> With the announcement of federated learning of cohorts [0], it allows Google to use their massive amount of information they have on you already, as well as their dominance in AI to cut out the easy way of tracking people, and make it harder for anyone else to personalize ads.

yes, yes yes yes. This proposal is that:

* the web have no individual tracking

* the browser, instead uploads our every activity across all sites into the cloud

* the cloud/browser-operators (google, firefox, microsoft, opera, brave) take these troves of big data & apply machine learning to generate cohorts

* the browser then places us into these cohorts

this would, as parent post suggested, absolutely "cement google as the best ad tracking company". it also makes it radically harder to create a new browser. Brave would have to become not just browsing software you run on your computer, but a cloud-service, gathering reams of data like Google, generating cohorts out new big data systems. how is Lynx/elinks supposed to become compliant with this new proposal? how is anyone other than the already existing giants of the world supposed to do this?

bold bold bold plan organize & own all the worlds data here. privacy on the web, but none in the web browser: brave new world Google. you madmen.