Has search gotten worse?

[White Devil]

Never had any success with Tineye unless the image is super common.

That sucks, I've uncovered a few catfishes with it before I got married. Maybe the people were just dumb and using common images.

 

[Michael Jacks0n]

I've used tineye to some success with this before but I don't know if it's fagified now too.

Try Yandex for context aware image search.

Eh, I've had mixed success for either of those sites. I've also used Bing reverse image search with mixed success, but none of them are as powerful and accurate as the original Google Image search was a few years back.

What also really sucks is how the original Google Image search was able to find copyrighted or stolen material pretty accurately. A buddy of mine is a professional photographer, and he used to upload his images in the search to see if people were stealing them from his portfolio. For example, back in 2013 my friend traveled to San Fran and shot a really excellent photo of the Golden Gate Bridge at sunset with a DSLR and wide angle lens. When he'd upload his picture to Google Images shortly thereafter, he'd occasionally get results of people using his exact Golden Gate Bridge shot on their blogs, Pinterest pages, or Facebook profile covers. Then eventually when he'd try again using the same picture in the following years, the search result just said "Golden Gate Bridge" and showed a bunch of stock public domain images of the bridge at various times of day, followed by random news and science articles about the bridge...but NOT his specific photo that he took. Pretty lame.

This! It's become nearly impossible to do any actual serious searching these days, because every single fucking search engine treats my search term as a mere suggestion by a retard. I'll routinely get tons of results where one or more of the search terms are completely missing, and tons more where the search engine substituted something sort of related or with a similar spelling for one or more terms, and there is no way to turn this bullshit off. Keep in mind that by "related", it can be anything up to "these two terms are direct opposites".

I think that it's largely a result of one of the worst features of capitalism, where everyone mindlessly chases the single largest market segment (in this case, complete retards who don't really know what they are looking for) and smaller market segments are always underserved, along with one of the worse features of human nature where everyone mindlessly follows the market leader because they wouldn't be the market leader if they didn't know what they are doing, right?

Here's my story with the same issue: a couple of years ago I was trying to find a certain obscure B-movie whose name I forgot, but there were some key scenes I vaguely remembered. There was a scene in the movie with sea lions, so I searched Google for movies with sea lions. I'm not making this up: about 75% of my results were news articles about Navy SEALs. After going through page after page of unrelated bullshit, I still couldn't find my phantom sea lion movie, and found myself dumbfounded why the results were so borked.

I concluded that Google's lobotomized algorithm was following this particular pattern of smoothbrain "logic":

1. Sea lions are related to seals
2. The seal animal is spelled and pronounced the same as SEAL (Navy)
3. There are currently far more websites and search results relating to Navy SEALs than there are for information on the animal, let alone movies about seals or sea lions
4. Therefore, I must have meant to search for Navy SEALs the whole time, and not sea lions

And that's just the beginning. I constantly try to get help with issues for tech and whatnot and most of the top results are for similar things I'm having issues with, but Google assumes they're the same thing. Original Google search used to be very specific to the thing you were searching for, and now it shows you irrelevant things that have more common results than the specific query in question (see my sea lion/seal/SEAL frustration).

 

[byuu]

I still can't get over how fucking dumb and annoying it is that Google just ignores keywords I typed in if it leads to more results.
I don't want a billion results that no human can go through, I want a few that are actually about that specific thing I searched for.
Don't make me put every single thing in quotes to get what I actually ask for.

Altavista gave me better search results decades ago than Google does now.

 

[The Littlest Shitlord]

I still can't get over how fucking dumb and annoying it is that Google just ignores keywords I typed in if it leads to more results.
I don't want a billion results that no human can go through, I want a few that are actually about that specific thing I searched for.
Don't make me put every single thing in quotes to get what I actually ask for.

Altavista gave me better search results decades ago than Google does now.

I loved AltaVista's "near" search where you could require that the given keywords were within 10 words of each other, it was good for finding search phrases where there was a lot of room for variation in word order and you could have various small words like "of" or "the" present or not. This is good for finding a technical subtopic that isn't common enough to have a universally-agreed upon name, but the name is instead a combination of descriptors.

 

[Jones McCann]

You get different results depending on what you use (eg. using Google compared to using any of the Searx instances)

The problem is that a lot of search engines lack their own index and rely on other big ones for their search results (i.e. the ones doing all the hiding and misinformation in the first place.

Here's a link regarding it if you're interested.

since we will often share some really sensitive data with them (such as medical, travel, or even that we're interested in loli or shoplifting) - they better not be doing dirty stuff with it.

 

[Not Really Here]

We've gotten to the point where we need that one engine that opened 8 windows and searched 8 different engines again.

 

[Miller]

I've noticed that Google doesn't show an IMDB page at the top of the results anymore, it's always Wikipedia or some shitty website.
When it comes to news or politics, Google shows the results based on the MSM narrative. It's all about shaping people's opinion, especially since 2016.

 

[Jump]

I think yandex.com uses the same engine that bing uses?
My process is ddg, bing, then yandex.RU
The .ru one is good for finding old technical documents and PDF's.

Google definitely has some sort of war on old style forums and mailing lists.

 

[Michael Jacks0n]

Alright, so just a little update here.

I gave Yandex a 2nd chance, and surprisingly it helped me find a Tinder girl was a catfish. A few days ago I saw a decently attractive blonde on my Tinder queue, but she had a blank bio. I wanted to swipe right, but her face looked kinda familiar, instead I took a screencap and uploaded it to Yandex. Sure enough, the first result was the Twitter of some pornstar I never heard of. The photo from the girl's Tinder bio was a selfie from the actual pornstar's Twitter, so it wasn't all over-the-top sexual or like a typical thot pose...rather it was just some cute blonde in a t-shirt. Figuring it was a fake bio, I swiped left, especially since the ages didn't match up (the girl on Tinder was listed as 30, and the pornstar is like 25).

 

[Smaug's Smokey Hole]

We've gotten to the point where we need that one engine that opened 8 windows and searched 8 different engines again.

Webcrawler 3.0 or whatever the latest would be.

Google image search went from hashing the image and searching for locations of it, to some machine learning classifier BS. It would even work with cropped or scaled images, and find a higher quality version of the same image. Now if you upload a photo of a person it'll pass it to its lobotomized AI and return "Human Male" and a bunch of stock photos of humans.

Guaranteed the old search box is still available to spooks and troon employees.

I'm still confused why Google Image Search went from completely useluess to suggesting "Brienne of Tarth nude" and "Brienne of Tarth naked" across different browsers. It was a very specific suggestion and had nothing to do with the picture I wanted to know more about(I think it was from the cursed pictures thread).

 

[moocow]

I used to screencap questionable Tinder bios, do a reverse image search, and find out if they were real or bot accounts. It doesn't work anymore, and now when you reverse image search a girl's Tinder photos, the results are all generic stock images of women's photos with the search result saying, "Image results for 'woman'."

And don't get me started on how the Google Image search function was completely fucked in 2018 when Google kowtowed to Getty Images. That still pisses me off.

Maybe Google intentionally fucked up reverse image search to stop stalkers, or because some troon got deadnamed by a reverse image search, sounds like something a woke company would do.

It's probably the 1st one. I noticed that it started to get less and less comprehensive in the past few years, so maybe something in the post-2016 culture wars kicked it off, like concerns over doxing of ANTIFA members from the mysterious 4chan hacker.

Google image search is utterly worthless now. Has been for a couple of years at least. I have seriously never had any success (seriously, not even one correct result) doing an image search on a person to find out who it is in the last two years. It always just returns image results for "woman" or "girl."

I suspect another factor in it turning completely braindead was that kerfuffle a few years back where it started classifying black people as gorillas. That was fucking hilarious. I'm pretty sure their solution to that problem was to just disable classifications entirely -- they apparently couldn't figure out how to actually fix the issue. The lobotomized leftovers of that classification system is probably what they're plugged into their image search feature, and it's completely broken.

Yandex is the only image search I have regular success with. Crazy Russians doing a better job than the woketards in silicon valley. Gotta love that!

 

[Bongocat]

This seems like a good week to give DuckDuckGo a real, dedicated trial. The results I get for technology related searches are so consistently terrible from google that I'll be really interested to see if I can have better luck somewhere else.

I've been trying to learn a new tool this week, and every single query is leading me to unrelated topics. I can't imagine DDG can do much worse. I'll let you know how it goes.

 

[The Real SVP]

I've been trying to learn a new tool this week

Have a look at this next week: https://searx.me/

 

[Michael Jacks0n]

Remember Facebook mosaic search? Like around 2010-2013 you could find most anyone based on their interests and likes. Nowadays the most you can do is search by name and then narrow it down by school, work, or location...and even then it's not very comprehensive. Also the Facebook search in general isn't as comprehensive as it used to be. Like of you try to find a girl named Marlene, most results are girls named Maureen or Maura, or even guys named Marty. Has anyone else noticed that?

LinkedIn search has gotten shitty as well. Let's say you're trying to find a woman named Kelly Johnson (made up name): about half of the results are women whose names are something completely unrelated, but they work for a company like Johnson & Johnson or Kelly Services Inc. and it's really fucking annoying. Also it doesn't show you their photo unless you're in their network, so it's hard to see if it's the actual person you're looking for.

 

[stares at error messages]

I think search has gotten worse. Maybe it's due to the increase in machine learning. Like how you can read text from a Marchov string generator and it sort-of look like a really book or what ever source text, but It's actually not meaningful. If the search results are just trying to mimic what people are searching for and not actually trying to identify the concept of the query it would fallow that the results are not human-meaningful. For example, if I search of chocolate cake and I get pictures of chocolate and cake but not chocolate cake I'm really not getting back what I'm looking for. It seems the algorithms have gone in the direction of just throwing a bunch of suggestions out without trying to match the query correctly.

All day I've been looking for this video, I think it was on Internet Archive, of lecture about programming and playing the electric clarinet. The lecture was about how music has changed it's timing abstraction a few times thought history and this can be metaphoric for how to write programs that use dependency injection. So the same way a piece of music can change timing systems so can your software change what it operates on while not changing the core logic. The guy doing the lecture looked like KingCobraJFS in his 50 or 60s. I really wish I could find it.

 

[LatinasAreTheFuture]

You get different results depending on what you use (eg. using Google compared to using any of the Searx instances)

The problem is that a lot of search engines lack their own index and rely on other big ones for their search results (i.e. the ones doing all the hiding and misinformation in the first place.

Here's a link regarding it if you're interested.

That was pretty interesting!

 

[Smaug's Smokey Hole]

I think search has gotten worse. Maybe it's due to the increase in machine learning. Like how you can read text from a Marchov string generator and it sort-of look like a really book or what ever source text, but It's actually not meaningful.

For many years I've told people "just type your question into google [and stop calling me]" and google used to be good enough to give relevant results when doing that, maybe machine learning aquired stupidity because people asked too many dumb questions and it had to come up with answers recursively based on what links moron clicked on.

 

[stares at error messages]

For many years I've told people "just type your question into google [and stop calling me]" and google used to be good enough to give relevant results when doing that, maybe machine learning aquired stupidity because people asked too many dumb questions and it had to come up with answers recursively based on what links moron clicked on.

If you don't know what your looking for and just want a plausible rationalisation to satisfy a question, then maybe google is acceptable. I'm not looking for any results, I'm looking for correct results.

There is no such thing as "what good for the goose is good for the gander." Ganders exist so that when the English come and start eating geese, the few geese that do survive have a buffer for the English to eat before they get cooked. Evolution preferences replication of what ever is level can be tolerated because it's very expensive to start back for nothing; anything surviving is better than nothing. It's a bad idea to take search engine pointers for cellular automata. When I search for "flying cats" I want airlines that allow pets not floating fur-balls with cat-like faces. No such creature exists, except in the minds of photoshop-trannies.

 

[⠠⠠⠅⠑⠋⠋⠁⠇⠎ ⠠⠠⠊⠎ ⠠⠠⠁ ⠠⠠⠋⠁⠛]

To some extent, is this because of the expansion and democratization of SEO?

A few select people used to make a lot of money doing this stuff. But nowadays, any random pajeet can do it, know what they can get away with, know what will start to knock them back.

The fact that one can sometimes get better results from a search engine with a lot less investment like Yandex or Bing can get better results than Google can suggests the fact that the web has been perverted to optimize for sitting high up in the Google results may be part of this.

 

[Dick Justice]

To some extent, is this because of the expansion and democratization of SEO?

A few select people used to make a lot of money doing this stuff. But nowadays, any random pajeet can do it, know what they can get away with, know what will start to knock them back.

The fact that one can sometimes get better results from a search engine with a lot less investment like Yandex or Bing can get better results than Google can suggests the fact that the web has been perverted to optimize for sitting high up in the Google results may be part of this.

SEO was an inevitability. Search, and particularly google is the portal to (supposedly) all knowledge. That's too tempting not to try to manipulate, and too exposed to be impervious to manipulation whether by hook or by crook. Ergo are we better served by democratized SEO, or would we have been better off if SEO was only available to major governments and the biggest of business?