(untitled)

Is it just me, or nowadays search engines are useless, badly design garbage? Every search engine I've tried spits literally the same generic results from exactly the same sites (this problem also applies to images, videos and news). And, on top of that, it's becoming pretty common to see that stupid and annoying "you requested too many searches, try again latter" error; I mean, I know it's a way to prevent automated requests, but it's just plain irritating. Also, I personally believe that it is becoming a problem for search engines to generate quick responses using A.I. It's very common for the thing to generate responses using outdated, unreliable or highly biased sources. It looks like the people who design searching algorithms just copy-paste what Google does and create something not so different or even worst than that crap. It's like soda companies trying to replicate the Coca-Cola recipe; it's a damn mess. There are more things I would like to bring up, but the post would get long and boring... Sorry to throw my dumb anger here, but honestly today was my breaking point.

It's not just you. Finding anything has become a two hour work. We had a saying: the best place to hide something is on the page 2 of google because nobody ever go there. The result was always on page one. Google used to have an I am feeling lucky button that just teleport you to the first result, most of the time it was exactly what you wanted. Try to look for CYOA pictures for instance. Internet being dead is one factor. Everything is on 10 website now, but that's not the entire story. I will say the problem is with tech in general. When was the last time a big tech have produced anything of value? Take microsoft. 11 is a repulsive spybot and nobody want it. 10 is meh-okay and we got it only because 7 was getting old. And 7 was only XP with gimmicks and draining twice the ressources of XP. It's the same with all their lines of product, at best it's meh-okay. The last good product that Microsoft has made was in 2001. It's my opinion but I think tech simply don't hire skilled people anymore and fire and blacklist skilled people over wrongthink. The bosses just want easy money from investors and creating a product that sell isn't the goal anymore. Like other industries, tech sell to the investment funds, not to their customers.

>>84413 >11 is a repulsive spybot and nobody want it. 10 is meh-okay Lol what? Yes 11 requires an account now but I can't see how and why it spies less than 10. Except the account.

>>84414 10 send automatic reports on your activities. 11 check the DRM of every program and every file you open. And every minute, it tell MS what you are running and what is open.

It's been soykaf for years due to seo spam bullshit, ai gen garbage and many other reasons, but at least in the case of google there was also was a very deliberate enshittification effort that eventually turned it unusable. https://www.wheresyoured.at/the-men-who-killed-google/ I find duckduckgo with ublacklist and a huge blocklist to make it kinda tolerable. With all the dgg ai soykaf turned off. You can find the good blocklists on the github repo readme for ublacklist I think.

The internet is completely unusable. There is nothing of value on here anymore. The combination of midwits spamming inane garbage, and seo parasites spamming literal spam killed the internet over a decade ago now. It just keeps getting worse though. 99% of things created organically by humans today are posted to proprietary platforms like facebook and X, which block crawlers and unauthenticated access. So they lose discoverability almost immediately after creation. Everything else is for-profit clickbait ad-sponsored trash. (for-profit isn't bad, but the current business models are) This is what web-crawlers are allowed to index, and what makes up the majority of Google's results. This soykaf needs to be nuked from orbit, it cannot be salvaged.

>>84414 Don't forget that in 2015 after W10 was released, microsoft backported the telemetry to W7 through multiple patches that automatically got installed through windows update. I've been a GNU/Linux user for 10 years now thanks to that.

It is bleak but I think things can change to be more like the past >>84428 Xitter will for sure die off in coming years as people switch to BlueSky and Mastodon and Misskey at least. Facebook probably will be around a while still though. Just be sure to remember to poison AI datasets to help end the AI slop investors and selfish techbros are pushing that is accelerating the death of the internet. A lot of search engines are popping up son a few years people might be using Google less so SEO content may diminish hopefully. Maybe things would get better faster if hackers target SEO sites or something.

>>84434 >Xitter will for sure die off in coming years as people switch to BlueSky and Mastodon and Misskey at least. The most uninteresting people on the entire planet use those platforms, they will die long before X or facebook. But it doesn't matter. I was trying to make the point that the concept of an open Internet cannot work because it will turn into an open sewer eventually. Any discoverability system will always be gamed by parasites shiIling penis-enlargement pills and hair loss prevention systems. This applies to other things on the internet than just the WWW. See multiplayer video games; All of them are plagued by cheating subhumans who eventually find a way around the technical measures put in place to stop them. Some form of social user-driven gatekeeping the only way I think you could reverse the trend, but I have no idea how to implement it or what it would look like.

>>84414 >>84432 11 has new tech "windows recall" to record everything you do that is still missing in 10. And the telemetry in 7 was trivial to block. Since 10 you don't even have full control over updates as a countermeasure to this and updates reset privacy settings meant to stop spying. They went full malware since 10 and doubled down with 11. Btw the image viewer update also added generating fingerprints for images to send back home. So they track now what images you look at when even in 10 if you using the their viewer.

>>84438 >And the telemetry in 7 was trivial to block No it was not. Blocking telemetry in W7 required you to download a batch script from some rando on the internet that uninstalled an arbitrary list of updates and blocked random domains via the hosts file. https://gist.github.com/xvitaly/eafa75ed2cb79b3bd4e9 Every script uninstalled a different subset of updates, some of which disabled certain necessary services. (I remember Forza Horizon stopped working when some of these "telemetry" updates were uninstalled) A lot of the commentary for disabled updates is like: >Miscorosft Claims it's a security update but also has a key logger on the Kernel Level) >Microsoft claims it's a security update but it contains IE spying >Fixes an issue regarding long wait while searching for Windows Updates but also has Windows 10 Upgrade preparation for Windows 8 So in order to disable telemetry, you also need to remove security patches and bug fixes. These updates would also randomly reinstall when you rebooted your computer. (and force you to sit through the unbelievably slow windows update process) Maybe it is "simple" for you to download and run a random batch script, but you have no way of knowing what is actually happening on your computer. You are trusting some NEET with wireshark randomly toggling updates on/off and seeing what breaks.

>>84439 >turn updates off >no telemetry trivial as i said... was simply retardation to keep updates on and waste time hunting down random blogposts which updates are save and which need blocking

>>84434 twitter seems to be growing more and more doesnt it? is it just my bubble? every day i see dozens of tweets shared and nobody gives a fuarrrk about other platforms

>>84442 >every day i see dozens of tweets shared and nobody gives a fuarrrk about other platforms Obviously, since discussions about these other platforms get derezzed. Or so I've been told

Can some testify their own recent struggles? Not to invalidate the complaints, I think they are fair and I had my own struggles myself. I haven't had a reason to search further than the first page of a search result for a while, though. My most recent struggle i can remember is trying to find a specific meme to use as a reaction but failing to remember the specific content of it. Or, just getting the meme I want, but with its parts cutoff. Actually, Why do all websites crop their images so fuarrrking poorly? god damn.

>>84442 The west's perception of twitter is growing more and more right wing meaning it'll soon be socially unacceptable to use it. Numerous businesses and even municipalities in my area are leaving it for this reason and because of this I have less and less of a reason to keep an account there as time passes. Polite sage for off topic.

First, I must thank you all for taking this thread seriously even though after reading it several times myself I said “wow, I'm such a pitiful buffoon”; that's why I consider lainchan one of the last places on the web where you can talk seriously about these things. This is just my opinion, I may be wrong and I invite someone more educated than me to correct me if so: Saying “The Internet is dying” is wrong; I think they rather mean “The World Wide Web is dying.” I know, I think I'm coming across as pedantic, but my support for this is that the Internet is not dependent on the Web, as there are other protocols such as, e.g. Gopher, for using the Internet. Of course, there are many more protocols but sadly not as well known as HTTP. I definitely agree that the Web is ruined thanks to unscrupulous and greedy measures of parasitic corporations that disguise their ego and self-interest as "technological advancement." That's why we humans cannot have nice things; we end up turning awesome, useful stuff into a stick covered with excrement: wherever you grab it, you get stained. At this point, the last alternative would be to go back to something unconventional and niche (obviously not HTTP-based) as USENET was at the time. After all, companies want to get their hands on something that is mainstream and profitable for them; something that everyone is using... like the Web. Yet, there are many disadvantages, such as the fact that it would be unfair to people who are noble and who have good intentions but not computer savvy; it is easier and more accessible to create a Blog using Blogger (which is from Google) or to create a Facebook group. Many people are intimidated by even seeing the command line appear by mistake. I say this primarily because of this comment: > twitter seems to be growing more and more doesnt it? is it just my bubble? every day i see dozens of tweets shared and nobody gives a fuarrrk about other platforms People might migrate to something other than Twitter (or X whatever), but Twitter is more popular and accessible than, say, Bluesky. By this I mean that people, especially the ones who want to be popular, prefer something that ensures them exposure to a large audience, even if it sometimes means abandoning their principles (like the people who claim to hate Musk but continue to post vanities on that vile social network). A bit conformist, to be honest. But I'm already starting to get into social issues and honestly I'm mentally lazy to dig deeper into it, besides the fact that I have already extended myself unnecessarily. Sorry if this thing is worded bizarrely, but I haven't slept like in 2 days.

>>84448 >unfair to people who are noble and who have good intentions but not computer savvy; lack of entry hurdles is how the internet declined. you need a filter to keep the stupid out or the stupid will attract the wolves who hunt them. Spam is the easiest example, only 0.01% click on it and everyone suffers because we share space with those idiots.

>>84412 It is not just you. Modern search engines have become bad enough that I resorted to running DeepSeek R1 and V3 locally. It's slow, error-prone and an insufficient substitute but still allows me to do things other than endlessly trying to refine queries. I've been thinking for a long time about a "manually"-curated local search engine, but I've never been quite sure about the details. One of the things that fascinate me about DeepSeek is that it's essentially a very lossy compressed version of the entire internet which fits on your disk. I also believe it shouldn't be limited to just HTTP(S) websites but include all types of resources. Perhaps structured more like a dictionary or a compact wiki. The most important aspect would be sharing these manually-curated datasets, which should have a simple but stable format. I have some idea on how to approach it, I should just commit to a prototype at this point and see if it is workable.

Hello, This is my first time posting here, so please forgive me if I’m making any mistakes or coming across awkwardly. I found my way here while looking for people who might share similar thoughts. I’m based in Japan and regularly use Japanese websites, but over the past decade or so, I’ve watched the World Wide Web (WWW) change dramatically—often for the worse. There was a time when the Web was a space where we actively went to seek information. Now, it feels like we are constantly being shown things we didn’t ask for. Search engines are flooded with SEO-driven noise. Social media and video platforms serve up endless algorithm-selected “recommendations,” regardless of a user’s actual intent. Even when we think we’re choosing freely, our choices often feel hijacked by predictive systems. Of course, the content we’re truly looking for still exists somewhere on the Web. The problem is that it’s fragmented and buried under massive amounts of irrelevant or low-quality information, making it increasingly difficult to reach without a great deal of effort. I believe the core issue stems from the explosive growth in both data and users. There was once a sense of intentionality—you had to sit down at your computer and search with purpose. But with the rise of smartphones, billions of people now casually generate and consume information, constantly and without much friction. As a result, the WWW has become a sea of noise, where discovering anything meaningful requires reliance on external filters, not personal curiosity or judgment. To address this, I’ve been considering a simple idea: a dedicated web browser for people with specific tastes or values. The idea would look something like this: Websites detect the browser via User-Agent When accessed through this browser, the site provides an interface or experience tailored to a particular context or worldview On regular browsers, only a limited or abstracted version is shown—or the site may appear inaccessible Technically, it’s very simple. It could work on top of existing HTTP/HTTPS protocols. The site signals, subtly or directly, that a different informational layer exists, accessible only through the designated browser. Naturally, this kind of system would require human curation rather than algorithmic filtering. That is, decisions about what to show and to whom would rely on subjective judgment and the values of a given community, not impersonal metrics. It would be imperfect, inconsistent, and undoubtedly biased—but I believe that’s a strength. If we remove human intent and nuance, we risk surrendering everything to large-scale algorithmic homogenization once again. Perhaps we are all just drifting in a vast sea of noise. But if somewhere within it, there are small lighthouses—sites that can only be found through deliberate tools— then even if they don’t “save” us, they might still offer a quiet refuge. This isn’t meant as a call to action or a manifesto. Just a fragment of thought. Thank you for reading to the end.

>>84457 Interesting post, but I think the problem with this is that there still isn't an effective method of gatekeeping different "versions". If a version was successful and attracted intelligent users making high-effort posts, it would eventually develop a reputation as the "good community". Then everyone wants to join, but the influx of users just degrades the average post quality. This dynamic has played out many times before on niche interesting discussion boards.

>>84457 Is the Japanese side of the internet any better than the English speaking side? The algorithms pushing things we are not interested in is least of their issues even though it is one issue. They can cause people to lose their sense of self entirely succumbing all their opinions and thoughts to some wealthy tech company owner or government entity. I see the issue to me isn't so much smartphones and larger number of people online but rather more so how everything became more corporate business modeled and centralized where near everyone is using the same sites and consuming the same media. Unfortunately curated content has as much a risk as algorithmic content in being abused by an authority. I am not sure what is the best solution to that though. I do agree on layout customization but I could see that as not likely since I've seen sites do this and often trying to code for it they have more design flaws and code bugs. >>84449 PC literacy is not a sign of intelligence or how someone behaves. A lot of PC nerds are emotionally stunted and awful to interact with. No, the issue is lack of individual spaces for different individuals to choose their own communities. Too many people are all crammed into one place.

>>84457 Browser addon's allow you to set the user agent to a desired value (this is called spoofing) user-agent locked websites equate to knowing a passcode that you broadcast as your user-agent over addons This is already a thing in a bad way - i had to spoof my pale moon browser to chrome to access some mainstream sites That aside the core of your idea kinda exists in the form of Tor - Tor websites are limited to Tor users. Of course tor has its own issue: normal people are too afraid to use tor due the association with drugs&crime. Lastly i would like to look on the webs history: At the start of the web, websites would have a link section listing other websites the site owner liked. Lacking search engines the web was connected through collective effort. When search engines released they simply crawled those links and provided faster access than doing it manually. But as search engines became the standard the effort of creating link sections was dropped since why bother when everyone is gonna use the search engine? This is imo what destroyed engines - they relied on copying link sections websites creators provided and as the websites stopped providing a link section the search engine had nothing to copy anymore. Falling back to the original idea i think the solution is very simple: Return to hosting link sections, listing websites you like. Going back to a crowd sourced interconnected web is likely the best counter to engine degradation. You can do this on mainstream sites for higher visibility too. Posting collections of links on reddit/github will give it high visibility in search engines. The "share good link" threads are probably the biggest throwback to old crowd sourced web.

>>84450 That last part sounds like you're trying to reinvent kiwix. I doubt of the feasability of a curated index though, the internet becoming too big was the reason search engines became so popular.

84459 Even if a small online community starts off well, once it begins to grow, its culture and shared context inevitably start to shift. This is something we’ve seen many times in the Japanese web space as well—and I believe it’s an issue that affects all communities, not just isolated ones. What I personally find most problematic is the way users are thrown into massive, borderless platforms, where there are no fences, no context, and no sense of shared rhythm. People from vastly different backgrounds and values are lumped together into a single space, which naturally leads to confusion and conflict. 84460 I don’t use the English-speaking web enough to make a direct comparison, but I can say that the decline of Google search and general fatigue with social media is definitely a shared sentiment in the Japanese internet. Many Japanese users no longer trust Google as a reliable search engine. Instead, they’ve turned to alternatives like X (formerly Twitter), ChatGPT, or TikTok—especially among younger generations. In particular, X dominates as the primary platform for many Japanese users, while a number of tiny Discord communities exist scattered across the web. The once-dominant anonymous forum 2channel (now 5channel) has long fallen out of the mainstream. Currently, the most active anonymous boards are the “may” and “img” boards of Futaba Channel (2chan)—although their userbase is far smaller compared to X. You made a good point earlier—just blaming algorithms for “forcing content” onto users doesn’t explain the whole problem. What’s more dangerous, I think, is that this forced exposure has become normalized, and many users have gradually stopped making choices for themselves. The potential issues with “human-curated filtering” are also very real. For example, one could imagine feeding a majority-voted standard into a large language model (LLM), and letting it curate content based on that. (Whether or not that would actually work is another matter entirely.) Even if a community like that does emerge, it will likely evolve and shift over time. But ideally, new, smaller communities could naturally splinter off and form their own ecosystems in response. 84461 You’re absolutely right that it’s possible to spoof your User-Agent with a browser extension. But the core of what I’m proposing is intentionally adding friction to the experience itself. Using a custom browser and spoofing a User-Agent are essentially similar in that both introduce a deliberate step or inconvenience. That’s the point—it’s not just a technical trick. Rather, it’s an attempt to reintroduce intentionality and selectiveness into how we approach information—by raising the threshold of access on purpose. My interest in this kind of system began when I came across a small Japanese community hosted on Tor some time ago. It was rooted in the culture of old-school crackers, with connections to forums like 2channel or platforms like Nico Nico Douga. What struck me most was that some of the users were building chat tools and manual link directories by hand. It reminded me of the early Web—when people built things for others to find, not for algorithms to crawl. That sense of personal taste and handcrafted curation made a strong impression on me. I also completely agree with your point that the decline of personal link directories has drastically altered the structure of the modern Web. We now live in an ecosystem where it’s increasingly difficult to share and discover “personal treasures”, or connect meaningfully through them. That’s exactly why I think we need to bring back a sense of ritual and conscious effort in how we “choose, search, and access” content. In practice, that could mean not displaying everything by default, but instead requiring the reader to choose to enter a certain context, or opt into a different way of seeing the same content.

>>84460 I strongly disagree with everything you said. Blaming "corporations" is a cop-out. People aren't manipulated into watching trash, they make an active choice to do so. Companies, just like scammers, look for easy marks. Yes, anybody can be scammed by a skilled social engineer with access, motivation and time; However the vast majority of scams are carried out by trawlers casting out massive nets to identify and exploit the easily manipulated. On the internet you can listen to (almost) every piece of music ever recorded, watch every movie ever filmed, and read every book ever written. You can choose to learn a new language, a new skill, pick up a new hobby, anything you want. And all of this is available either for free or at a very reasonable cost. Yet a large percentage of people disregard the vast riches available to them, and exclusively consume trash targeted at the lowest common denominator. It's 100% a human issue. There are significant variations between individuals and you are willfully ignorant if you pretend otherwise. The internet cannot be fixed if people refuse to recognize this. And PC literacy is strongly correlated with intelligence.

>>84467 Platforms decide what people are allowed to see and manipulate thereby what users consume. YT is insidious because it still allows some good content creators but then bombards you with garbage. Their entire engagement design is to slowly convert people from good content to consuming old media slop. You can fault people for being not more resistant but it remains true that all the social media platforms use psychologists and addiction research to manipulate people as best as they can.

>>84412 At least for me, I noticed the downfall began when search engines started ignoring search operators like " ", inurl, intitle, -, filetype. This combined with the increase in spam, AI-generated content, and mass advertising being indexed has been the nail in the coffin so to speak. I have yet to find a search engine that has the same index size and quality as Google used to around 2010 with support for advanced search operators to allow for narrowing down on what you're looking for and where. Currently I use Brave Search because of its independent search index but the results are not great and it doesn't respect search operators.

>>84412 I've basically replaced search engines with AI since they give more detailed answers faster and it's a lot better than dealing with 50 useless links that take me an hour to fuarrrk through hoping to find a crumb of relevant information(which will have ended up being wirtten by AI anyways.) This is sufficient for basic information like wikipedia article type stuff, hardware/software/coding information, hobby information, stuff that in general is readily available but takes too long to compile by hand. Even for more esoteric soykaf if you know how to lead the models you can get some pretty interesting/useful soykaf from them or at least it will get you thinking in a direction you weren't before. For anything AI can't manage I just look for more esoteric knowledgeable content creators on less popular sites/video platforms and then kind of network through them, see who they talk to, what they read, what they bring up etc. This is the only useful way to handle the modern internet because just using a search engine has become too tedious and ineffective and most of the internet has devolved into midwits or utter half wits spamming their ebonics and retardation like pigs in soykaf I've come to realize I don't really want to socialize on the internet, I hate most of it, it's way too parasocial and empty and I can literally feel myself becoming more mentally ill the more entrenched I get in this garbage. That's the future imo, treat this soykaf as a tool and look for the most optimal ways to enrich yourself then cut it off. This should not be a way of life

>>85574 AI is pretty good now, but I still try not to use it because it is always (confidently) wrong. Even using the best models like Gemini 2.5, GPT o4, and Deepseek r1 this is still the case. For programming it will just invent nonexistent documentation, stdlib functions, github repos, compiler flags, etc. I only know that it's wrong because programming is my area of expertise, and it's easy to validate the AI's output by compiling and running it. I wouldn't feel comfortable using it at all for something I'm less knowledgeable about, or where a compiler cannot validate the output and there's more of a grey area as to whether a response is correct.

I have heard very good things about Kagi but it is a subscription service which is hard to swallow for a search engine. They try to offer a bunch of different AI services to sweeten the pot, but I just can't get over the idea of paying for a search engine. They offer some free searches as a trial and I use those when I really need good results. I think search engines outputting trash is the natural result of everything else being trash. Every service was trying to maximize profit and abuse SEO before AI gained in popularity, and now it's just gotten unusable. Something like 50% of all site traffic is bots now, and with the few big sites people actually use starting to lock public access, there's not a lot of good stuff left to search for. I think the realistic way things are going will mimic VPN sales. You can try to get by on the free ones which are horrible and a privacy nightmare, or you can self host or pay someone else if you want something usable. For search engines you can look at self-hosting something like searxng or paying something like Kagi to filter the slop out for you, but public search engines will become more and more poisoned by AI slop and SEO abuse. It is like kicking dead whales down the beach, but I can't see it getting better in the future.

>>85577 yeah you have to be careful but personally I find search engines more frustrating. At least the AI will respond to the topic I brought up whereas a lot of times I search for something and I get results that aren't even tangentially related to what i was searching. we just have to make do at this point

Search engines were dead before AI spam was a thing. See picrel.

>>84412 I strongly believe that AI bots will replace search engines as they become better and better I, of course, don't think this is a good thing but I do believe it'll happen

>>85579 >Kagi I'm on the $10 plan and love it. Good search results and you get a lot of control over them.

>>84460 >Too many people are all crammed into one place. Why not think of it as a form of containment? if everyone is crammed into one place then go and make your little community somewhere, where, hopefully, only people with some brain cells exist (not because it's hidden or something, but mainly because people don't even think of leaving the space they're in, because it's "the thing people use").

>>85556 Searching for images has become specially annoying, too much AI generated slop

>>85585 From what I hear about Kagi it's probably one of the few subscription plans worth the subscription, I just hate the payment model. I don't want to encourage more services using a subscription model because everything is moving that way and it's easy to end up spending too much each month on a dozen subscriptions. I wish they adopted a "credit" system like openrouter has but I know the whole point of the subscription model is that occasional users subsidize the power users. I think Kagi is more "ethical" with their subscription, since they don't charge you if you don't use it at least once in a month, but I still haven't bought in yet. I'll probably give up on DDG and sub at some point anyway.

>>85585 How is it for academic research? As in, a Google scholar replacement?

>>84467 >Blaming "corporations" is a cop-out. People aren't manipulated into watching trash, they make an active choice to do so. Companies, just like scammers, look for easy marks. Yes, anybody can be scammed by a skilled social engineer with access, motivation and time; >However the vast majority of scams are carried out by trawlers casting out massive nets to identify and exploit the easily manipulated. I would suggest you study psychology because I think you're underestimating what the average person is susceptible to. Also it's often the case where they do not really have a an easy choice. >Yet a large percentage of people disregard the vast riches available to them, and exclusively consume trash targeted at the lowest common denominator. Most people don't even have time for that or if they do they aren't given good guidance on how to approach things and get frustrated and wind up giving up or moving on to something else. >And PC literacy is strongly correlated with intelligence. Even a chimpanzee can use a PC though. I think this is just an ego trip some people tell themselves to feel special. >>85586 > Why not think of it as a form of containment? if everyone is crammed into one place then go and make your little community somewhere, where, hopefully, only people with some brain cells exist (not because it's hidden or something, but mainly because people don't even think of leaving the space they're in, because it's "the thing people use"). The problem is there is a certain social expectation. Even many schools and workplaces expect everyone to have certain social media and messaging programs and apps as a requirement. >>85588 Sometimes filtering the search with "-AI" can at least slightly reduce it. I have been trying to find a way to automatically identify AI images but haven't quite figured out the forensic identifiers in their data. Sometimes they have inside the image the prompt terms but this is rare.

>>84412 Almost a decade ago, I used to use reverse image search for everything. I like to hoard as many images as possible of various things like fanart, photographs, memes, visual assets from video games, ect, but I'm also very particular about sorting and grouping them so when I need to find the source of an image I'd reverse image search it. Ever since Google pushed out Google Lens, their reverse image search function was utterly gutted. That tool rarely brings up related results, and if it does they're in the handful and only marginally related at best. The sites it pulls images from are hot garbage too since you're always guaranteed to get pinterest. I switched to Yandex which was a useable alternative for a while, but they implemented their own lens-esque feature and the results feel significantly worse over there now too. I really loathe this era of the usability of online tools and websites worse for the sake of advertising.

>>85592 For general image search you should use Tin-Eye in cases Yandex doesn't work well. If you're looking for anime screenshot sources use trace.moe and if looking for either that or an artist who drew something then try SauceNAO. >>85594 Lol nice.

Are just now starting to notice? Google has been soykaf for years. There is a reason why everyone started to append "reddit" to their search queries. Reddit obviously has its own problems but that aside. I also love how Luke Smith's based.cooking is still most easy to use and to the point recipe catalog on the internet. Finding recipes with Google was awful.

Kagi user here. I've been using their paid services for a few months now, and I wanted to share some of my thoughts: Firstly, not getting ads and sponsored links, having AI results (including images) mostly filtered out, and the additional manual preset filtering or pinning of certain websites is already making the search experience much better than Google, which should be expected. Their own AI-generated quick answers are actually usable if you're knowledgeable about the subject matter (like >>85577 described), or if you're sure that what you're looking for is such common knowledge that AI can't possibly mess it up (but even then, I recommend still checking the sources). Their video and podcast search isn't bad either; it's certainly a good way to avoid their algorithm slop from the platform they're hosted on. The recently introduced full AI assistant (not just quick answers) is somewhat comparable to Perplexity, as you can give a vague description of what you're looking for, and the AI will start a suitable web search and give you the results back in human language (but with sources). While it may still lack Perplexity's ability to provide solutions to complex problems, the assistant is usually the first AI tool I reach for, provided I can anticipate that it'll be able to handle what I throw at it. They also use a bang system like DuckDuckGo, which I use extensively. The webpage/video/podcast/document summarizer works well enough and is often sufficient to help me decide if a piece of media is worth my time to sit down and consume properly. You may have noticed that they use optional AI features quite a bit, but judging by their statements about using AI in a reasonable manner, strictly to enhance their search functionality, I'm not too worried about it. I can relate to feeling weird about paying for a search engine, but like >>85579 mentioned, I think it may start to become a necessity sooner or later as a consequence of the SEO/AI infestation that's taking over the web, and free search engines are no longer viable. I'm just starting to get a bit of a tech bro vibe from the company itself, considering their questionable use of finances in the past (like their god-awful trailer video for their AI assistant, or that one time they bought an entire t-shirt printing factory in Germany for a limited run of promotional t-shirts). Also, I'm worried about how they're trying to expand into the territory of other services unrelated to searching. I would be happy to give them money if I could be sure that it 100% goes into perfecting their search index/engine and supplemental AI tools, but then they started to development on their own Google Maps replacement, and I think I read somewhere that they're thinking about launching their own email service. While I don't doubt that these could be good standalone products if given a dedicated team and funding, the company is more likely just getting greedy and wasting money in an effort to replace the Google ecosystem entirely. I do, however, like their small web initiative and how they've expressed the desire to help create a more human, organically crafted internet as opposed to walled gardens. That's why I'm confused about why they restrict most of the user discussion around their products to their Discord server, considering how they should be all about openly searchable traditional forums. I'm happy with using Kagi. Right now, I'd rather pay a bit of money instead of wasting my time on frustratingly bad search results, but in the long run, I hope that the selection of paid/free tools diversifies enough to achieve the same functionality as Kagi, so I'm keeping my eyes and my mind open.

Yeah classical search engines such as Google, Yandex etc... are pure garbage nowadays because of SEO "optimizations" fuarrrking over their ranking system and the search engines themselves prioritizing certain websites and hide others. As an alternative you can use LLMs --- such as DeepSeek and the various ChatGPT --- to search the web. Think it like this, during learning the model is indexing the a large chunk of the web storing a sort of index in its weights' matrices, when you formulate a query in natural language it looks through its index, cross-references the results with those from a classical search engine, looking for the more likely results and produces an output I have had a high degree of success searching for information in niche areas. If it doesn't find any good result you can continue refining your results by re-formulating or adding information to the query

>>85574 AI often gives a sup-optimal answer. This is obvious when you just forgot a detail and try to lookup again and then have to prompt 10x to force your desired info. Extrapolating from this you can deduct any subject you dont know you will get likewise a bad reply you cant correct because you dont know whats right. Old forums are better sources because people provide context as they discuss solutions. AI doenst do that. It just posts something without context and nuance. The problem is of course quality is harder to rate than speed and it seems even corps are now pushing ai to employees because speed is easily verified metric while quality loss is hard to quantify.

>>85634 When you think about it wouldn't it be wiser to fund search engines like Stract and Marginalia and maybe Right Dao so they can eventually become a lot more usable not just only paying for a service from a private company that could easily just start acting like every other search engine as they get greedy? Well i mean you can do both though. >>85637 The only problem with looking through old forums is some have threads spanning 20 or more pages over many years time and it takes a lot of time to read useless posts to find the useful ones.

>>85634 It does feel like Kagi wants to be Proton without taking the same care and caution. Either they are way too ambitious (which almost going bankrupt buying a t-shirt factory hints at) or they are trying to attract new members by hitting every market they can. Likely both. I'd much prefer they focus on search and do it right rather than risk blowing the company up on a side venture. Proton is painfully slow to implement new features but they are trustworthy enough that I'm fine relying on them for the long term. Kagi seems so investor-chasing that I wouldn't want them to have access to any sensitive data or to have my digital life relying on them staying afloat. Another concern along the same lines I have with Kagi and a big reason I'm hesitant to subscribe is that they're running out of the US which has less than great data privacy laws. Kagi's privacy policy seems to be very "we don't want to log anything, so we don't" but not necessarily that they couldn't if they changed their mind or were coerced into doing so. Proton has a lot of focus on encryption to separate them from your data while Kagi's policy is mostly based on good faith. I'm hesitant to start relying on them, especially with email and such, and then the investors or courts come knocking. Cockli has (had?) a similar privacy policy but it was more excusable due to the nature of the service.

>>85648 Kagi is a publicly traded company. So yes iis likely not simply people buying services and products they are trying to get but investors. Investors are who keep failing companies afloat until they turn a profit or they bail on them then go bankrupt. What they are doing is diversification of their assets to help attract investors who see them as a fully functional tech company like an up and coming Google rather than a company just doing one thing.

>>85634 henwo lainon. you of made me decide to of try the kagis tank u very much plox saar

>>85590 I don't do academic stuff so I can't really answer that. They have a free trial for 100 searches if you want to try it out.

>>85692 y-you too >>85590 Yeah not exactly my area of expertise either but from what I could gather so far it's not quite on the same level yet. Perplexity should be more suited for that kind of research, but I haven't had a chance to properly try it (beyond the few pro searches you get in a day) since I don't use credit cards. Please, test a few alternatives and let us know if either of them meet your expectations.

Yeah I've noticed the same thing over the past few years. It feels like there are very few places left where you get genuine, human written answers that aren't just regurgitated corporate backed content. Searching for a specific image has become a nightmare too... they're rarely in the same resolution or aspect ratio anymore, and now everything's in annoying formats like .avif or .webp. What makes it worse is the flood of AI-generated images cluttering the results. It's hard to find anything authentic or specific now. The fact that there are many people sharing information within Discord makes the situation worse. These days I just use AI for quick definitions or small lookups… at least they save me a few clicks when I'm not in the mood to sift through junk.

>>84442 Could be bots to create a false sense of larger community and consensus. Then Musks pretends fighting against bots as part of his war on false accounts while at the same time shadow banning people by massively curtailing their reach if they tick secret boxes that correspond to parameters that twiddler doesn't like.

>>84448 >I definitely agree that the Web is ruined thanks to unscrupulous and greedy measures of parasitic corporations that disguise their ego and self-interest as "technological advancement." Weird way to say religious tribal racist motivations.

There's a new one, with an entry fee, called MetaGer (Ger for Germany I suppose). could it be seriously considered?

>>84437 I once thought it would be great if access to the internet (or computers in general) was granted after passing some kind of exam and getting a license, just like we have for cars. the problem is anonymity dies completely though, but maybe it's not that bad? at least we could get some comfy zones without retards, idk

>>85721 This does make me wonder: is there something like a database or a search engine that actually values and promotes non-AI, high-resolution images? AI is threatening to kill the whole semiotics of objects. I only care because I'd like to look things up to draw without suffering, and recent struggles have made me heavily consider taking my own goddamn pictures instead.

>>85963 >drivers license drunk driving still exists... nor did licensing prevent kids driving cars or people using them for attacks running people over. Legal measure only contain obedient people which fails to address targeting problem personas. tldr: license are ineffective to curb intentional misbehavior You know what likely would have solved almost all drunk driving cases? Car asking you to solve some math questions for ignition. You need hurdles your target wont be willing to cross, it doesn't have to be impossible it just needs to fit the persona you trying to keep out. In that sense i believe any hurdles that take patience to cross will automatically kill low effort type. Secondary you need to have something AI cant solve easily because the modern problem isnt stupid people but chatbots that use full browser proxies to bypass anti-bot filters. There is a new startup going big right now offering proxy browser for ai as a service. I expect its gonna get worse so we need to arm ourself adequate. >>85965 Just filter by date, anything older than 2015 cant be ai. That aside im more a fan of posing characters,buildings,... in blender to get a realistic example when i want to have a correct scene. Easier to pose exactly what you want then find a perfect example when it comes to correct perspective.

>>85968 thanks, but the only reason I'm into art is to train my psychospatial abilities and get a better feel of what 3d objects actually are in my mind beyond vague symbols. posing in blender gives you too much information and we're lazy, if you can trace a 3d model you're not going to bother with rotating stuff in your head as accurately as possible

>>85968 >tldr: licenses are ineffective to curb intentional misbehavior but they filter many unintentional cases. they might not filter *all* kids but most of them with at least 2 working neurons will follow the rules (at least here in Europe it seems so, likely because fines are huge). with intentional it's kinda similar in that you might be willing to break a law but are you sure it's necessary to break 10 at the same time? if you're fighting while drunk it's one thing, but driving drunk is more expensive after you are caught. and let's not forget that after such cases license can be revoked, so we can prevent repeating accidents from same person >Car asking you to solve some math questions for ignition still not preventing intentional misbehavior though. also not gonna happen, too hard to tune specifically for everyone. drinking prohibition seems more realistic but even that with all its side consequences doesn't stop intentional stuff. y'know, if someone *really* wants to make soykaf they will do it, no way to stop it, only reduce rate of happening and prevent accidents. and the best way is of course education but you see where it goes >Secondary you need to have something AI cant solve easily yeah so captcha isn't worth it. always. sooner or later new captchas will also be automated so it's not a permanent solution, we need a completely different approach

>>85963 Or use IQ tests as uberCaptchas, that will filter the retards too.

What upsets me the most is how most search engines seem to have abandoned the "advanced search" option or have hidden it behind some kind of arcane interface. So you either remember all the little intricacies of special search syntax or you have to settle for garbage like constantly being "corrected" on a word or phrase when you know damn well what you meant to search.

>>86013 I find myself frequently repeating words that I would previously put in quotes....sometimes both repeating and quotes....it helps but its not great

> search quoted phrase from some book > 1 result > "no ai overview available" Google search is fake

It's not a fault of search engines themselves. They're actively losing a battle against aggressive SEO algorithms and mass-produced AI slop. What you're witnessing is the internet being completely drowned in pure garbage.

It's a number of things, but I do believe in most of them to be intentional dark patterns, not just righteous search engines losing a battle against the evil slop. One of the things I've noticed, is that specifically looking up any kind of obscure technical problem regarding particularly popular software is basically impossible. The engine basically discards your query and rewrites it in a way it thinks you meant to write it, based on what the majority of people in the past have written. Statistically, most people are idiots, and also statistically, most users who were having issues with this popular software were having this particularly dumb issue, meaning you are also more likely to be having the same issue, meaning our re-written query and re-shuffled results are statistically more likely to be helpful to the average user. Quantitative strategy over a qualitative one - assume every user is an idiot and that you know better than them. It has gotten so bad that I've learned to instinctively recognise issues that I know will be impossible to find through search engines, before I even try. There's other things too, like western search engines filtering out anything that's not a legit paid/licenced streaming service, while Yandex happily gives you links to illicit stream sites for western content (and their pres even encourages it), but that's mostly a political issue.

I think the most reasonable way to still use search engines is to 'hack' them with queries and also using alternative options. The first option is to build a ridiculously complex search query, where you literally put more search operators than the search term itself. Something going like -inurl:(".gov"|"reddit"|"wikipedia" ... [and so on]) Doing this way in a trial-and-error and give you a somewhat decent one–two links result after several queries. Also, if you search for some extensive article that is written for 'deep-dive' and not brief overview (like the most slop articles on the first page) you also can try including "before:2010", or any year that you want. It seems that it eliminates most of new SEO garbage pretty decent, and pretty quickly (instead of chaining 'negative' queries). I don't know, maybe it's me, but I don't have seen any mentions of this operator as to serve a filter to junk results (or new agenda, which increased its presence from near 2018). Btw, also I've made a simple ViolentMonkey script to add a button to filter by language under "Tools" in Google. You can do it manually in the URL by adding param "&lr=lang_en" (or another language code). Really useful knowledge if you use two (or even three) languages; google simply suggest you to filter results only to english, as in your browser's language. The second option is to search a somewhat decent alternative, as mentioned above. I personally didn't use Kagi (although I'd want to, just can't afford it lol), but you can try using free alternative engines like Marginalia and Mojeek. The first one has quite small index, since it uses own crawler, but have nice features like filtering by 'Small Web' or 'Forums', which I use pretty often and thus find niche small-web sites with later WebRings to follow. Mojeek is an alternative search engines that have its own index, crawlers, and — as they claim — cares about privacy. It's looking like the most 'Google-like' alternative, but has much better results in the matter of ideological (or generally mainstream agenda) bias. Besides, very nice feature is 'Focus' that allows you to limit your search only to a list of domains, like a list of dictionaries or *.neocities (small-web). Also I'm pretty sure everybody knows about Wiby, but their search potential is really limited only to 'old-era' of WWW (and a portion of neocities). Nevertheless, it already seems that it's easier to use an AI to search something. Not just because it understands contextual queries, where you can explain better your query, but also because it doesn't give you 40 links with the same mantra, just in different wordings. Looking for something *new* with AI works good, like 'what [algorithms, concepts] I can use to build a search engine' — it'll give you *starting points* to later find details by yourself. In the claimed 'age of information' the information is buried under kilotons of junk, repeated articles 'for-dummies', agenda propaganda & censorship. I'm pretty pessimistic about the future of mass-WWW: I think that most users will comply with the established situation — most probably doesn't even notice - hence will come back to the intellectual (as in knowledge, not IQ) level, perhaps even lower than pre-internet era. Why dig deeper, when you can get an instant answer and trust? But for those who strive for new information, strive for real human opinions and small personal blogs, those ought to group into small communities, and 'spread the word link'. Creating WebRings and directories. Literally coming back to the time when your friend sends you some link to a cool website, from where you follow the rabbit hole. Btw, I encourage you to not only bookmark pages, but also save them (as .html/pdf). Create your archive of the fading human-WWW, preserve the pages. Sorry for such much text, just want to share some possible escape routes for which I strive for already year, everyday feeling deep pain from using modern search engines and sorrow for the old days of WWW.

>stackoverflow now makes you solve a cvckflare fingerprint capcha to browse the site It's over. It's never been more over. The www is officially dead. I will exclusively use AI and libgen as my sources of information from now on.

Obviously the most important question of all has not been asked yet. What are the search engines that still parse the many chans and provide honest results remevant to their content?

>>86117 how big is stack? since its mostly short text i imagine it would be feasible to have a local copy of the entire website to use offline. Might be a good idea regardless since the continued existence is not guaranteed.

>>86336 Less than 100G apparently. Last official one: https://archive.org/details/stackexchange Last community release : https://archive.org/details/stackexchange_20250331

>>85577 >>85574 The entirety of the World Wide Web is caught in a Malthusian trap; Web 2.0's core incentive model built on making our attention a commodity itself incentivizes everyone on it to pursue engagement regardless of what activity you are participating in. Literally everyone is in it for the (you)s, and content or structure that was not aligned was crowded out, and the means of reaching it dismantled. It's a fundamental, structural problem. This is the ensoykafed web we inhabit today. Make no mistake: using AI, particularly those trained on engagement, to organize and explore the web is antithetical to curating a better web. Fundamentally. You've baked the seeds of your destruction from the start. You've already ceded the ground you thought you were defending. Enjoy this old hidden gem of Scott Alexander's that hit the core issue of the use of AI like LLMs before LLMs were a thing. https://web.archive.org/web/20121008025245/http://squid314.livejournal.com/332946.html >It is well that we are so foolish, or what little freedom we have would be wasted on us. It is for this that Book of Cold Rain says one must never take the shortest path between two points.

God... and in fact this is already all over the freaking Web. It's annoying. I'm not exaggerating when I say that this day alone I've come across this thing about 10 times (and I haven't even been online for an hour). I understand it's for fear of DDoS attacks, but several sites (ahem, 4chan) used this service and it did nothing to mitigate the attacks; it just wastes my time or maybe worse. This crap, among other things, make it clear to me why humans can't have nice things.

>>86021 The search engines are focused entirely on profits from advertising especially Google which itself owns an ad service and same with Microsoft with Bing ads. Only way to end these is end capitalism itself or at least break up tech monopolies so they do not control too many different aspects of the internet at once.

>>86349 >The entirety of the World Wide Web is caught in a Malthusian trap; Web 2.0's core incentive model built on making our attention a commodity itself incentivizes everyone on it to pursue engagement regardless of what activity you are participating in. Literally everyone is in it for the (you)s, and content or structure that was not aligned was crowded out, and the means of reaching it dismantled. It's a fundamental, structural problem. This is the ensoykafed web we inhabit today. It's a whole artificial economy where at the bottom of it there's a legion of data collectors who make soykaf ton of money while surveying people and gathering human content for the mimicking machines. >>86355 This and the Google popup because their services are not enmeshed within most websites. >>86365 The business side is used to fund the creation of the next generation of obedient AIs while keeping tabs on citizens on behalf of friendly governments.

>>84467 Most breakthroughs in psychology and neurology came from either 3 letter agencies mapping the brain to understand it better for psyopping the population or from gigantic marketing firms to take advantage of the brain to make people buy their soykaf, pay attention to them, listen to their jingles, etc. People are actually manipulated on every level that is known including you and me even if we don't realize it. The real trick is just being aware of it, which majority of normal people aren't, hence they fall for it and pollute everything on behalf of these corporations/propagandists.

>>86380 >People are actually manipulated on every level that is known including you and me even if we don't realize it. Even through grammatical changes and new words or new versions of old words.

Google in 2025, outcomes. And it couldn't be possibly an error, so they're intentionally just manipulating search results not only by what pages are shown to you, but also by literally giving you wrong metadata, just to make you click. Predicting, that in a 5–10 years Google will just remove the search bar and replace it with AI-answers with "sources" from curated lists, and will obtain money by including sites into curated lists. Ask a question: here's two product-placements in the answer; ask another: here's sources from governmental sites, UN, Wikipedia (with its source/reference policy, aha).

Is it just me or have duckduckgo's results been garbage for a couple of months?