This thread was brought to my attention, and I think it’s worth answering. My name is Josep M. Pujol, I’m responsible for Brave Search. Prior to joining Brave, I spent 6 years building the eventual foundation of Brave Search at Cliqz.
Brave Search does not censor. The fact that some expected results aren’t returned is not the result of censorship; it’s the result of search-engine-development being hard, and forever a work in progress.
There is some confusion between censorship and bias. Brave does not practice censorship; there is no intentional suppression of results due to politics or ideology[1]. Regarding bias, we do have “biases”, but not the type being discussed in this thread. There is no bias towards liberal or conservative media—we don’t even have information regarding the ideological leanings of sites across the Internet.
The strongest “bias” is towards popularity, but there are many other influencers to consider as well[2]. Popularity of course does not equate to truth—everybody knows that—however statistically speaking it works in most of the cases. Where does it fail? On controversial topics where the opinion of the majority might not be correct. Search engines tend to favor the most popular content, provided that the terms of the user’s query match the content with confidence. In the majority of cases, different views and opinions will be represented in the results. But in some cases, one view may take the whole page. It’s as simple as that, Occam’s razor.
Couldn’t Brave Search detect certain cases, and make sure the results are balanced? That’s easier said than done. As of today, we do not have the technical capability of doing that. But hear me out, we do not want to have it either. That would actually be dangerous, and it would open the door for real censorship. The moment we start using whitelists and blacklists we stop being a service and become the ministry of truth.
How does Brave intend to compensate for biases introduced by popularity? By enabling users and communities to correct existing biases with their own with the upcoming Goggles project[3]. In our opinion, this is the best approach to not becoming a single source of truth, as well as ensuring that users and communities remain in control of their experience and exposure in the future.
I hope this response helps to shed some light on the discussion.
[1] Suppression of results is limited to the following cases: CSAM content (child pornography), DMCA notices on copyright infringements, and of course, compliance with relevant laws for the countries in which they apply.
[2] Some domains you find to be missing (or underrepresented) could be more difficult to crawl than others. Some pages you judge to be more relevant could be more difficult for our crawler to parse. The list goes on and on; there are hundreds of reasons why certain domains/pages rank lower (or higher) than what you think they deserve. And none of them are ideological—at least not in Brave Search.