@kagihq is the very interesting project for a paid search engine, without tracers and with an accuracy in identifying results such as to exclude all Google spam.
Those who believe that #Kagi's costs are too high, should reflect on a small detail: if Google lets all those searches be done "for free", who pays those costs? The answer might seem simple: "advertisers".
Yet this would be an incomplete answer: like saying that rain is caused by clouds!
In reality, those costs are paid by users, by being milked and letting Google extract their "value", a bit like in the human farm in Matrix...
We first heard about Kagi on the @lealternativewebsite (unfortunately, since then the prices have increased a lot, raising many doubts about the sustainability of the project) and recently Cory Doctorow also talked about it on@pluralistic
In any case, we are really happy that a service like Kagi's, effective and respectful of users' privacy, has landed here in the #Fediverse.
I’ve tested kagi and agree that the search results are great. What I don’t like is that it’s making anonymous searching impossible, since I have to be logged in to use it (or use my unique token as part of the url for mobile searches).
Ultimately this means to me that in a private window mode (or even logged out with a fingerprinting resistant browser) I do not have the same degree of anonymity I enjoy even when using Google, let alone DDG or others.
I like the idea of not being dependent on google, but exposing my entire search history to one single entity is not my answer of choice.
Thanks, that’s an interesting read.
I know that’s one person’s opinion and not a thorough research, but that’s still plenty of red flags.
I’ve used the 100 searches in the free trial, thought the search was fine, better than Google’s these days. The subscription is a bit steep so I held off, kinda glad I did after digging more into this.
Having what little employees they have also make a mac-only browser, AI stuff and email that their user base doesn’t seem to want is all a bit weird.
Buying a t-shirt factory (wtf) with the money they could have used to potentially lower the subscription, but decided to burn through it to give out free t-shirts. That just screams narcissism-driven to me.
Their vague statements on privacy isn’t convincing at all.
Some variation of “we don’t care about your data” isn’t in any way compelling evidence that you care about protecting the privacy of said collected data.
In my opinion they lack focus, commitment and conviction into what I thought was their primary mission at first glance: being a privacy-focused no nonsense search engine.
Although that’s probably on me for reading what I wanted to see between the lines and that never was their stated mission, which would explain a lot.
Aaah, now I get it. You guys made an account to post on. I saw the headline and I was busting my head how in the world a search engine would federate within the fediverse…
Mastodon posts tend to get funky when they federate, because Mastodon has this (annoying) norm of starting posts with mentions. So OP mentioned the community on the first line, which became the first part of the title - @fediverse. Second, it mentions Kagi as a tag instead of name, which gives the @kagihq. And then comes the first sentence as the rest of the title.
It's a great example of Lemmy/Mastodon interoperability working, but not being quite there yet.
Mastodon posts tend to get funky when they federate, because Mastodon has this (annoying) norm of starting posts with mentions. So OP mentioned the community on the first line, which became the first part of the title - @fediverse. Second, it mentions Kagi as a tag instead of name, which gives the @kagihq. And then comes the first sentence as the rest of the title.
I'm not writing from Mastodon, but from Friendica
It's a great example of Lemmy/Mastodon interoperability working, but not being quite there yet.
The mentions problem seems to be only for PieFed, but it doesn't happen with Lemmy
I think it should be reported as an issue to PieFed developers
I subscribed to the lower tier for a while, but I kept running out of searches early on every month, and the price of the higher tier is just not excusable. So I found myself adding the !ddg bang most of the time to avoid spending my Kagi quota.
And as good as Kagi is, it's still primarily a meta search engine, organizing results from the dominant actors. So it's not like the price is justified by them having to crawl the entire web themselves. Their own crawler, Teclis, is currently small web only and can probably best be described as an interesting project.
Instead of making search cheaper or more affordable, they spend subscription money on creating AI services and various other non-search distractions. Maybe that's good for some people, but I don't want that shit. I just want a good search engine at a justifiable price. And for that, sadly, Kagi fell short.
For me, it took me several months to use the 100 trial searches. Paying $5/month for like 20 searches isn’t worth it for me. Also just not into getting more subscriptions… wouldn’t matter if it was just 50 cents/month for the basic plan. If they had one-time-payment no-expiration option like 1000 searches for $10, I’d probably lean towards getting that.
I can only speak for myself, but I like to keep posted when important people and organizations are opening accounts on the Fediverse. :)
It's funny that Kagi included "fediverse forums" functionality in their search before they decided to join Mastodon - one would think the search functionality required a bit more.
I pay 5 euros for a VPS at Hetzner and self host my own SearXNG metasearch engine without compromising my privacy. Unlike Kagi, I can search without limits and without the AI stuff that even Kagi has started getting into.