Data on search engine market share is available, but I wonder what that looks like for Lemmy users in particular, who I would assume lean more technical than the average user, so probably use DuckDuckGo and alternates more than Google.
I use a mix of DuckDuckGo and Kagi. I’ll also use ChatGPT, which can be good if you’re careful to verify the answers it gives you as a check against hallucinations. It’s useful for short, direct answers without ads or SEO bullshit.
This article on Ars (and if you’re not a subscriber, you absolutely should be, as they are the best tech journalists out there) inspired the question: https://arstechnica.com/gadgets/2023/06/google-admits-reddit-protests-make-it-harder-to-find-helpful-search-results
Fucking Reddit. Enshittification ruins everything.
I don’t understand why lots of you answer with chatGPT. It’s not a search engine! And you shouldn’t use it like a search engine.
If you pay there’s an option for chatgpt4 that can use Bing to search. There’s also various plugins that can let it interact with all sorts of additional data sources. Not that you should use it like a search engine exactly, but it can be useful for search if you configure it correctly and understand that it doesn’t “know” anything.
Bing has gpt 4 for free, there’s a button for it on bing.com. I do pay for GPT Plus but there’s no web search option there for me, I have to use bing for that.
This is interesting, do you know where can I understand more how it is supported to work?
I can see a usecase for where you don’t know where to start or search with, and then verify with actual searches.
I recently used it to explain for a friend what is the difference between wheat and ale beer, and it gave a very good summary. With DDG I might not get a direct explanation and would need to read a few articles and then word them in a comprehensive way.
I think we’re making some confusion here.
If you need answer to a question, you can do it by searching for sources, using a search engine, or you can do it by asking a knowledge base (e.g. reading and Encyclopedia).
If that’s the usecase, you can successfully use a generative AI to obtain an answer (providing the fact you are willingfull to accept a possibly bad answer). You are not using the AI as a search engine.
If you need a search engine, if you need to search for places and sources into the web to access it directly, you should not use an AI, because it’s not how it works. It does not search and it does not provide you searching results, it “simply” generate answer based on training, Which is a completely different task.
- tenet ( @tenet@kbin.social ) 4•2 years ago
Except it IS a search engine and that’s basically all it’s good for. By its very nature all it can do is collate information. It’s the only thing AI is good at.
No it’s not. To search is a specific task, and generative AI can’t do that. It can fulfill some need that we are used to fulfill by searching the web, but this doesn’t mean it’s a search engine.
If you lost the key of your car and have access to an AI that can (sometimes) start your can without a key, you can be happy about it, but you still can’t say the AI searched the key for you. It can’t do it.
Edit: btw, we are talking about generative AI here. I’m not saying there isn’t and could not be a search engine that use AI to better its result.
Bing implemented ChatGPT with their own layer on top. It works like ChatGPT and can give some ChatGPT like responses, but it does so by showing the sources and links.
Therefor, i’m happy to say - I actually like this new bing and use it a lot.
BUT… i also use bing rewards to buy all my video games so it works there too :D
Yeah the new bing/chatgpt thing has its uses. It’s not my first choice, but I do occasionally use it.
They even got me to install edge on linux in order to get early access to it.
Bing is literally the only site I’ve ever opened in edge though. lol.
Maybe people mainly search for answers to simple daily life questions or something.
I guess, but it’s still not a search engine and I think it’s a bit problematic if that’s the usecase.
To be fair, some implementations like bingGPT and perplexity can provide some search engine functionality, since they actually find and provide links to the user.
We just had a case of lawyers getting fucker over by chatgpt making shit up, it makes things up all the time! I don’t understand why so many people use it like that and trust it like some oracle.
ChatGPT-4 can use a search engine (Bing) via a plugin.
Mostly duck duck go.
Same here. I know a lot of folks don’t like the results, but to be honest, I don’t find Google any better these days.
Agreed, Google has been ruined by seo.
Self-hosted Searxng. It’s shared to multiple people which kills a lot of the usefulness in Google or others trying to track my instance.
I tried this, but it kept saying ‘Engine failed’ or something on every other search. I never could figure out why. I might try again
Edit: Actually it was Searx I used. I’ll spin up Searxng and see if it’s improved
I had some issues with searx… Things are a bit better in my experience with searxng. Sometimes I still run into the error messages. But usually it’s my fault more than anything (server bogged down, too many requests/searches across all my users, or internet blips)… I just rerun the search a few seconds later and it’s usually good again.
Nice. I’ve stood one up now and it seems to be working nicely:
Awesome! Only thing I see wrong… I see version 1.0.0, I believe (and could be wrong) that they changed the versioning of it to a date-based system about a year ago… If so you might be running an older version. I run the docker version and I do in this specific case use the latest tag.
Same here
I’ve spun up my own SearXNG instance, and it’s the best search engine I’ve ever used.
Is this like when people say arch is fun?
Seeing as I absolutely hate using arch linux (sorry to those who like it I guess) I would say no.
There is even one script that does all the work for you, so it’s dead simple.
To compare, PeerTube was a bit more of a slog with many steps and a few chances to break things (the fun happened when I had to uninstall the official npm because despite the installer saying it would install 16.x, it installed 18.x, and that really upset PeerTube). I then had to use nvm, and that was fine until a step failed as I forgot to soft link my local npm binary to /usr/bin or something like that.
tldr: please for goodness sakes fediverse, make your setup scripts a once and done affair!!!
Kagi. Very happy with it. Best $5 it recently invested. Gives me much better results than Google and all the others.
How do you come by with just 300 searches per month? I tested the trial period and used up the 100 searches in just a couple of days
Yes, that limited number of included searches is my only criticism I have with Kagi. They are aware of this, and are trying to offer customers more searches for the same price by improving their costs. I am glad they decided to do this by reducing their costs and have decided to not go the road of monetizing their users by selling ads and customer data.
However, I try to use Kagi only for serious search requests. For other very trivial searches, I use Startpage. For me, works OK. But I hope that one day Kagi offers enough searches, so I can just use it everywhere as my default search engine without having to thinking about it.
@eight_byte @monotrox How do you differentiate ‘serious’ search requests?
I’m considering Kagi but I’m a very trivial person.
With trivial search requests, I mean stuff like entering the name of a company as a search term, where you could have easily just entered the direct URL in our browser instead. There is almost no benefit for using Kagi on this. Almost every search engine will give you the result you are looking for as the first search result.
Kagi is perfect people doing a lot of research in the web. I am a software developer. When I try to find solutions for programming related topics, Kagi gives me much better results than Google since they don’t show me all the ads and don’t do weird ranking stuff. Also, I am able to ask Kagi to only show discussion from public forums or even let them summarize the results via AI. Doing product research is also a lot more helpful with Kagi.
@eight_byte That’s a great answer, thanks.
I think I’ll give it a whirl and see how I get on. Search seems to have generally gotten a *lot* worse lately and I’ve been guilty of using ChatGPT to augment some of my searches for work-related stuff. Maybe Kagi is a better answer.
It should. As far as I know, ChatGPT is not connected to the internet and therefore doesn’t have access to recent information.
@eight_byte Oh yes, I mean I use ChatGPT more semantically, to help me figure out how to search for the right things.
Not sure how anyone gets away with 300 searches in a month, but I’ve been on a ~$10/mo plan (might be grandfathered) that has 1500 searches and I’ve never reached my limit. And I search a lot
Why didn’t they raise the limit or kept that for $10 subscription?
It’s a bit weird to me how much simple queries can cost. It’s basically amount to 10 searches a day for $5 subscription and about 33 searches a day for $10 subscription. Depending on circumstances, I would be searching a lot for very difficult to find information related to programming, so I could see $25/mo as a steep price to pay.
Another Kagi convert here. Switched over late last year after hearing about it on the Python Bytes podcast. Thanks to doing it that early I’m on the “Early Adopter Professional” plan, so 10 bucks per month for 1500 searches which I’m happy to pay. Just being able to fully mute shitty sites is already so much worth it.
DuckDuckGo, but mostly because of the !bangs. I do 90% of my searches through StartPage (!s), and the rest directly on a few websites (Wikipedia, YouTube, Arch wiki…).
I switched to DDG almost entirely because of the !w bang — Google massively downranking/hiding Wikipedia links made it a lot less useful to me.
Yes, the bangs are a game changer! They even work well on niche sites like Jisho and Scryfall.
Bangs to search Scryfall might actually be a selling point for me. I’ve been experimenting with Kagi, seems like it also supports that as well. This is amazing, thank you for this useful information. :D
I see a lot of people like duck duck go. Could you tell me the reasons why? What makes it good?
What are !bangs and how do I use them?
Take a look at this page.
Exclamation marks!!!
!auk searches Amazon UK
!aw searches the Arch wiki
!w searches Wikipedia
There are many, they are useful
Ecosia. I like trees.
I’d use Ecosia still if it weren’t for the fact that the filter is missing the “last year” setting. I’m a software engineer - 9 times out of 10, I want to find the bugs for a very specific version of a software, so having the year filter helps.
I now use Brave Search.
Fair point. I use ecosia to search everyday stuff on my phone and I’ve planted a lot of trees just by that. The results are fine.
At work for specific enquiries I use Google because I know how to get the details out of it at first. I don’t remember doing a Google search without additional tags. F.i. At work I use visual basic a lot and since it’s a shit language I have to Google everything including “vba excel”. There’s not a lot of logic to it, I just need to dive into the right post head on.
It points at a deeper problem though. There’s sooo much garbage on the internet that I have no idea of how ordinary people are supposed to use it anymore. I’m not disenfranching anyone, but for someone like my mom. How the fuck is she ever going to get an answer of any search? Google just plain sucks by now.
Yikes. I vaguely knew people didn’t like Eich, but it knocked Brave right out of any consideration as even a backup browser. Thanks for that!
Fucking same, last year is a game changer for finding results relating to more modern and relevant (yet mature) answers and approaches
Yay Ecosia!!
Do you know if ecosia is still using google under the hoods?
It uses Bing, similar to Duck Duck Go.
Afaik they use the Bing engine
Apparently not, but the (usable) results are much the same. Ecosia was borrowing from yahoo and bing, but they changed that some years ago.
I have no idea of what or how they’re doing it now, but then I also only search explicit things. I never liked the idea of a search engine trying to understand what I was actually looking for, so I state myself clearly in that regard.
I use Ecosia too. I find that Ecosia is good enough for ~80% of my searches and then I use Whoogle/Google for the more pressing/granular searches. I do the same w/ the free chatbots as well
Currently down for updates, but does a great job of avoiding SEO abuse/blog spam/etc. Takes you back to the earlier days of the internet when it felt like there were more forums/individual sites/etc. They’re still out there, just hidden under all the junk.
Thanks I look forward to trying this.
I use mostly either ddg or brave search. I miss the google of pre 2010, when the majority of its results were good.
I also use Yandex whenever I’m looking for pirate stuff, the only engine that doesn’t block those kinds of results.
God damn it maybe I’ll have to go back to Firefox and DuckDuckGo.
Thank you for these links! What a bummer.
I’ve been using Ecosia for a while and liking it. I think the results are usually better than Google and the image search is way more useful, still gives you direct links to the image files. Though most importantly I like planting trees.
@SemioticStandard Kagi. I used DDG for a long time, and Kagi is strictly better. Specifically, it’s very snappy and I trust the privacy guarantees even more since I’m a paying customer.
Kagi, hands down, is by far the best search engine I’ve ever used (next to Neeva, which got bought and shut down) without looking for Reddit results all the time.
Just simple searches like “Best gaming headphones” or “Realtek Driver Download” and comparing them with Google, Bing, DuckDuckGo, Brave, Startpage, etc. shows how the quality of the results are far superior.
And you can directly define, which sites you’d like to see higher / more results of or less - or even completely block or pin them to the top.
Also, it also shows you directly, before visiting a site, in colors if a site has a very high number of ads and/or trackers.
And they support for power users custom CSS to adjust everything, URL rewrites (e.g. change all Reddit URLs to old.reddit or to automatically open libreddit or archive.org versions), DDG and custom bangs, and much more.
Lastly, I created a so-called “Lens”, which allows me to search Lemmy / Kbin content only (also still have one for Reddit).
Meaning with one click, it shows me results from only sites or keywords I’ve defined - see image.Very satisfied with it, can only recommend.
(copied from another thread I replied to)
@Nankeru @SemioticStandard ohhhh can you share your Lemmy lens?
Kagi is sounding. But looks I’d have to learn to use crypto in order to use with anonymity
Kagi is not a purely anonymous service. They keep track of how MANY queries you run for billing purposes. (They never track WHAT you search for). No special crypto stuff needed.
What plan are you on? Did you adjust your usage behavior to not waste search queries?
Didn’t adjust my usage at all. I used the plan with 1000 searches, but since I work as an IT administrator and literally make searches everyday throughout the day multiple times, I changed to the ultimate plan.
For normal (home / mobile) usage, 1000 searches are more than enough for 2 people.
I use Kagi too, it’s surprisingly snappy! Like seriously impressive for a small org. They talk about speed optimisation being critical for them as well. I find the result to be excellent as well. A true Google replacement/feels like Google in its prime.
I believe they have their own index and bot as well?
…I trust the privacy guarantees even more since I’m a paying customer.
How effective is this stance? While using a service for free basically guarantees there are privacy concerns, paying doesn’t directly provide any assurances. They can both charge you money and sell your data.
But Kagi seems pretty expensive, and too reliant on Bing/Google search pricing to be a long-term solution.
certainly it’s not a panacea. But Kagi’s entire value proposition is that they charge you to cover their costs. If they sell my data, they lose all credibility and the business dies more or less immediately. That creates a strong incentive for them to build in privacy by default.
Plus it just feels better to me. I’m sick of the hidden cost of free digital services. I sleep better when I pay for the things I need.
- tenet ( @tenet@kbin.social ) 2•2 years ago
“I pay them to lie to me, I don’t get lied to for free!”
+1 for Kagi, seems a great value to me, well worth the price to not have any ads, no tracking (leap of faith here) and great search results.
I’d never heard of Kagi before. How do their results compare to DDG or Google?
In my opinion, much better than DDG and Google. With DDG, I had to fall back to !g relatively often, and with Google the results became wildly inconsistent in the past few years, often showing several sponsored posts before showing me something useful.
The plan with a 1000 searches per month is imo well worth the cost.
I do a lot of searches so for the time being my default search engine is Ecosia and when it fails I use Kagi. That way it doesn’t get too expensive :)
I’m also a Kagi user, and it’s completely replaced Google for me for everything except the occasional image search. It’s awesome, and I consider it worth the price.
Yaay kagi! Totally worth it’s money for me, I only have to fall back to Google once every few days (mostly for very local, obscure searches) and am very happy with the results otherwise.
I’ve been using DuckDuckGo since, at least 2010, maybe earlier. If its results aren’t up to snuff, I’m not aware of that because they’re what I’m used to. I fall through to Google ( !g) if I think there might be more out there. The bang commands are so good. I use DDG as my main search in my search bar and then I can use the bang commands to get to whatever specialized search I want from there. It’s a meta-search-engine.
I run my own searx instance
As someone who’s only recently heard of SearXNG, why searx and not SearXNG?
I looked into it for a bit and it looks like there’s basically no reason. Searx development has mostly stagnated, while SearxNG is a fork with a lot of improvements and much more active development. You should just be using that nowadays.
I think it’s no more used searx. There’s only searxng
Duck Duck Go is the only search engine I use. Switched away from Google for privacy reasons and haven’t missed it a bit.
I am a long time DuckDuckGo user. I came for privacy and stayed because of the features.
I typically use StartPage, sometimes DDG. Occasionally I pop in and check out how Brave Search is progressing, out of curiosity.
I would love to use Searx, but I’ve never found an instance where functionality wasn’t breaking all the time or it just randomly goes offline. As much as I want to be, I’ve learned that I’m not much of a self-hoster. So, yeah, every time I try Searx, I wind up back at StartPage. If anyone has any solid, reliable instances they know of, I’d love to check them out.
I circling between few “trusted” instances and always at least one of them works well. You can do it quite easily with Firefox’s keyword bookmarks like “!searx2 <searchterm>”