Maybe google images needs a toggle for AI generated content. It won’t be perfect, but it should filter out a chunk of it just by excluding pages that have it
Upping the original as a matter of course.
Sauce: https://the.me/making-the-iconic-1989-tank-man-photo/
Look, all I’m saying is that I have yet to see any definitive proof that China really exists. A flag with 5 stars and 2 colors? Who do they expect to fool with that?
Also the borders are so inconsistent. One minute it’s 9 dashes, next they’re claiming it’s 10 dashes. Can’t even keep a story straight.
I saw a picture of China’s purported leader but it was just Winnie the Pooh.
You can’t fool me with cartoons.
Google’s becoming pretty terrible anyway, it only seems to return pages that are selling things. I’ve switched to Kagi at this point and it seems to work better, it’s subscription only, but you know you’re the one paying for it and that means that you’re the end customer.
I wonder if the verb Google will stay with us when its origin is lost in history
New Kagi user here too, been very happy so far. Though it turns out I do a lot of searching and blew through the 300 searches in the $5 plan in like 2 weeks…
I just signed up with them, too. They were able to get me a link to the age of a small, nondescript lava flow near my town on the third hit. (5000 years old! A youngster!) All the other search engines gave me unrelated crap.
I have a hard stop set up for when I hit $10, so I’ll switch tiers if it comes to that. 😅
I don’t necessarily like paying for search, but I couldn’t take ad-driven search any longer. Big waste of time getting through the chaff.
Make an effort to use bangs and I bet you’ll stay under the limit. Edit: bang searches don’t count towards the limit
Knowing I wanted a result from a certain site but using the search engine to get there was a (bad) habit I brought over from Google.
!imdb barbie
!w mattel
There’s even custom bangs, which is something DDG doesn’t give you: !libgen some book
Because last time I checked they just used Bing anwyay, while Kagi runs their own indexer.
Another “success” of CCP technology
When will people realise that google has tailored algorithms and we are not all experiencing the same search results?
The first thing you’ll see if you search Google for “tank man” right now will not be the iconic picture of the unidentified Chinese man who stood in protest in front of a column of tanks leaving Tiananmen Square, but an entirely fake, AI-generated selfie of that historical event.
No, this is the first thing the author saw. Probably because they are a journalist writing about AI.
When I google tank man I don’t even get the AI image on the first page. The top result is from history.com. If I go to google image search it is the 7th result on the page. The top result is from wikipedia.
When will people realise that google has tailored algorithms and we are not all experiencing the same search results?
You’re right. This is the real problem with search engines like Google and one reason I use SearXNG instances and Mojeek instead. Where I live, the algorithm is more likely to net content that is biased toward right-wing conspiracy theories and problematic agencies because of that algorithm. Any search engine that does this is not a valid search engine, in my opinion.
I suppose it’s invalid in the context of showing you what you exactly searched for. But it’s pretty valid in the context of showing you what you’re looking for. For example, someone with a disdain for science when searching for the terms “big bang” or “evolution” is probably not looking for scientific articles detailing the rigor of the prevalent theories. If the point of a search engine is to find what you’re looking for, it’s pretty effective by that measure. It just so happens that what you’re looking is biased in its own ways.
Rather, educating people how to use keywords and syntax is far better than teaching people to depend on an algorithm. This would net one the results that you describe without any of the problematic aspects of an algorithm.
I don’t see any need for an algorithm whatsoever, but I see many ways it can be used to frustrate or manipulate users. It is my strong opinion that a valid search engine only responds to deliberate input from the user, and things like algorithms or location-specific results are an endless source of discouragement and frustration to me.
Yes I had a family member in a right wing conspiracy area. It was infuriating because his friends would tell him their nonsense and he would be skeptical and google it, only for google to seemingly support what they were saying.
I couldn’t replicate his results at all and it would take a lot of searching to even find what he was talking about so I could debunk it for him.
I became hyper aware of it the first time I tried using Tik Tok, and I was served nothing but alt-right hate videos. Obviously, search engines aren’t usually quite so obvious, but for instance, people in my location are certainly more likely to net results connected to far-right ideas and agencies, because of the interests of people who live here.
I don’t like that idea at all. Search engines should only respond to deliberate input from the user imo. I know that’s a big ask for people to acclimate to appending “in [location name]” if they expect location specific results, but the convenience of just saying “hey Google, restaurants near me” is not worth the consequences.
Unfortunately, almost all search engines are complicit, including supposedly privacy friendly ones like Kagi, Qwant, and Startpage. I’m no longer recommending those to people. SearXNG and Mojeek are the only ones that don’t lean into the algorithmic and locational fuckery, but even that’s with a lot of tweaking the settings.
It’s no great mystery why things like fascism are on the rise. And people will say I’m in the minority for caring about this, and … yeah, that’s the problem.
