privacy-focused users who don’t want “AI” in their search are more likely to use DuckDuckGo
But the opposite is also true. Maybe it’s not 90% to 10% elsewhere, but I’d expect the same general imbalance because some people who would answer yes to ai in a survey on a search web site don’t go to search web sites in the first place. They go to ChatGPT or whatever.
It’s not that wild of a conspiracy theory. Hard to get definite proof though because you would have to compare actual search results from the past with the results of the same search from today, and we unfortunately can’t travel back in time.
But there are indicators for your theory to be true:
It’s evident that in UI design the top area of the screen is the most valuable. AI results are always shown there. So we know that selling AI is of utmost importance to Google.
The Google search algorithm was altered quite often over the years, these “rollouts” are publicly available information, and a lot of people have written about the changes as soon as they happened.
Page ranking fueled a whole industry which was called SEO (Search Engine Optimization). A lot of effort went into understanding how google ranks its results. This was of course done with a different goal in mind but the conclusions from this field can be used to determine if and how search results got worse over time
It’s an established fact that companies benefit from users never leaving the company’s ecosystem. Google as an example tried to prevent a clickthrough to the actual websites in the past, with technologies like AMP or by displaying snippets.
If users rely on the AI output Google can effectively achieve this: the user is not leaving the page and Google has full control over what content the user sees.
Now, all of the points listed above can be proven. If you put all of that together it seems at least highly likely that your “conspiracy theory” is in fact true.
Have you seen the quality of google searches the last few years? I’m not surprised at all. LLM might not give you the correct answer but at least it will provide you with one lol.
Oh im definitely thankful for that and personally i dont use google, but alas many people are not tech savvy enough to switch to a different search engine if they even know that others exists.
Most people don’t even know the difference between an URL bar and a search bar, or more precisely: most devices use a browser that deliberately obfuscates that difference.
For some issues, especially related to programming and Linux, I feel like I kinda have to at this point. Google seems to have become useless, and DDG was never great to begin with but is arguably better than Google now. I’ve had some very obscure issues that I spent quite some time searching for, only to drop it into ChatGPT and get a link to some random forum post that discusses it. The biggest one was a Linux kernel regression that was posted on the same day in the Arch Linux forums somewhere. Despite having a hunch about what it could be and searching/struggling for over an hour, I couldn’t find anything. ChatGPT then managed to link me the post (and a suggested fix: switching to LTS kernel) in less than minute.
For general purpose search tho, hell no. If I want to know factual data that’s easy to find I’ll rely on the good old search engine. And even if I have to use an LLM, I don’t really trust it unless it gives me links to the information or I can verify that what it says is true.
I agree that happens, but it has nothing to do with what op said. They didn’t want a solution, they wanted a link to where the problem was being discussed so they could work out a solution.
People seem to really confure the difference between asking an llm how to patch a boat vs where did people discuss ways to patch a boat.
Most likely because if they came directly with their problem to whatever platform you are on, they would have been scolded at for not trying hard enough to solve it on their own. Or close the post because it has already been asked.
Yup this is a great example. LLM for non opinion based stuff or for stuff that’s not essential for life. It’s great for finding a recipe but if you’re gonna rely on the internet or an LLM to help you form an opinion on something that requires objective thinking then no. If I said hey internet/LLM is humour good or bad, it would insert a swayed view.
It simply can’t be trusted. I can’t even trust it return shopping links so I have retreated back to real life. If it can’t play fair I no longer use it as a tool.
How often do you check the summaries? Real question, I’ve used similar tools and the accuracy to what it’s citing has been hilariously bad. Be cool if there was a tool out there that was bucking the trend.
Yeah, we were checking if school in our district was canceled due to icy conditions.
Googles model claimed that a county wide school cancellation was in effect and cited a source. I opened, was led to our official county page and the very first sentence was a firm no.
It managed to summarize a simple and short text into its exact opposite
I also sometimes use the Kagi summaries and it’s definitely been wrong before. One time I asked what the term was for something in badminton and it came up with a different badminton term. When I looked at the cited source, it was a multiple choice quiz with the wrong term being the first answer.
It’s reliable that I still use it, although more often to quickly identify which search results are worth reading.
I can’t speak for the original poster, but I also use Kagi and I sometimes use the AI assistant, mostly just for quick simple questions to save time when I know most articles on it are gonna have a lot of filler, but it’s been reliable for other more complex questions too. (I just would rather not rely on it too heavily since I know the cognitive debt effects of LLMs are quite real.)
It’s almost always quite accurate. Kagi’s search indexing is miles ahead of any other search I’ve tried in the past (Google, Bing, DuckDuckGo, Ecosia, StartPage, Qwant, SearXNG) so the AI naturally pulls better sources than the others as a result of the underlying index. There’s a reason I pay Kagi 10 bucks a month for search results I could otherwise get on DuckDuckGo. It’s just that good.
I will say though, on more complex questions with regard to like, very specific topics, such as a particular random programming library, specific statistics you’d only find from a government PDF somewhere with an obscure name, etc, it does tend to get it wrong. In my experience, it actually doesn’t hallucinate, as in if you check the sources there will be the information there… just not actually answering that question. (e.g. if you ask it about a stat and it pulls up reddit, but the stat is actually very obscure, it might accidentally pull a number from a comment about something entirely different than the stat you were looking for)
In my experience, DuckDuckGo’s assistant was extremely likely to do this, even on more well-known topics, at a much higher frequency. Same with Google’s Gemini summaries.
To be fair though, I think if you really, really use LLMs sparingly and with intention and an understanding of how relatively well known the topic is you’re searching for, you can avoid most hallucinations.
I use Perplexity for my searches, and it really depends on how much I care about the subject. I heard a name and don’t know who they are? LLM summary is good enough to have an idea. Doing research or looking up technical info? I open the cited sources.
For others here, I use kagi and turned the LLM summaries off recently because they weren’t close to reliable enough for me personally so give it a test. I use LLMs for some tasks but I’m yet to find one that’s very reliable for specifics
You can set up any AI assistant that way with custom instructions. I always do, and I require it to clearly separate facts with sources from hearsay or opinion.
it just makes it evermore obvious to them how many people in their life are sheep that believe anything the read online, i assume?
a false sense of confidence where one mught have just said 'i dont know"
I think the attitude of being virtuosos or preachy can seep in at times, especially when being part of a cause but IMO diplomacy, having conversations and opening their mind to objectivity has to be better than telling people they are wrong.
I know this is easy to say, and esp when so many people are just so addicted to social media and the internet.
I have had conversations with friends, family where they can have a clear conversation about how much propaganda is pushed on to them, and they then turn straight to their phone and hoover up and hour of FB. It does make you think wow sheep. But I have to remind myself we don’t get change by telling people ‘you clearly don’t know your own mind’
The main problem is that LLMs are pulling from those sources too. An LLM often won’t distinguish between highly reputable sources and any random page that has enough relevant keywords, as it’s not actually capable of picking its own sources carefully and analyzing each one’s legitimacy, at least not without a ton of time and computing power that would make it unusable for most quick queries.
Genuinely, do you think the average person tiktok’ing their question is getting highly reputable sources? The average American has what, a 7th grade reading level? I think the LLM might have a better idea at this point
First, its results are often simply wrong, so that’s no good.
Second, the more people use the AI summaries, the easier it’ll be for the AI companies to subtly influence the results in their advantage. Think of advertising or propaganda.
This is already happening btw, and the reason Musk created Grokipedia. Grok (and even other llm’s!) already use it as a “trusted source”, which it is anything but.
Okay but its a search engine, they can literally just pick websites that align with a certain viewpoint and hide ones that don’t, Its not really a new problem. If they just make grokpedia the first result then its not like not having the AI give you a summary changed anything.
The article already notes that
But the opposite is also true. Maybe it’s not 90% to 10% elsewhere, but I’d expect the same general imbalance because some people who would answer yes to ai in a survey on a search web site don’t go to search web sites in the first place. They go to ChatGPT or whatever.
It still creeps me out that people use LLMs as search engines nowadays.
That was the plan. That’s (I’m guessing) why the search results have slowly yet noticeably degraded since Ai has been consumer level.
They WANT you to use Ai so they can cater the answers. (tin foil hat)
I really do believe that though. Call me a conspiracy theorist but damn it, it fits.
SEO has been fucking up searches long before LLMs were a thing.
You mean Google.
And Bing, and searches that use google and Bing results (DDG, ecosia)
It’s not that wild of a conspiracy theory. Hard to get definite proof though because you would have to compare actual search results from the past with the results of the same search from today, and we unfortunately can’t travel back in time.
But there are indicators for your theory to be true:
Now, all of the points listed above can be proven. If you put all of that together it seems at least highly likely that your “conspiracy theory” is in fact true.
Search results have been degrading for a lot longer than LLMs have been a thing. Peak usefulness for them was around a decade ago.
They WANT you to use Ai so they can
cater the answerssell you ads and stop you from using the internet.Have you seen the quality of google searches the last few years? I’m not surprised at all. LLM might not give you the correct answer but at least it will provide you with one lol.
Thankfully Google is not the only search provider.
Oh im definitely thankful for that and personally i dont use google, but alas many people are not tech savvy enough to switch to a different search engine if they even know that others exists.
Most people don’t even know the difference between an URL bar and a search bar, or more precisely: most devices use a browser that deliberately obfuscates that difference.
For some issues, especially related to programming and Linux, I feel like I kinda have to at this point. Google seems to have become useless, and DDG was never great to begin with but is arguably better than Google now. I’ve had some very obscure issues that I spent quite some time searching for, only to drop it into ChatGPT and get a link to some random forum post that discusses it. The biggest one was a Linux kernel regression that was posted on the same day in the Arch Linux forums somewhere. Despite having a hunch about what it could be and searching/struggling for over an hour, I couldn’t find anything. ChatGPT then managed to link me the post (and a suggested fix: switching to LTS kernel) in less than minute.
For general purpose search tho, hell no. If I want to know factual data that’s easy to find I’ll rely on the good old search engine. And even if I have to use an LLM, I don’t really trust it unless it gives me links to the information or I can verify that what it says is true.
I’m seeing almost daily the fuck-ups resulting from somebody trying to fix something with ChatGPT, then coming to the forums because it didn’t work.
I agree that happens, but it has nothing to do with what op said. They didn’t want a solution, they wanted a link to where the problem was being discussed so they could work out a solution.
People seem to really confure the difference between asking an llm how to patch a boat vs where did people discuss ways to patch a boat.
Most likely because if they came directly with their problem to whatever platform you are on, they would have been scolded at for not trying hard enough to solve it on their own. Or close the post because it has already been asked.
Yup this is a great example. LLM for non opinion based stuff or for stuff that’s not essential for life. It’s great for finding a recipe but if you’re gonna rely on the internet or an LLM to help you form an opinion on something that requires objective thinking then no. If I said hey internet/LLM is humour good or bad, it would insert a swayed view.
It simply can’t be trusted. I can’t even trust it return shopping links so I have retreated back to real life. If it can’t play fair I no longer use it as a tool.
I use kagi assistant. It does a search, summarizes, then gives references to the origin of each claim. Genuinely useful.
How often do you check the summaries? Real question, I’ve used similar tools and the accuracy to what it’s citing has been hilariously bad. Be cool if there was a tool out there that was bucking the trend.
Yeah, we were checking if school in our district was canceled due to icy conditions. Googles model claimed that a county wide school cancellation was in effect and cited a source. I opened, was led to our official county page and the very first sentence was a firm no.
It managed to summarize a simple and short text into its exact opposite
Depends on how important it is. Looking for a hint for a puzzle game: never. Trying to find out actually important info: always.
They make it easy though because after every statement it has these numbered annotations and you can just mouse over to read the text.
You can chose different models and they differ in quality. The default one can be a bit hit and miss.
I also sometimes use the Kagi summaries and it’s definitely been wrong before. One time I asked what the term was for something in badminton and it came up with a different badminton term. When I looked at the cited source, it was a multiple choice quiz with the wrong term being the first answer.
It’s reliable that I still use it, although more often to quickly identify which search results are worth reading.
I can’t speak for the original poster, but I also use Kagi and I sometimes use the AI assistant, mostly just for quick simple questions to save time when I know most articles on it are gonna have a lot of filler, but it’s been reliable for other more complex questions too. (I just would rather not rely on it too heavily since I know the cognitive debt effects of LLMs are quite real.)
It’s almost always quite accurate. Kagi’s search indexing is miles ahead of any other search I’ve tried in the past (Google, Bing, DuckDuckGo, Ecosia, StartPage, Qwant, SearXNG) so the AI naturally pulls better sources than the others as a result of the underlying index. There’s a reason I pay Kagi 10 bucks a month for search results I could otherwise get on DuckDuckGo. It’s just that good.
I will say though, on more complex questions with regard to like, very specific topics, such as a particular random programming library, specific statistics you’d only find from a government PDF somewhere with an obscure name, etc, it does tend to get it wrong. In my experience, it actually doesn’t hallucinate, as in if you check the sources there will be the information there… just not actually answering that question. (e.g. if you ask it about a stat and it pulls up reddit, but the stat is actually very obscure, it might accidentally pull a number from a comment about something entirely different than the stat you were looking for)
In my experience, DuckDuckGo’s assistant was extremely likely to do this, even on more well-known topics, at a much higher frequency. Same with Google’s Gemini summaries.
To be fair though, I think if you really, really use LLMs sparingly and with intention and an understanding of how relatively well known the topic is you’re searching for, you can avoid most hallucinations.
I use Perplexity for my searches, and it really depends on how much I care about the subject. I heard a name and don’t know who they are? LLM summary is good enough to have an idea. Doing research or looking up technical info? I open the cited sources.
For others here, I use kagi and turned the LLM summaries off recently because they weren’t close to reliable enough for me personally so give it a test. I use LLMs for some tasks but I’m yet to find one that’s very reliable for specifics
You can set up any AI assistant that way with custom instructions. I always do, and I require it to clearly separate facts with sources from hearsay or opinion.
lol, the random text generator does not understand what any of those things are.
what makes it creepy?
it just makes it evermore obvious to them how many people in their life are sheep that believe anything the read online, i assume? a false sense of confidence where one mught have just said 'i dont know"
What an absolutely arrogant attitude 🤣 You actually believe there is some gap here 🤣 just amazing.
Not using AI doesn’t mean your performing whatever task your doing better. It has nothing to do with being able to parse results for bullshit or not.
I think the attitude of being virtuosos or preachy can seep in at times, especially when being part of a cause but IMO diplomacy, having conversations and opening their mind to objectivity has to be better than telling people they are wrong.
I know this is easy to say, and esp when so many people are just so addicted to social media and the internet.
I have had conversations with friends, family where they can have a clear conversation about how much propaganda is pushed on to them, and they then turn straight to their phone and hoover up and hour of FB. It does make you think wow sheep. But I have to remind myself we don’t get change by telling people ‘you clearly don’t know your own mind’
So many people were already using tiktok or youtube as google search. I think AI is arguably better than those
edit: New business, take your chatgpt question and turn it into a tiktok video. The Slop must go on
The main problem is that LLMs are pulling from those sources too. An LLM often won’t distinguish between highly reputable sources and any random page that has enough relevant keywords, as it’s not actually capable of picking its own sources carefully and analyzing each one’s legitimacy, at least not without a ton of time and computing power that would make it unusable for most quick queries.
Genuinely, do you think the average person tiktok’ing their question is getting highly reputable sources? The average American has what, a 7th grade reading level? I think the LLM might have a better idea at this point
First, its results are often simply wrong, so that’s no good. Second, the more people use the AI summaries, the easier it’ll be for the AI companies to subtly influence the results in their advantage. Think of advertising or propaganda.
This is already happening btw, and the reason Musk created Grokipedia. Grok (and even other llm’s!) already use it as a “trusted source”, which it is anything but.
So literally the same shit as before with search but wrapped up in a nice paragraph with citations you can follow up on?
Okay but its a search engine, they can literally just pick websites that align with a certain viewpoint and hide ones that don’t, Its not really a new problem. If they just make grokpedia the first result then its not like not having the AI give you a summary changed anything.
I Prefer searx
Do you also use Arch btw?
Arch based on laptop , phone is grapheneos and lineage🤔
Nice