Recently I have noticed that when using DuckDuckGo to search, it's returning unrelated 'news' in my results. For example, I just searched < circle.co video calls > and in the first page of results got links related to the war and Speaker votes. This has been happening generally now when I search using DDG. (I'll put a link in the comments to an image)
I use DDG but actively of thinking switching to Kagi: https://kagi.com/ -- it's a paid service but I can justify it based on the amount of search queries I run everyday. Has anyone switched and not looked back?
I've been using it exclusively since beta and am completely satisfied! The key improvement for me over other services is the ability to boost and block domains—it comes with a decent set of filters for the worst offenders (I've never seen a GitHub issues or StackOverflow mirror), but it's two clicks to block a domain that I personally don't like seeing. Pinterest, geeksforgeeks, and a bunch of other low-quality high-SEO sites all get permanently blocked for me, which is a huge quality of life improvement.
I’ve been paying for kagi as well and am really satisfied with the results. One improvement tho: the image search is not there yet. I often have to use Google (or better: Yandex) for that.
Oh, totally. It's a sign of just how much better my search life has been that I'd completely forgotten about Quora when I was trying to remember what I'd blocked.
Kagi user. I think it's pretty dang good. After comparing its results to google for the first few days, Kagi really does appear to be as good by default, and you can make it better by up/downranking various domains. Its quick answer feature is also fantastic (if perhaps not as quick as its name would suggest).
Like siblings, I switched and never looked back. I used !g with DDG often, more as the years went on tbh, but with Kagi the only time I use !g is when the search results end in one page, which is rare and almost always the Google results are noisy trash. Maybe 5% of searches I check Google and 10% of those actually return something useful that Kagi didn't. On balance with how much better the other 95% of queries are, first time I'm excited to search for things.
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.
> with Kagi the only time I use !g is when the search results end in one page, which is rare and almost always the Google results are noisy trash
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".
I use a number of different search engines for different purposes. Yandex offers sites duckduckgo doesn't. Duckduckgo offers sites Google doesn't Google is good overall and at image searching.
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?
I'm on the same boat. Used to g! most of my queries with ddg. Since switching to kagi, I've fallen back to Google for only a few queries, and I think only one or two had better results
Once they brought back the unlimited searching, I jumped on board and couldn't be happier. I only had to use "!g" to get live baseball scores. The quality is fantastic and I cannot recommend them enough
I've been using Kagi since the beta, and used DDG for years before that. I would not switch to Kagi for search result quality. It is fine, but has never blown me away. I would switch to it for privacy, for not being the product, and for certain UX features they support but Google never would (like being able to map reddit.com -> old.reddit.com, or being able to block or downrank content from search spam sites like Pinterest)
I’ve been thinking of switching to Kagi too so this is good to hear. My biggest concern is that they’ll eventually sell out and flip on their current privacy policy as we’ve so many other tech businesses do.
I switched to Kagi a month ago, and it's fine but the experience of using it doesn't live up to the hype IMO. You don't really "feel" the privacy benefit, so it's just a paid search engine that mostly lines up with Google's quality.
kagi is amazing for non-technical searches. i'm going to go back to it once they figure out location-aware searches and improve results for non-technical searches.
Annoyingly, I find that Bing is the only “good” search engine of the 3 major ones (Google, Bing , DDG)
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
There's a fascinating split there. For a lot of people, Google's guesses seem to be mostly on-point. For a lot of other people, they're awful. (I'm in that latter group).
It would be interesting to see some sort of study about what the difference between those two groups actually is.
For myself, I feel like I was really good at getting precise results from Google years ago, but it took a specific way of asking that isn't natural to most people. I learned that way of asking, and was able to consistently find exactly what I needed, but the average user couldn't.
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.
This tracks with my experience as well. For a long while, I tried different ways of formulating my Google queries to try to get good results again, but I never was able to find the magic formula.
Google will often exclude ignore terms in my searches -- specific error codes and messages, for example -- making it pretty useless unless I add quotes around every term. Even then, it doesn't do a good job of ignoring content in sidebar divs (supposedly related content), which is even worse.
I have also noticed that when I search for something obscure, when pages related to my query are exhausted, there are search results related to my IP region (e.g. weather sites, wikipedia pages, linkedin pages mentioning or related to that city). These results appear even when the region switch is off or set to another country.
This happens to me as well. I wonder whether they just ignore that selection now, or if Bing somehow is able to tell the user region despite DDG not sharing it with them.
After months of arguing against using ChatGPT, I now find myself using it for probably 80% of the things that I would previously have searched for. I also find it's much different, instead of spending 3 minutes trying to reword my DDG search, I have a fairly useful conversation learning things. I keep my skepticism up as I get more specific and detailed, but I somehow started trusting it, and haven't missed search at all...
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.
If you like ChatGPT for search, I highly recommend checking out Perplexity, it is a search engine that uses the same model but can provide quite good information with source citations.
I've started subbing to Perplexity.ai recently and it's been great. All the benefits of interacting with an LLM with the grounding truth injected from search results. It's not perfect and sometimes falls on its face, but I use it way more than regular search these days. As a paying customer you can switch the underlying model to GPT4 or Claude2, which is nice.
"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.
I am not the person you’re replying to but what is great about ChatGPT is how it parses meaning and doesn’t get confused by a large amount of results with similar wording but different meaning.
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 have never seen anything even vaguely like this, and can’t reproduce it on that exact query. I’m more inclined to attribute what you’re observing to a temporary glitch of some kind, if true (not that I have any reason to disbelieve you).
I actually hesitated to post this because I want to support privacy-oriented tools and appreciate the education DDG has offered to lay people. But lately this seems to happen for me all the time. Perhaps they have testing segments and not everyone is getting the same algorithm, or maybe there’s some other factors like location, device or browser type, who knows?
Neither can I. I tried the same query and the first few results are exactly about circle.co . There are some news articles near the end of the first page but they're still relevant - about making video calls.
If DDG is truly privacy-oriented (as in, not using information about the user), then what inputs is it using to return different results to different people? Shouldn't it be returning based on a neutral ranking?
If I search the same - I don't even get sponsored ads with DDG with your search terms. I generally get with my own searches one or two sponsored ads at top of page before the real results begin.
This is the main links from the first page of results I get.
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
I use custom filters, small C programs, that extract the URLs, descriptions and other info from various SERPs, e.g. DDG, and output SQL or simple HTML. This makes it very simple to exclude/include results based on strings in domains, paths or descriptions. I like to mix results from different search engines into a single SERP comprised of simple, regular HTML. No nonsense.
Weird. I get different results. I also get the "Speaker vote" thing, but I can sort of accept that because the words "circle" and "calls" appear in the title. Nothing about the war, though. Instead, there's a bunch of companies with a similar name or product. It also offers me to search to "circle.co" video calls.
Everyone in the comments is mentioning Kagi, but I can't afford yet ANOTHER subscription service.
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.
I feel you, but running something like Kagi requires massive amounts of computing power, and something has to give. It's either we pay for the hosting costs and growth, or we get an ad-supported mess, i.e. Google v2.
Happy with Kagi and its 2-person unlimited "family duo" account for $14/month.
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.
Thanks for posting this. I had the same suspicion. Makes me wonder if they've teamed up with one of the other big tech companies to supplement their results.
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.
What alternative for meta search engines are there? I know about DDG and SearX (and SearX-NG). I don't really care of the bangless search quality because most times I have a particular website in mind to search (and when I don't I alternate between Google, Yandex and Bing, depending on content)
4get.ca, but instead of mixing results it gives proxies results for a single backend scraper. Of the three source engines listed, I believe only Yandex is implemented.
there is also LibreY, but my searches there fail more often than they succeed.
The only problem I’ve had is that sometimes there are 0 results for something that should definitely have at least some results. It feels like I’m hitting some prohibited combination of words but I’m searching for something tech related, nothing like “tank man”.
StartPage is my default engine on the search box, I use Google only when SP fails to find something, which is rare but does happen from time to time. DDG has been great for years, growing steadily in quality, then just when it was about to become a lot better than Google, its reliability got a lot worse. I can't confirm what the OP writes as I abandoned DDG like 2 years ago, but wouldn't be surprised at all if it was true.
> 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.
This actually isn’t at all relevant to what I’m talking about.
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.
