We don't aggregate anything into Federated Learning, all opt in (off by default) machine learning is local to the browser. Chaum blind signature protocol (Privacy Pass standardized this) to confirm.
I know it's fun to repeat misinformation on hacker news out of ill will, but many people have verified our claims from open source audits, network audits, and more. Comments like yours just look sloppy or even malicious, even on HN.
Brave doesn't do any of this cloud surveillance based ad-tech. We leave ads off by default, but when a user opts in, all the ad matching is local to the browser against a catalog that's the same for a large cohort by region and human language. Ad impressions are confirmed by a Privacy Pass variant protocol. Users get 70% of the gross.
There isn't a great category for what Google is, it already won its own "google it" verb. But among many other things, it is a huge ad-tech player. Brave is small, user-first, privacy-by-default, and ads are opt in. See the difference?
https://www.computerworld.com/article/3292619/the-brave-brow...
> Brave scrubs sites of ads and ad tracking, then replaces those ads with its own advertisements, which are not individually targeted but instead aimed at an anonymous aggregate of the browser's user base.
Sounds an awful lot like Google's https://en.wikipedia.org/wiki/Federated_Learning_of_Cohorts, no?
btw I don't know anything about Brandon Eich, but I still would never use a crypto browser