I find it ironic to see this brought up as the first example of problems, when this is exactly what Google's FLoC intends to solve. The same FLoC that has so vocally been rejected by HN recently.
I find it ironic to see this brought up as the first example of problems, when this is exactly what Google's FLoC intends to solve. The same FLoC that has so vocally been rejected by HN recently.
I don’t see irony there!
And there are also a lot of design questions remaining that Google hasn't answered or is not sure about.
Eg. How will ad bidding work with FLoC?
Finally, and most importantly as it relates to this post - Google pays an enormous amount of money to employees to spend their time and intelligence on building better profiling tools. Building profiling tools is something ethically questionable because of the damage that can be done.
I don't work on FLoC, but why wouldn't bidding handle it just like any other signal? For example, here's where it is in the OpenRTB docs: https://developers.google.com/authorized-buyers/rtb/openrtb-... An advertiser could take it into account in considering whether they want to bid and how much.
I think I have an idea. So by visiting site A, A gets your cohort and now there is somewhere a "central store" or register. Using that A can now do ad bids for that entire cohort to reach others.
But where is that central store? Is Google running that?
Might also work fundamentally different.
Some entity might choose to run a "store" like you're describing, but there's no reason there would only need to be one.
It looks to me like Chrome is proposing a low-level feature that anyone can build more sophisticated products on top of.
Same here:
- Google and Facebook have turned ads into 24/7 surveillance of everything people do
- Now Google roll backs some of it by saying, "with FLoC we can only surveil groups of people 24/7"
- we are expected to praise them for finding a "solution"