of ads. Almost invariably, they are loud (vocally and visually), intrusive, obnoxious and almost coercive in their use of language. They are intended to dumb down the target audience, rather than enrich their targets' lives.

I do have a problem with everything that attacks my ability to think freely.

That's just how "new" works.

Websites can move to more efficient and lighter methods of delivery to cover the absence of advertising revenue. Anyone skilled who works in the advertising industry is freed to offer his skills to better human endeavours. Everyone wins.

I don't want to be rude, but won't they just end up losing their livelihoods? If the advertising industry goes, most programmers will see their salaries drop hugely. Even those not working in advertising will see a drop due to the flood of supply.

(I do think advertising is net positive and prohibiting it would be a bad idea: https://www.jefftk.com/p/effect-of-advertising)

Consumer habits will just become more unpredictable- and small companies could compete with big companies that spend billions on marketing.

Even if you were not directly influenced by advertising, you probably were indirectly through word of mouth.

With harmful incentives removed, this would encourage websites to focus on good writing and less on being 'content farms' to drive up engagement for the advertisers.

All I want from most sites is the text. Not adverts, trackers and those damnable autoplaying videos that jump out and follow me down the page. I use the reader mode in most cases. Why is it necessary to use reader mode if not to cut the useless cruft? Why can't the web just be like that?

They can handle some theft, but if literally everyone stole their candy, the store would close down and nobody can have candy.

Also, the store isn't freakishly stalking me and noting down all my habits whilst sharing what it has collected with its friends, or indeed anyone who will buy the data.

Besides, candy is impossible to replicate in the same way that data is. Theft of physical objects is a materially different act from advert blocking and avoidance. I'm not obligated to view adverts in every public space because they fund the local government. I'm not denied access to the cinema because I arrive 10 minutes after the start time to avoid the adverts. Or back in the day when I taped TV shows and fast-forwarded the adverts, had I a responsibility to view them?

They're useless noise that contribute no value to anyone beyond whoever pays for them, and therefore I find it a moral responsibility to limit my exposure to them and restrain the harm they do to others. And curiously, I am much more content and much less impulsive in my spending habits since I installed uBlock.

One might argue that you are then not entitled to view the content on ad-supported pages as well.

I use an ad blocker too because many ads are just horribly intrusive, but I honestly can’t blame any page circumventing ad blockers, nor do I believe that I am somehow morally entitled to an ad-free, compensation-free browsing experience.

Narrowing all of this down to CPU time misses the point entirely in my opinion.

> They're useless noise that contribute no value to anyone beyond whoever pays for them, and therefore I find it a moral responsibility to limit my exposure to them and restrain the harm they do to others.

They‘re literally paying for the content you get to view for free.

Then they shouldn't serve the page in a way that it can be viewed without also viewing the ads. Ad-blocker-blockers are a thing, and I am happy to close that tab and never visit the site again when I see one.

> They‘re literally paying for the content you get to view for free.

Hopefully someday they'll be forced to find a business model that doesn't include destroying the privacy of their website's visitors.

This is true, and websites are entitled to try to detect and block users who use adblockers. And users are entitled to use better adblockers that are harder to detect.

Nobody frames, say, taking a bathroom break when the ads play during a football game this way.

It's a bad metaphor, a better metaphor would be reading a journal and covering ads with post it notes.

Many retail stores do, actually, through wireless tracking, cameras, and/or purchase history. They will buy and sell consumer data through the likes of Acxiom.

It would be great to have a ublock equivalent for the physical world.

No one is under any obligation to accept advertising. Only since the modern web have advertisers and content providers made this an argument on ethics.

We could all go around in circles like this for hours.

I'm not really sure what crowd of us you think claims that ad-blocking is "morally responsible". I certainly don't see it as a moral issue at all. I just don't want so see ads, period. I think they're psychologically manipulative garbage, and I don't want them in front of my eyes, infecting my brain.

If someone wants to publish something behind a paywall, I'll pay if I think it's worth the price. If someone wants to put up an ad-blocker-blocker and refuse to serve that content to me, that's fine too; I'll live without their content.

If a server is going to send me bits over the wire, I am going to decide how I view those bits, and which bits I do and do not want to see.

I suggest you are incorrect to assert that "no one is entitled to my CPU time." I agree, however, that advertisers do not have that moral claim right.

> I suggest you are incorrect to assert that "no one is entitled to my CPU time."

Very well, please elaborate. I'm always up to delve deeper into reason and philosophy (time-permitting of course).

I'll start with ~~a humorous point for an autistic person~~ the worst kind of pedantry. An object cannot be no one's and mine at the same time. If I own it, then it isn't no one's. To the degree the CPU is indeed yours, then it is someone's CPU. Obviously, there's some conversational implicature embedded in your claim here, and that's what we're here to clarify.

For the record, I'm on your side. Your overall argument is something everyone needs to hear. So, I'm not here to destroy your assertion, but I think it should be weakened. There's a lot packed in there.

I take your proposition to mean, roughly, that the CPU in your possession is something only you have a cluster of moral claim rights to (which corresponds to a cluster of moral obligations of others to you regarding that CPU), with no possibility of overriding reasons, caveats, or provisos. Now, you could be speaking about political claim rights, but I don't think that's the case here (correct me if I'm wrong). Is that a fair interpretation, and do you wish to maintain that? Can you think of any exceptions? Are there any adjacent possible worlds in which that is not the case?

Perhaps it is time for an advertising model to die out as the default.

I will not render your ads. I’d be happy to put that in the first http request to your site. If that means no site for me, then quite honestly 90% of my browsing is a waste of time anyway.

They encourage people to believe their self worth is linked to the things they own. That a persons status in society is somehow associated with how much money they have to spend.

Who's pretending? Ads naturally and quite effectively price-discriminate, and so they do make internet content cheaper for people with less money.

You may have problems with excessive consumption, but many people in developing countries do not. They desperately need the free content (educational, informative, and otherwise) that the current internet model provides them.

Well funded public broadcasters like Australia ABC provide news and entertainment content free. (both text and video)

Well funded public libraries can provide a wide range of alternate content.

While I love libraries and public broadcasting, I learned to code as a poor kid through library books, they are not the full solution and they never provided free, instant, and any time international communication for people.

If websites really find that ad blockers are killing their revenue, they will switch to a paywall model, and maybe micropayments will finally happen (or perhaps a service that allows users to subscribe to large numbers of websites at a time). Whether or not that is a good thing depends on who you ask.

These reductionist approaches don‘t really help with understanding ads (and communication in general).

However, I do run an ad blocker and I'm completely fine if every company that relies on ads for income goes out of business. In fact, I'd appreciate it.

This sounded good when I first heard it, but by now I understand 'relevant' to mean 'tailored to optimally influence my behaviour', eg showing me articles about the futility of voting if my political preferences are likely to lean a particular way.

It says it's about respecting the privacy of users better but in practice it fails very bad at doing so the only thing it archives is allowing Google to disable cookie based tracking and with that makes it harder for 3rd party trackers to be google independent (it not harder for 3rd party trackers if they use/abuse the FLoC Id and generay there are many ways to track users which are not cookie based so especially when it comes to the very bad offender of invasive tracking it doesn't make a difference or becomes even easier, but some of the smaller google ad (and analytic) alternatives will have it harder).

A philosophy blog is the first thing that comes to find but even then that can have ads for universities, textbooks, quirky t-shirts, etc.

A choice of least worst is not necessary.

You have no idea how hard people work. Ad browsing pays the same no matter how hard people are working.

I’ve never done anything with the crypto part. I just use them as a clean, ad-free browsing experience. There’s ad blockers for other browsers but brave is the easiest for me to use.

There’s over a million web creators accepting bat just from brave [1] and while that’s not everyone or even a big percentage it’s not “few web creators.” If you don’t want to give to web creators you’re free to cash out and donate cash to charities. So that’s a plus that’s not possible with other browsers.

Note, I don’t use the crypto feature in brave, I’m just annoyed by people spreading FUD about stuff.

[0] https://www.coinbase.com/price/basic-attention-token [1] https://batgrowth.com/

That just seems like hyperbole actually, and from a biased opinion or vested interest.

Brave seems to handle privacy very well considering the choices available, and has made a number of good privacy decisions from what I can see after watching browsers evolve for 25 years. I have switched to using Brave and installing it for those I support.

The crypto thing is 1) easy to turn off, 2) an attempt at a possibly better business model where users aren't just sold with it. Regardless I simply turn the crypto setting off.

Who knew that ship designers could buy special plugins for CAD environments? Even though I could have guessed it (In fact, I made this example up), it’s still fun and interesting to be surprised.

where is the better part? is it the fact that this makes fingerprinting easier? or that floc makes more data available to advertisers than cookies?

For example if I visit a website that follows trends in computer hardware... advertisers may pay to put ads in this site to PC parts, etc. If I visit a site about popular camping locations.. advertisers can purchase ads that are relevant to camping,gear, travel etc.

The very idea that we need a tracking identity that monitors and identifies behavior from site to site is ridiculous.

If you get rid of user tracking, anyone or their brother can sell ads, and then why deal with Google?

There isn't really any hard bound on how much information could be leaked via these algorithms.

> There isn't really any hard bound on how much information could be leaked via these algorithms.

With ~33k cohorts, there is literally a hard bound of just above 15 bits per visit. That's still theoretically a lot if you have some other stable identifier, but practically speaking most users on most sites will have an identical cohort and it will drop off rapidly as cohorts stabilize into groups appropriate for ad targeting.

Barring the other considerations I've mentioned in the comments here, that's still immeasurably better for individual privacy than ~infinite bits per visit from stable third-party cookies.

"This API democratizes access to some information about an individual's general browsing history (and thus, general interests) to any site that opts into it. This is in contrast to today's world, in which cookies or other tracking techniques may be used to collate someone's browsing activity across many sites."

Can the user detemrine whether a site has opted in without sending an HTTP request the site (and thereby recording the access in her browsing history).

"Sites that know a person's PII (e.g., when people sign in using their email address) could record and reveal their cohort. This means that information about an individual's interests may eventually become public.

"As such, there will be people for whom providing this information in exchange for funding the web ecosystem is an unacceptable trade-off. Whether the browser sends a real FLoC or a random one is user controllable."

Can advertisers determine whether a FLoC sent is real or random.

I can imagine some FLoCs if sent would trigger ads that were undesirable, perhaps damaging to the user's reputation, embarassing, etc.

As described, this system compels users to accept targeted advertising (what Google and Facebook want every user to do). There is no opt-out. There is no such thing as non-targeted ads. Users of browsers that support FLoC can either send real FLoC or random FLoC, but they cannot send no FLoC. If they choose randomised FLoC, they might receive ads targeted to cohorts with which they do not want to be identified. As the next paragraph states, the FLoC when compbinaed with other data can indeed be used as an user identifier. Choosing "randomised FLoC" might mean being publicly identified with undesirable cohorts. For example, IP address might become associated with an undesirable cohort.

"A cohort could be used as a user identifier. It may not have enough bits of information to individually identify someone, but in combination with other information (such as an IP address), it might."

"The expectation is that the user's FLoC will be updated over time, so that it continues to have advertising utility. The privacy impacts of this need to be taken into consideration. For instance, multiple FLoC samples means that more information about a user's browsing history is revealed over time."

"Second, if cohorts can be used for tracking, then having more interest cohort samples for a user will make it easier to reidentify them on other sites that have observed the same sequence of cohorts for a user."

"A cohort might reveal sensitive information."

"Some people are sensitive to categories that others are not, and there is no globally accepted notion of sensitive categories."

"It should be clear that FLoC will never be able to prevent all misuse. There will be categories that are sensitive in contexts that weren't predicted."

"A site should be able to declare that it does not want to be included in the user's list of sites for cohort calculation. This can be accomplished via a new interest-cohort permissions policy. This policy will be default allow."

Default is no privacy. Nice.

"... a site can opt out of all FLoC cohort calculation by sending the HTTP response header:

Permissions-Policy: interest-cohort=()"

Proxy config:

http-response add-header Permissions-Policy "interest-cohort()"

"We will analyze the resulting cohorts for correlations between cohort and sensitive categories, including the prohibited categories defined here. This analysis is designed to protect user privacy by evaluating only whether a cohort may be sensitive, in the abstract, without learning why it is sensitive, i.e., without computing or otherwise inferring specific sensitive categories that may be associated with that cohort. Cohorts that reveal sensitive categories will be blocked or the clustering algorithm will be reconfigured to reduce the correlation."

Well, this appears to negate the claim that no browsing history is sent to the browser vendor. To conduct experiments and develop this system, Google needs to collect and process user browsing histories.

Why? I would expect a browser extension to be able to override document.interestCohort to throw an exception?

(Disclosure: I work at Google, speaking only for myself)

If that's an actual way to access it, "undefined" will have to be acceptable due to browsers that don't support it.

And undefined would still have to be handled anyway, as browsers without support wouldn't have the attribute at all (at least for several years after it was accepted, for everything to get the new feature).

The explainer indicates it's a Promise: "cohort = await document.interestCohort()" -- https://github.com/WICG/floc

One reason to make it async could be to give browsers the option to gate access behind a permissions prompt in some circumstances.

> And undefined would still have to be handled anyway, as browsers without support wouldn't have the attribute at all

Definitely. There are also browsers that are unlikely to add support (ex: Brave, as we're discussing)

But - if the regular, more accurate/personalized, mechanisms work in Chromium - why use the less-accurate ones?

2. The more accurate ones are going away.

[1] https://ai.googleblog.com/2017/04/federated-learning-collabo...