So, what's the move for search engine's these days?

Duck duck go is a wrapper on bing that censors itself.

Google is The Empire from Star Wars.

There was that little paid search engine that kept double down on AI.

Is there anything else?

@ajroach42

@thelinuxEXP was wondering exactly this, it's a mess 😞

(vid is at https://tilvids.com/w/3nfn5RgLU6b5J13rBS71HG)

Google Search sucks, AI is everywhere: what should you use?

PeerTube

@ajroach42 I use maapl https://maapl.net/ which is far from perfect. As I understand it, it's a classic "meta search engine", which aggregates the results of multiple engines.

I find it okay-ish.

maapl

SearXNG — a privacy-respecting, open metasearch engine

@ajroach42 Got me to double check what's up with kagi.com at this point, and in 2024 they became a public benefit corporation. But I think they're also the doubling down on AI one you mentioned. So... yeah, I too wonder what the still other options are at this point.
@ajroach42 I’ve started to use Qwant, which shows up as an option on Vivaldi. I don’t know much about it, so it may be as evil or compromised as others. https://www.qwant.com/?l=en
royaume d'israel – Qwant Search

Fast, reliable answers and still in trust: Qwant does not store your search data, does not sell your personal data and is hosted in Europe.

Qwant

@ajroach42 I've been using Kagi and it's been pretty great

yeah they've done some doubling down on AI but I find you can use it without all that - but I totally get not wanting to go with them because of that

@bds I don't want to signal to them that I am comfortable with their AI stance by giving them money.

@ajroach42 fair enough!

thanks for putting the question out there though, learning about a few other options I haven't heard of

@ajroach42 @bds yeah, I'm currently running down a subscription with them, but looking for other options for when it runs out because of how all-in on AI they've gone. Also their results are getting just as filled with slop as everywhere else, and that was one of the biggest selling points for me at first. I've been getting great mileage out of adding "before:2003" to filter it out, but that's not really a sustainable approach long-term.

I've been experimenting with self-hosting SearXNG, and it seems promising. Whoogle and YaCy also look worth trying (and are both easy to set up on my existing YunoHost server), but I haven't actually tried them yet.

@ajroach42 mojeek.com is a funky little UK-based-webcrawler search engine that I use as a *supplement* - it's pretty bare-bones compared to modern search engines (you still gotta use key words, no phrasing-as-a-question here) and it doesn't have the same breadth of or even necessarily up-to-date sites (some of the sites indexed are no longer extant) but it makes up for it in finding me stuff that would be buried pages back on other search engines, if listed at all

but again, not my *go to* search engine, more of a furthering-curiousity machine

@ajroach42 my order these days is: Searxng - Mojeek - Startpage /ddg as a last resort
@ajroach42 currently trying out Ecosia, based in Germany.
@ajroach42 I guess it depends on whether you want mainstream stuff or the niches you enjoy? If the latter, things like marginalia etc. could probably work, but for general stuff I still find DDG is the least worst option, though they seem to be trying hard to change that...
@ajroach42 If you want even more niche than marginalia, there's the project @amin is working on...

@ddlyh @ajroach42

Mojeek is perhaps the biggest of the independent engines. As mentioned Marginalia and Clew fill similar spaces, but with Clew I don’t crawl large sites like Wikipedia and I down rank pages with ads and trackers. I also collect page size data and display it though I don’t currently use it as a ranking factor.

I’m finishing up on a huge refactor of Clew right now.