I hate that search engine degradation is what’s lead me to use AI more. Instead of searching past pages full of 8 ads for a waffle recipe, I ask Copilot or something: “Give me a basic waffle recipe”.
So much computation to go back to what the web used to be great at.
I knew Google started ignoring double quotes for required text years ago, but I found out yesterday that it doesn’t even think “site:xyz.com” needs to be followed.
I was researching something and saw some Reddit posts. Clicked below it to view results from Reddit and a third of them were other websites.
Google has always respected my double quote and site: searches. Please share a screenshot of it borken, I looked online and don’t see examples. If you have the time for a silly little thing :)
There are literally tens of thousands of examples of them ignoring any and all of their operators on Reddit and Google help. You can find them easily if you look.
Fortunately, Google Search has a special operator for that: quotation marks. Put quotes around any word or phrase, such as [“wireless phone chargers”], and we’ll only show pages that contain those exact words or phrases.
Caveats:
Quoted searches may match content not readily visible on a page.
Quoted terms may only appear in title links and URLs.
Snippets might not show multiple quoted terms.
Quoted searches don’t work for local results.
I would be ticked if quotes didn’t work. My screenshots do show them working.
An example of them appearing to ignore quotes came up. When they pull this, I can ignore the results below the error/red line:
Im a massive proponent of FOSS, But I have not heard a single sustainable FOSS model for maintaining free search engines. It just takes so much capital to operate.
I think a paid model is much better than a privacy disrespecting / ad driven one.
FOSS and paid are not mutually exclusive, but Kagi is not FOSS and of dubious transparency/trustworthiness.
Also Kagi is not operating a search engine, but a search aggregator mostly dependent on Google. They don’t need much upfront capital to operate.
An actual search indexer competitive with Google is too expensive to be profitable without (tens of) millions of paid users or hundreds of millions of free ones (i.e. bing and maaaaybe yandex?).
True google alternatives are therefore only going to come out of big capital (MSFT), or less likely a government (EU?) funded company. There might be an argument to be made for decentralized search as well, but the only actual contender in that field right now is a crypto thing that probably relies mostly on bing/google. Still, a decentralized open indexer may actually make some sense in theory.
I’m liking $0 SearXNG: lots of instances if you don’t host your own (for max privacy I think)
Germany/Spain hosted instance with all the checkmarks (Vanilla, IPv6, 100% uptime): https://searxng.site
If you fancy paying for a decent cause, don’t see a problem with the paid option sometimes suspiciously mentioned on Lemmy. Free trialing it saw a pleasant experience.
Sorry, meant it as a reply to each segment of your sentence, to clarify what I think we like and don’t like:
Lemmy doesn’t like [astroturfing], but [Lemmy likes] great [search]
I haven’t seen anybody say Kagi’s search itself is bad! Oh, I should have mentioned some don’t like the idea of paid search period. That’s another complaint.
Overall positive impressions from many users here, is what I see. “Lemmy doesn’t like Kagi” is somewhat of a mischaracterization I think.
Kagi has a really neat feature that if you phrase your search in the form of a question and add a question mark to the end of it, it’ll summarize all of the top results and give footnotes to the pages that it evaluated. It saves me tons of time!
I hate that search engine degradation is what’s lead me to use AI more. Instead of searching past pages full of 8 ads for a waffle recipe, I ask Copilot or something: “Give me a basic waffle recipe”.
So much computation to go back to what the web used to be great at.
I knew Google started ignoring double quotes for required text years ago, but I found out yesterday that it doesn’t even think “site:xyz.com” needs to be followed.
I was researching something and saw some Reddit posts. Clicked below it to view results from Reddit and a third of them were other websites.
Google has always respected my double quote and site: searches. Please share a screenshot of it borken, I looked online and don’t see examples. If you have the time for a silly little thing :)
There are literally tens of thousands of examples of them ignoring any and all of their operators on Reddit and Google help. You can find them easily if you look.
The examples I’ve found fall into “caveat” territory.
From the adware company blog:
Caveats:
I would be ticked if quotes didn’t work. My screenshots do show them working.
An example of them appearing to ignore quotes came up. When they pull this, I can ignore the results below the error/red line:
Further discussion:
I can’t reproduce but I wanna! (Prolly not kids though)
They ignore any and all of their operators if there is more money to be made by doing so.
I know Lemmy doesn’t like it, but Kagi is really great
Lemmy doesn’t like it for a reason.
Im a massive proponent of FOSS, But I have not heard a single sustainable FOSS model for maintaining free search engines. It just takes so much capital to operate.
I think a paid model is much better than a privacy disrespecting / ad driven one.
FOSS and paid are not mutually exclusive, but Kagi is not FOSS and of dubious transparency/trustworthiness.
Also Kagi is not operating a search engine, but a search aggregator mostly dependent on Google. They don’t need much upfront capital to operate.
An actual search indexer competitive with Google is too expensive to be profitable without (tens of) millions of paid users or hundreds of millions of free ones (i.e. bing and maaaaybe yandex?).
True google alternatives are therefore only going to come out of big capital (MSFT), or less likely a government (EU?) funded company. There might be an argument to be made for decentralized search as well, but the only actual contender in that field right now is a crypto thing that probably relies mostly on bing/google. Still, a decentralized open indexer may actually make some sense in theory.
Right, but nobody hates google because ofits results. They hate that its privacy invasive.
If I wanted to pay to talk to people, I’d go to a therapist.
Lemmy just likes shitting on popular things to feel superior
You keep on Kagi’ing
Lemmy loves Kagi. At least this Lemming does.
Astroturfing bad
Good search good
The former is unconfirmed to be sure
I’m liking $0 SearXNG: lots of instances if you don’t host your own (for max privacy I think)
Germany/Spain hosted instance with all the checkmarks (Vanilla, IPv6, 100% uptime): https://searxng.site
If you fancy paying for a decent cause, don’t see a problem with the paid option sometimes suspiciously mentioned on Lemmy. Free trialing it saw a pleasant experience.
Bruh, you think Im a bot?
Responding to the human who typed:
Explaining many Lemmings seem to like it, and attempting to explain why some may bristle at its mention
That’s all :)
You said it was astroturfing, implying I was astroturfing.
Sorry, meant it as a reply to each segment of your sentence, to clarify what I think we like and don’t like:
Lemmy doesn’t like [astroturfing], but [Lemmy likes] great [search]
I haven’t seen anybody say Kagi’s search itself is bad! Oh, I should have mentioned some don’t like the idea of paid search period. That’s another complaint.
Overall positive impressions from many users here, is what I see. “Lemmy doesn’t like Kagi” is somewhat of a mischaracterization I think.
Totally fair, sorry for the misunderstanding
Stop being reasonable and fight more for our amusement!
JK, good on you guys for being civil.
I friggin love this site
Hope you have a nice day & weekend ‘round the corner
I just tell AI to google stuff for me and link me to the best results…let it wade through the ads and spam.
Kagi has a really neat feature that if you phrase your search in the form of a question and add a question mark to the end of it, it’ll summarize all of the top results and give footnotes to the pages that it evaluated. It saves me tons of time!