Previously, I used DDG but pretty much reflexively g!'d every query.
I’m surprised I normally stay within the $5 plans limits, but when I’m regularly above I’ll gladly bump to the next tier.
You seem like someone who has used a number of search engines. How does it treat sites Google blacklists like nsfw sites or porn sites or torrent sites? Does typing in movie/book/game torrent bring up listings the way Google previously did? How does it treat older non-mobile friendly sites? Is the image search comparable to Google?
My biggest complaint with Kagi is sometimes when I search for something the top result is a pinned domain that maybe hit the right keywords but mentally to me it was unrelated to the search I was performing. That's a pretty good problem to have at the end of the day.
That might be because Google won't admit when it doesn't have results responsive to your query. It will just fill the results page with nonsense instead of saying "I couldn't find any results".
Google has gone so far into the “well we think you really meant this” zone that it refuses to show me pertinent results unless I start adding quotes and modifiers to my search.
DDG is veering there.
Bing for now is the most literal. I search something and it gives me results that have tighter matches to my terms
It would be interesting to see some sort of study about what the difference between those two groups actually is.
It appears that they've "improved" search over time to cater to the lowest common denominator, in the process neutering all the functionality that I used to get precise results. So most people will have seen an improvement, because it no longer takes any expertise to use the search field well, but the skill ceiling is so low that there's no way to do much better than the mean.
I wonder if Google / DDG will start to notice this drop in searches or not.
Also, I find myself using GitHub search a lot, it returns actual code someone was using, which is always helpful.
Example: I wanted to know how tall Frank Lloyd Wright was. But if you google something like “Frank Lloyd Wright height” you’ll get a bunch of results about his buildings.
"I run every single day, how have my heart and muscles changed in the last x years."
"Can I pressure cook seitan in two sessions, so I cook it along with my soup as well?"
"The character from my book gets a little annoying in chapter 16, does that get better (no spoilers.)"
Either of these would have you wading through tons of random websites to collect the necessary information, and in the case of the cooking questions through a load of noisy SEO'd cooking websites, ugh.
Wgen asking the LLM I get a specific answer to my exact query. It comes with uncertainty (hallucination etc.) but either it's a topic where fuzziness is irrelevant or a search enabled LLM grounds the truth and gets rid of most hallucinations.
Either way, sifting through search results is not what I want to do if I want a specific question answered, that's a machine's job.
No idea if that's the case though.
https://circle.so/ (followed by a number of child entries and links)
https://go.circle.so/coaching/
https://guruscoach.com/circle-so-review/
https://help.circle.so/c/getting-started
https://redeemingproductivity.com/circle-review/
https://community.circle.so/c/tutorials/
https://community.circle.so/c/product-updates/introducing-li...
I would guess even if DDG protects your privacy, it would use information (perhaps search and or site history) the browser is broadcasting regardless to filter results it pipes to the end point.
For example I never see quora results (honestly it's been some years so I'd forgotten they existed still) as some members in this thread were annoyed with and paid to be done with them ... then again I'm cheap and would use search terms -site:quora.com
What happens if disable Javascript, cookies and use lite.duckduckgo.com
Also what happens if try the .onion address (can set exit node country in torrc)
Something like
tor -f torrc &
x=duckduckgogg42xjoc72x3sjasowoarfbgcmvfimaftt6twagswzczad.onion;y=$x;
socat tcp4-l:443,fork,reuseaddr,bind=127.0.0.42 socks4a:127.0.0.1:$x:443,socksport=9050 &
#!/bin/sh
read x;x=q=$(echo "$x&p=&s=${1-0}&dc=$(($1+1))&api=%2Fd.js&o=json&kl=wt-wt"|tr '\40' +);
{ printf 'POST /lite/ HTTP/1.0\r\n';
printf 'Host: '$y'\r\n';
printf 'Content-Type: application/x-www-form-urlencoded\r\n';
printf 'Content-Length: '${#x}'\r\n';
printf 'Connection: close\r\n\r\n';
echo "$x";}|sed w/dev/stderr \
|socat stdio,ignoreeof ssl:127.0.0.42:443,verify=0
# usage: echo circle.co video calls | 1.sh > 1.htm;firefox ./1.htm
# usage: echo circle.co video calls | 1.sh 22 >> 1.htm;firefox ./1.htm
In the meantime, you should also check other free search engines. DDG gets it's results from Bing. Startpage.com is another good one, it gets it's results from Google and just has a privacy layer. I recently switched to Brave Browser and have been using their native Brave Search & it has been working pretty well. They use their own bot to actually go out & index stuff (from what I understand).
There's also Metager.org which is a meta search that gets it's results from several different sources. There's also a few others but I really haven't found much use for them or just get poor results.
Check out searchengine.party if you want to try out the really esoteric search engines.
Fair. For me, Kagi and YouTube (which I'm ditching) are the only subscription services I have, so it isn't a burden.
Default search engines have a massive amount of inertia behind them. Anything that disrupts the Google near-monopoly is a good thing, in my not-so-humble opinion.
Edited to add: What sort of use cases do people have for the Kagi API? I was just curious about its use.
When Google was new, we used to play a game to see who would run out of answers first on a particular search term. Then that morphed into who got a porn link first, even though the search term had nothing to do with porn.
But now Google has gotten so bad, you might have to play to see who gets a reasonably correct answer first.
there is also LibreY, but my searches there fail more often than they succeed.
Seems to be fixed now, but it reminded me of the early search engine days.
Funnily enough, when I used dogpile.com, it returned relevant results
https://techcrunch.com/2022/08/05/duckduckgo-microsoft-track...
$ whois $(dig a duckduckgo.com @1.1.1.1 +short) | grep OrgName
OrgName: Microsoft Corporation
An analogy to your claim would be to say that Hertz drives their rented cars.
You can encrypt your data with your own key with trivial effort on all the major cloud providers.
Believing that this actually protects you from the provider is wishful thinking. What keeps them honest is that it's mostly not worth it for them to mine cloud customer data, but if they wanted to they absolutely could.
Indeed. Besides, there is a whole genre of research and practice on the introspection of virtual machines.
I’m talking about who holds responsibility for the actions taken with the virtual machines that Microsoft sells.
If I buy a VM from Microsoft and run a scam on it, Microsoft didn’t have control of that action. They don’t make my scam website or sell the data I collected. I’m the one responsible for scamming people, not Microsoft.
We can go down the hypothetical rabbit holes on whether Microsoft is running an incredibly sophisticated operation to memory dump all their customers’ VMs to try and steal data, but that’s basically beside the point.
Elsewhere it would be hilarious if anyone would think so, but on a technical forum called "Hacker News" it's downright embarassing.