Brave was made by an alt-right crypto bro that resigned as the CEO of the Mozilla Foundation after everyone there called him out for opposing gay rights. Fuck that guy and fuck his crypto farming browser. Do not trust that guy with your data.
I've replaced everything else, including migrating multiple services from AWS, DO, GCP to Scaleway, but Kagi is the only thing I haven't satisfactorily replaced.
I'm using Ecosia as Qwant isn't available in my country, but it's just not the same. Something no one else seems to do is the domain rewriting, so that I can send Reddit links to Redlib, etc.
This infograph should probably also include that Kagi uses the Russian search index Yandex also. It basically says that Kagi is the best search engine and that it fits all the requirements, but puts Brave in a bad light due to their CEO. If ethics are to be included, I believe Kagis use of Yandex should also be included.
Looking forward to it, but might take a while before the English internet is indexed by them: "Starting in 2025, our new index will be added into the database pool to serve results in both the French and German language. We are starting small and in the home countries of both Qwant and Ecosia respectively. "
"Crossposting" my comment from the feddit.uk post:
Interested in a far-out-there project, that is still very early, lacking in manpower and money, may never pan out, but has some very interesting ideas and vision?
Check out mwmbl - a project aiming to become a truly FOSS search engine, with Wikipedia-like volunteer curating of results/training of the search algorithm, as well as volunteer scraping of the web, to build it up.
If you want to participate, their non-experimental, older and more barebones interface is easier for that:
I currently use it as my go-to "first search" engine, and if I can, I help curate the search results, and then switch to Ecosia for a second search if they were useless. That already helps in slowly training the algorithm, as well as changes the results in the index real time.
You can also support the web crawling and index building efforts with either a Firefox extension, or a CLI script.. Not to brag or anything, but letting the latter run on my server has netted them a pretty hefty increase in crawled addresses, without slowing things down here on my Fediverse servers too much.
THIS IS NOT A SUITABLE DROP-IN REPLACEMENT FOR ANYTHING YET. But if you love the idea, go ahead and check it out.