dialtone

It's not just clear, it's Cristal Clear™.

Here's a review of all the relevant literature on the topic: https://docs.google.com/presentation/d/1PKHVtO6hgwBJS1vafLyv...

I think parasites are those that want the web to be simply free and have free access to whatever they want regardless of the infrastructure cost of running the internet. It's the actual definition of parasites to feed off of a host organism.

Pay and subscribe to all the sites you read or video sites you watch and let's see how many sites are left that you read. Incidentally I subscribe to a lot of online news sites and video sites to avoid ads where they bother me or to support the journalism, but not everyone can afford this, especially smaller publishers.

Because they aren't lying... The source code of this stuff is available in the open and you can verify it. There's no FLoC in the merged code in the main branch.

All good points, but they don't change the conclusion I wrote though, you effectively further specified the lengths to which Google is going to prevent any potential bad correlation to sensitive categories. Of course there will be correlation of some sort, they aren't random groups, but so long as you can't tell sensitive categories, you can only observe that a group seems more interested in a particular advertiser products. You can't tell their browsing history, just their interests.

Regarding the trial, yeah I know that origin thing is up, and the minutes for yesterday's W3C IWABG meeting aren't going to be out until next week, but the team working on this stuff at Google said that the trial up there is kind of fake and the code isn't merged in Chrome yet and won't be until 91, at some point I have to trust the engineers working on this stuff more than the rest.

[edit]: To be clear, I don't mean that there's no experiment going on, I just mean that it's not the browser running it, that the removal of 3rd party cookies as a way to opt-out likely is due to the fact that they may be relying on server-side computation to do this experiment rather than the browser, I wouldn't see why the browser would care about cookies to track your history, it's already stored in the browser.

I don't think you are being constructive or bringing arguments that support your thesis so I'm going to stop here to reply.

Only during the transition, when all is said and done there are no other identifiers of merit, and since this one can change pretty frequently it doesn't really identify much.

One day I receive a request that says you are group 12345, in 1 hour you are 12346 instead because the algorithm recalculated your FLoC... Not sure what anyone can tell about that without knowing the group ids meanings, which likely not even google will know since this is going to be a clustering algorithm anyway.

It's currently part of the conversation at W3C, advertisers aren't particularly happy about FLoC either, they'll need to build a lot of ML to correlate the FLoCs with their customer's campaigns, one by one, until they find the group that performs. They'll need to develop techniques for this, for example using machine learning and train their models. That's not how this stuff is typically bought today.

In particular they aren't happy also about the 3rd party cookie comparison because it just makes no sense, not sure why the PR focuses on that given that this is minimum 3-4 orders of magnitude less precise than 3rd party cookies, targeting a FLoC is going to be like targeting a zip code, hardly THAT privacy invasive.

This thing explicitly does not send your history to anyone so I'm not sure what you are commenting here.

Nah you don't need to decompose it. The only thing that matters is if a given FLoC has more probability to see its members click on ads, or convert on ads, from a given advertiser. This can be discovered with ML with enough data/observations. Nothing else is really necessary.

Those parasites pay for the vast majority of the internet you likely use and benefit from.

You could do that, technically there are other APIs like Trust Token API to protect against malicious clients, but one client in a giant cohort is unlikely to cause much damage anyway so it doesn't matter... And this is anyway going to be machine learning driven so unless you poison all of the FLoCs a lot it may end up being discovered or the FLoC simply not considered performant by a specific ad campaign.

How? How would fingerprinting use this? It's formed by data used in the past 7 days and the bigger the space of FLoCs the more likely it is that it changes frequently, for example every day if not more frequently.

The fingerprinting would fail to recover the same code at least once a day unless you always visit the exact same sites every day and never visit a new one.

fingerprinting needs a stable state from the browser to work, not something that changes arbitrarily.

No, you can't really do any of that. A single FLoC ID, while having access to no other identifier, is not gonna say anything about your past history to anyone. It's just useful to correlate to actual advertising campaign performance.

Most traditional adtech, used to targeting attributes, is quite puzzled at how to use this at the moment given they'll all need to invest in machine learning because these IDs aren't interpretable.

The specifications of this stuff are all in the open.

This isn't how FLOC works though, the EFF persists in the misunderstanding of this technology.

A FLoC is a single ID per browser and it's not human readable, the algorithm to determine this will run in the browser so your data will never leave it. There's no real way for someone to "reverse engineer" the ID, it's just a random set of bits not "people-who-like-shoes".

Last, and I may be wrong here given this announcement, FLoC isn't part of Chrome until Chrome 91 in june, at least I think this is what the Chrome team said Tuesday at the W3C meeting so I'm not sure what this refers to, but it shouldn't be the browser side.

Have you considered paying the $13/mo to not see ads?

One day ycombinator will be full of people that talk after informing themselves, but that day is not today.

The Chromium Edge doesn’t support 3rd party cookies already and thus isn’t really subject to targeted ads like Safari or Firefox.

Now please continue with your reasoning.

I don't do mobile advertising but this is just incorrect... Most of Google's revenue comes from Search and that has nothing to do with IDFA.

Then how would third parties with no access to first party data compete when there's no more third party data would be something I'm curious for you to explain.

I’m not sure how you can say that. The first examples of anti trust action have been about vertical integration and ownership of critical infrastructure by a market player not allowing entrance by others because it’s not easy to lay down more railway or more phone cables or telegraph cables. That’s the case for the web advertising situation.

As much as I agree with the core of what you are saying, it isn't AdMob that is the issue. It's that Google controls both the marketplace, the pub-side and the buy-side, as well as the browser and the devices used to experience ads. Apple is now aiming for a similar level of control in the ad space very blatantly using their ownership of the platform to do what they want to power themselves up.

This vertical integration in the space is what drives the anti-competitive behavior, if the Google buy-side were a separate company that happened to acquire AdMob there would be no real issues as they don't control the whole stack.

FB on the other hand sells a service and lets companies use their APIs to integrate as well as competing with them via their internal sales team.

These companies can do this only because you are forced to work with them because they own such a large part of internet traffic, not because they offer particularly good services or have better technology or prices.

I'm really not sure that makes you right. It's a fact that FAN is 2% or so of revenue for FB. And it's also a fact that FB, with their analytics/ads SDK and all the data they ingest from their own applications, already owns a very important data source for their targeted ads. These are plain as day facts. If you ever bought ads, you'd know that FAN, for the typical buyers, represents a tiny fraction of spend on FB (less than 5% typically). Then journalists can write what they want, as much as I like Arstechnica and pay for it, that's just plain wrong and countered by facts. It's also frankly funny how little journalists know about ads when in most cases it's the source of their income.

Furthermore, it's absolutely hilarious to think that the most valuable company in the world, with enough cash reserves to be able to almost buy FB outright, would somehow be willing to walk back their core value message because their, allegedly, value adversary asks. Apple made the change because they would have cratered their app economy and they themselves weren't going to play by their own rules, FB maybe played a role because $2B are a lot of money for the free apps, but they aren't a lot of money for FB. It's really hilarious to think that Apple gives a crap about what FB writes on their site, specifically when it's about privacy. Apples cares about its ecosystem, not about what FB says. Totally hilarious.

You deliberately choose to ignore all of the studies that disagree with you except the one that you agree with that has very obvious flaws. Here's a summary of the literature[1]. Personalized ads have become popular because they perform better for the buyer, and that makes buyers willing to pay more to show them to their customers. Buyer goals align with conversions. The history of advertising online started with contextual, there is zero evidence to support the claim that revenue will only drop 4% when any company submitting bids for contextual vs personalized ads can show you that the CPMs they are willing to pay can easily be 2-3x if not even higher for high worth users.

[1]: https://docs.google.com/presentation/d/1juu6UBguR7ru1Rhfyor9...

This would have been true for this change as well, clearly if Apple chose a different path means that it's not as clear cut as you imply here. And yeah contextual and stuff is nice but will simply not drive the same revenue for publishers.

Facebook didn’t complain... they said they would shut off FAN which generated a whole of $2B in revenue for them in 2018 when they make $80B this year.

They were telling their customers to find an alternative monetization strategy since this wasn’t going to be worth it for them.

I've been working with tens of thousands of them for 15 years. To name one area: last click attribution being the primary way people measure web performance is a prime sign that they are at least confused with how performance works. Advertising measured with last click tends to focus towards people that have visited your site less than 4 hours earlier, hardly something that is incremental to the sales of a business, and basically all research and experiments that you can run point to this, but it doesn't change the fact ad buyers all believe last click is the one true way. And to show that it's not just me saying this: https://research.netflix.com/business-area/marketing-and-gro... .

GDPR and CCPA give you control to stop tracking altogether, not allowed to drop cookies to you. I fail to understand how that would be different.

There's a limit to how many campaigns you can be a part of as well since one could display an arbitrary set of campaigns to you and understand who they targeted once again. In any case "show this guy an ad" like you quote, won't leave signs or campaign IDs in the apple framework unless there's a way to track that user explicitly. And tagging people that viewed ads is not anywhere in the scheme for apple. As an aside, attribution of online ads is vastly more sophisticated than you paint it and it involve statistical and causal models. https://dl.acm.org/doi/abs/10.1145/2020408.2020453 https://pdfs.semanticscholar.org/eaa4/875ae2376d681e156dcc3b... https://research.netflix.com/business-area/marketing-and-gro...

You can 100% easily disable tracking with Google and Facebook. They both are required by GDPR and CCPA to provide the tools to do that. In both cases you can go in your account and disable targeted ads. Do it.

FB isn't really alarmed... The majority of their revenue comes from their own app which they'll continue to be able to target you on because people give data to it willingly. They won't be able to use their partner mobile apps to display ads to, that was a decent business but obviously will become less valuable, but it wasn't that big for them. They are signaling to their customers that things are changing and to expect that they'll change dramatically. FB however remains with your real email address on file and lots of accurate data about you, with their SDK and other mobile apps their accuracy in data collection won't change a whole lot.

Yeah the DNT header didn't work due in large part to middle men (like routers or anti-virus tools) flipping that bit without asking the user in automatic, or again broad decisions, such as from Microsoft, of flipping the default of that header without asking users. Obviously if one side chooses to just do what they want, the other side is free to do what they want as well. This whole situation needs people to sit down and talk these issues through, like it's happening inside the W3C. It doesn't need a few companies to strong arm whatever is better for them financially (Apple makes no money off of current ads, but they'll force themselves in the picture with their attribution tool and will get more paid app revenue out of this, pretty clear incentive).

I can buy ads generically on a site, but what happens after the user clicks on the ad or just views the ad and then buys the product? How do you keep track of this when you can't measure the effect? The only way is to measure some form of before/after analysis of your conversions with or without the marketing campaigns... Too much noise to be able to do that on the web.

There is a reason why the news paper industry moved in the direction that they did. Or do you think they were forced to move this way? Measurement on the web can be more precise and allows to understand the impact of revenue, it's also vastly more distributed than magazines and TV channels of which there's only thousands or so for each while there are literally millions of websites. There's too much noise and fraud on the web to be able to make marketing decisions with little tracking. The W3C is working with Google on establishing what it would take to accomplish this stuff without sharing data, this isn't the approach apple took with its platform in plain disregard for all the apps that run on it, but than again at this point this shouldn't surprise anyone.