Not sure if this is just me, but Google Search results are getting less and less relevant for my queries and at the same time, provide more and more spammy sites as results.
Finding a useful site on page 2 and beyond? Good luck!
I've switched to DDG half a year ago or so and haven't looked back much. I only once in a while use the !g bang and almost always come back disappointed.
Sure, there are also some spam results on DDG (and probably there are going to be many more once it gains more relevance to be spammed) but for now and my searches i exclusively use DDG on my private computers as well as at work...
So, to summarize, i think the relevancy of Google as a search engine will, imho, continue to drop and also the perceived "truthiness" of statements found thereon as well.
To add, i also don't necessarily think of this as a "Google" problem - the same people that will believe in a statement gathered from a quick search will also believe in the same thing when they see it in a Facebook post, Whatsapp chainmail or whatever....
Personally I find other search engines (DDG included) quite impossible to use, the results just are not what I was looking for, while Google is an always first page hit. Engines like Bing just give me straight up crap which isn't even related to (seemingly ofcourse; it somehow did find a relation obviously) what I am searching for.
I use web search almost exclusively for compile issues, coding searches and whitepaper searches. I would like to use something else but it would be very inefficient for the work I do at this moment.
What I noticed as well (and is sommewhat related); I work in China a lot and my colleagues there have no (real; they have it when they need to but they don't use it like we do) access to Google and, when solving hardware/software problems they are far less efficient than us. Thye have an advantage (I find) that they memorize a lot more than we do, but when we just need to solve something that they did not encounter before, Google always delivers while Baidu, Bing etc do not even get close.
Even if you often use Google search, you should switch the default to DDG and make the Google search command explicit so that Google isn't always "listening" to everything you type into the URL text field. Otherwise, even when you think you are just entering the complete URL and not doing a search, you might be sending it character by character to Google as you enter it. A Google employee I know warned me years ago that they were always studying what you typed, including keystroke timing, accuracy, backspaces, everything including what URLs you were going to without needing a search, to see how much they could learn about you, even when you weren't (knowingly) doing a Google search.
You could just turn off autosuggest, but then unintended items (ex: typos, things that don't match the browser's "complete URL" regex pattern) would still be sent to Google. Best to not make Big Brother your default listener in the first place.
Or you know use a browser respects your privacy, I don’t trust chrome isn’t sending my data to google even after changing my default. Its really handy that the same browser keeps the search and url bars separate so the intent it never mixed, which is really handy if you’re not working with public sites with FQDNs all the the time.
I rarely notice a difference in quality between Google and DDG; usually DDG's widgets are more useful (stack overflow snippets, etc), and Google's search seems slightly better at using context to pick the right idea behind some overloaded search term. For example, if I search "nix provision", I want results about provisioning a system with Nix, and Google does this correctly while the top results for DDG are "Congress nixes provision..." and "Virginia may nix key provision..." and "provisioning a nix node"; DDG isn't far off--the correct results are still on the first page, and this query is understandably challenging, but Google seems to know that if I'm searching for "nix" and "provision", it should prefer entries for which those terms are significant. In other words, the words "nix" and "provision" are very significant in articles about provisioning with nix; they're not* very significant in articles about trimming legislation.
That said, I still prefer DDG because much of my searching is code, DDG seems less likely to snoop, and Google's handling of the Google Memo seemed sufficiently authoritarian to make me prefer suitable rival products where available.
works. More tips under "duckduckgo search syntax".
This said, I think google is more accurate. But duckduckgo works well enough and is my default search engine now (Google lost me with THE ONE GOOGLE aggregation since quite a while ago)
DuckDuckGo even informs you of their changed query under the assumption of a spelling mistake, and offers to search, in quotes, exactly what you typed instead. Right at the top of the page.
> but Google Search results are getting less and less relevant for my queries and at the same time, provide more and more spammy sites as results.
I'm getting many sites that are missing one or more of my keywords listed above sites that do have all of them. I get this odd feeling an AI is trying really hard to draw inferences from my search patterns and is failing horribly to understand what it is I am asking for most of the time.
When that doesn't work, I type the desired word three times after the search e.g. "Old books for sale old old old" -if, say, all the results are dominated by new books due to "books for sale" being so strongly attached to those websites.
If that doesn't work I type "BING" and don't return for a while.
If Google's algorithm doesn't keep track of frustrated searches for BING I will eat my hat.
if I correctly understand what you're up to, it's brilliant. this pattern of leaving "well f* you then!" breadcrumbs for behavioral-pattern matchers of the future is something I'm quite fond of doing myself.
here's hoping some future intelligent system will be a lot smarter and more capable of tracking the breadcrumbs of my evidently-permanent log-trail of passive-aggression than the human UX-dillholes who are currently running things.
At what point does google admit g+ has failed and stop crippling their primary search tool to accommodate it? I know "you are the product" etc, but if the tool becomes frustrating to use and someone else can provide a better user experience, it's going to cut into those ad dollars.
> I've switched to DDG half a year ago or so and haven't looked back much
Same for me. I specially like their approach to technical documentation.
Search for some web development stuff and DuckDuckGo shows you the MDN site while Google points to some outdated (and low quality and spammy) w3schools site.
Ask a question that is answered on StackOverflow? DuckDuckGo points you to StackOverflow, and even shows relevant parts of the answer. While Google points to some spam sites that are a bad rip-off of the StackOverflow content.
Google has personal blocklist addon for chrome that you can use to block unwanted sites from the search results.
I just wonder why this obvious tool is not account-level functionality that works across browsers and devices. If it is because the sites that people block bring revenue to Google, then Google truly is evil.
Even if that features worked, it would still be user hostile to force them solve a global problem locally.
If Google really cared about user experience, they would already have downscored all those low-quality sites years ago. But they don't - neither now nor back then. So either they are profiting from those low-quality sites, or they simply don't care.
The more I think the more intentional this feels. I mean, if Google really wanted to filter out low quality websites out form their search results, I can't imagine much more reliable algorithm than to ask from the users to indicate the results that are bad.
Oh, w3schools. I hate that they are always near the top of Google searches for anything even remotely web related, when what I really want is the official spec or MDN.
Just tried it right now - the top three search results for "html canvas" are all w3schools on google. They're still the top in DDG, but there's only one entry as opposed to three. Bing also only has one, but they also steal the "here's your answer" type card text from w3schools.
Congratulations on your SEO, but f' you for polluting search results.
There was a way to remove results from Google 5 years ago that worked and I've used before, but now you need to use a chrome plugin apparently this says. Post comments afterwards are still a few years old but may still apply today
Isn't exactly that the job of a search engine? I.e., that don't have to specify the exact site? Moreover, many searches are done be beginners who don't (yet) the best resource for what they search for.
But let's assume you use a search engine only for stuff you aren't intimately familar with - how could you trust the quality of the results? If they are so bad at the topics you know much about - why should they be better at the other topics?
Another way to look at it is people are expressing a need to have a single good search tool rather than having to learn every individual site's search tool. Site hosted search varies wildly in quality too since many don't get much direct use (perhaps largely because most people uses general search tools like Google). I know I've built up a (usually unconscious) decision making bias that "search on site always sucks" that makes me shy away from them in. It feels like a reasonably rational decision to do so vs being garden-variety lazy.
But now I have a better workaround which doesn't only work for mdn, but any other high-quality resource: I edited my browser settings to use a different search engine by default. (currently DDG)
It's a bit more work but, alternatively, you could also add a "new" search engine to your browser that uses a custom keyword (e.g. "!mdn") and passes your query to the search interface on the MDN site.
> Not sure if this is just me, but Google Search results are getting less and less relevant for my queries and at the same time, provide more and more spammy sites as results.
Personally I noticed that a few years back there was always talk of changes to Googles' algorithms to de-rank automatically generated SEO/ad pages, yet this exact kind of site dominate search results in many topics. This makes Google virtually useless for them.
On the other hand, sometimes Google still manages to dig up some information that I'd never expected to find. For example, sometimes when searching a part number Google will direct you to a page that doesn't even mention that number, yet is still about the same part (under a different label) or mentions it - rather impressive. However, most of the time the majority of search results are useless sites that just have endless lists of everything to get traffic and show ads.
Google also became notoriously bad in discerning (or rather susceptible to manipulation?) which page(s) are "official" (which should rank very high for certain terms) and which are not.
It's so frustrating trying to evaluate SaaS products for a particular domain when all of the search results are affiliate listicles and sneaky first party marketing.
Was going to comment the same thing. Lately, if I, say, search for four words, the top result will only contain one of those four words. Almost feels like being back in 1996.
This is one of my pet htes with google at the moment. They helpfully tell which of the search terms are missing in each result, and every time I see that "missing" tag, I am starting to draft an email in my mind:
"Dear Google AI. Why do you think I wrote that word into my query? Best Regards, Beefield."
(Second one is that there is no way to make specific queries to return results only in a certain language (and even adding words in a specific language does not help because it tries to translate them), but you need to change the settings for the whole google search)
Try searching for things like part numbers and datasheets for obscure electronic components, along with schematics and service manuals. That's one area where Google's search results used to be far better than now.
I've almost become accustomed to always putting my queries in quotes, just because it would otherwise mean completely useless results.
The other thing that's gotten worse results is... adult content, especially when searching for very specific terms; and even with "SafeSearch" off, which is supposed to mean no filtering.
To add insult to injury, repeating minor variations of a query, combing through "all" the results (or as many as Google is willing to show you), and in general trying to extract what you're actually looking for also tends to get you blocked with a CAPTCHA or just completely banned for a while.
> To add insult to injury, repeating minor variations of a query, combing through "all" the results (or as many as Google is willing to show you),
Oh yeah. When Google proudly proclaims that there are about sixty quintillion results. And the thing you want isn't on the front page, but you know it's there somewhere if it was ever indexed. You start combing through the results. 25 pages in, no more results.
The most obvious change I have noticed is that they are far more aggressive about dropping terms from the query. I don't this has ever produced a useful result for me, I always have to go back to the search and add quotes around the dropped term.
...and the most irritating part is that it's almost always omitting the term that is most important because it's the one that reduces the number of results the most, under the (ridiculously flawed) assumption that more results are better; and in that case, why not go the whole hog and make everything "suggestion"-based? People are searching for what they want, and specifying terms to get that.
To make an analogy, it'd be like going into a pet shop, looking for a specific breed of dog, and the owner trying to convince you to buy cats instead. "But you were just looking for a pet, right? What's the difference?" A pretty idiotic situation.
I've switched to Bing, which was a step I wouldn't have even entertained five years ago.
Google needs to get back to its minimalist beginnings. It's really apparent that the advertising department has taken over while the search department atrophies.
In retrospect, I think the turning point was when they murdered Labs and Reader.
I have the same experience I feel like Google is becoming less and less useful for finding things. There have been several times where I tried to find things I've come across before and Google was no help.
However it's still the best search engine hands down. There is no way a search engine like DDG could provide better results. If Bing can't do it there is no way DDG can.
It would like it if there was a comparable alternative to Google but there really isn't.
I think it would be cool if some system or protocol (eg. IMAP, RSS, OStatus) came along to provide a way to create a user programmable search engine, that's decentralized but comparable to Google.
Why do you think there is no open and free search engine? I don't think it's because of hosting costs, because a distributed search engine could be run with local hardware.
A search engine is many things in a giant package. Most wholesale search companies have a database text/images and algorithm data that is in the terabytes if not more. They also have many servers for fast and efficient processing of queries and giving results.
This database could be made freely available (see CommonCrawl) but without a datacenter big enough to store and run queries on it, it's meaningless as a "open" search engine.
A distributed search engine runs into all of the scalability and quality of service problems that researchers of Peer-to-Peer Information Retrieval have been trying to solve since the early days of the internet.
Even putting politics & world events aside, Google Search doesn't seem to get better.
For me the biggest disappointment is when they turned off the ability to search for "an exact match within quotes", as well as the ability to use +plus and -minus to refine results. (EDIT: could be wrong about that, I thought it was disabled at some point)
But the most dumbfounding is the fact that you still get several year old results whenever you search for a bug, hardware/driver issue, or even simple programming questions related to a not too old framework or API. Every time I have to go into the advanced options to restrict the results to one year, or one month.. you'd think they would make recent activity the default and then let the user know they can "search older results".
My bad. I could swear it was removed at some point but I can't find articles.
A quick test reveals maybe the change is that you can't do an exact "ascii" search, ie. google searches for an exact sequence of words, but the casing may be different , there-may-be-dashes and so on.
There's been so many times that old results have saved my bacon. I like that they give me everything and allow me to make the decision to narrow it down as needed. Sure, it can be annoying when you're searching for results that are relevant for recent versions of software or libraries but even in those situations old results can be great because they lead to solutions that are still applicable or relevant GitHub issues that lead you where you want to go.
That said, it would be nice if it wasn't buried and was exposed as a slider but I'll take what they're giving.
If I type "GitHub" in DDG, then the first result is the Wikipedia page, the second is some ad, and the third is the actual GitHub website. Seems in the wrong order if you ask me.
Anyway, I also use DDG, but I always end up using "!g". At least this gives DDG some metrics on how well they are doing.
Your commemt doesn't make much sense to me. If a random person is searching for github, the top result probably should be from wikipedia. Someone actually using github on the other hand, would just type the address.
If we expand that, however, to non-technical users; no, they don't always just type the address. They type "github" into the address bar, and click on the search result for the actual site.
This behavior completely derailed my mom when her search bar was hijaacked by some just-barely-not-malware addon installed in the browser. Suddenly, she couldn't reach any of her sites.
Frustrating as hell, but at this point there's nothing I can do to change her behavior.
I just checked on mobile, too. The first is a card listing the info from Wikipedia (is that what you meant?), the second is an ad, the third is GitHub, and the fourth is the Wikipedia entry. Seems like the right order.
Just for comparison I did this search on Google (UK site, logged in with my work account) as well, the first page of 'github' results are listed. It seems their are much more relevant and useful than the DDG results...
1. a clearly marked relevant ad at the top, for the circleci github plugin
2. an infobox at the right with details of the company, their web address, a link to their twitter account, and some related search queries
3. (first full-width result) link to github.com and summary text
4. a sub-search box to continue searching only on github.com
5. a short two-column list of six links to github.com sub sections
6. a link to the github twitter account and three square boxes across the page with its most recent tweets
Wow that's quite a load of useless crap, sensory information overload. For lack of a screenshot I imagine this is not presented in a consistent manner.
Apart from this inconsistent and confusing presentation, there is not much difference in actual results:
duckgo offers results all with the same consistent appeareance:
Actually, it's pretty consistently designed and laid out - that's not something you can fault Google for these days. Here's a screenshot of Google https://i.imgur.com/tiVr0bd.png to compare with DDG https://i.imgur.com/01Lkf2z.png (with no ad-blocking on mine) - interestingly, the top adverts are the same, both CircleCI, but I prefer the Google version, which has pricing as well, much more likely to lead to a click. Also, DDG has a second advert, under the infobox on the right, so actually 200% of the advertising that Google serves!
My guess is that you live in US. I live in Greece, and trying anything but Google for localized results is pretty much useless. The same goes for many European countries. That's why Google's share in Europe is above 90%. There's practically no alternative.
That’s also the very reason I switched to ddg: getting non local, non trending results in Google is a PITA. It’s a constant fight against the AI that tries to steer the results towards what 90% of the people in that country are looking for.
Google Search actually works really well for me, and I think it's been getting better over time. Just this week, a colleague asked me why the following code does not work:
Even a few years ago, answering an arcane question like this, where it is not clear what keywords you need to search for, would take you hours of googling. Most people would just not bother. But in this case, it took me literally 30 s to find an answer on Stack Overflow.
Google search does tend to still do well for areas where they see there's no money to be squeezed.
There's a stark difference if you try to search for something where Google can net a big cut of the outcome. Try searching for "vegas hotels". Once upon a time, google returned a healthy mix of results that directly benefited them, and ones that didn't.
Here's the portion of the page (yellow rectangle) that now appears to be presented to actually help visitors, and not directly pay Google: https://imgur.com/a/hAFas On my PC, you have to scroll down 2 pages to see that.
I don't see how you can justify the claim that a scrollable map with hotel locations and prices isn't "presented to actually help visitors".
Of course there's some synergy between helping visitors and profit to Google, that's how business models work. And if you think that following the links to hotels.com and tripadvisor.com will get you more altruistic and unbiased information than the Google links then I'd like to sell you this fine bridge with air conditioning, great views and swimming facilities, which 93% of visitors rated as 4 stars or higher.
Actually, I think the Google-provided map of hotels is a lot more useful to me as a search user than almost everything else likely to be returned from a spammy-hellhole-producing search term like this one.
I don't know how to make such search terms useful again. Maybe by aggressively delisting anything that reeks of SEO, has ads or otherwise wants to sell something. If that filters out everything but the personal homepage of Johnny S. Hotels of Las Vegas, then so be it.
> Google Search actually works really well for me, and I think it's been getting better over years. Just this week, a colleague asked me why the following code does not work:
I feel like you've posted this comment mid-thoughtstream. How is the compiler error relevant to Google Search?
Edit: > Even a few years ago, answering an arcane question like this, where it is not clear what keywords you need to search for, would take you hours of googling. Most people would just not bother. But in this case, it took me literally 30 s to find an answer on Stack Overflow.
Makes a bit more sense now you've edited that in.
Yeah, I'd totally agree. I've noticed it can help you find films based on description ("film with adam sandler and kate beckinsale") or code syntax ("c++ io operator") using terms that aren't really at all obviously connected to what you're trying to find out.
Personally I find the opposite, I always start with DDG, but sometimes it doesn't find what I'm searching for. Typically though the fallback with !g does work.
However for me that's fine, it removes most of my use of Google search.
But also the !g search is less personalised than the regular google search so you'll get less of the search bubble.
Even better is to always do the !g searches in a private browsing tab or use the Firefox containers addon to keep the tracking cookies separate. Then you're even less likely to get search bubble results.
I am the same. The fact is Google knows a fair bit about my searches and knows when I type Django that I mean the web framework (not the film). Duck Duck Go needs extra term in the search to know that I am looking for the software and not the film.
You know... I also had a hunch Google "profiles" me from my searches and because of that, seemingly returns better results.
But now we are back at the greater discussion... the feeling that your searches happen within a bubble. The odd realization that you haven't discovered anything new in a long time.
This has been discussed on HN many times. But usually people talk about how Google doesn't index the sites the same way.... But here we are pointing out a much more specific culprit: Google actually tries to help you, but does it? (I mean, with the SAME index it will show different results to different users. So the INDEX itself is affected by their algorithm, and then on top of that, Google filters and tweaks results to fit your own bubble.)
You know I may be walking around the block to look for new appartments... and maybe I stumble on a nice bookstore. With this "help" from Google, this doesn't happen anymore.
Anyone know if you can turn off this profiling entirely while reamining signed on?
( I have the same issue on YouTube... With YT I have got into the habit of using Incognito (ctrl shift N) for random youtube browsing.. that way when I go to the home page I get recommendations I tend to like, because I only allow certain content in my youtube history. )
Well yes, this is the expected behaviour of a search engine.
You ask for django and you get results for django[1]. When what you search has several possible meanings duckduckgo allows you to choose which one in the "meanings" top bar[2]. Thus preserving your freedom of choice and your privacy.
This is actually the whole point of duckduckgo, it has no filter bubble.
Then you also have all the usual search engine options: adding keywords, filtering out by keywords, etc.
I just put this to the test by entering "Django" into DDG and the search results page literally alternates between Django, the framework, and Django, the film name: Each odd result is about the framework, each even result is about movies of that name.
> To add, i also don't necessarily think of this as a "Google" problem - the same people that will believe in a statement gathered from a quick search will also believe in the same thing when they see it in a Facebook post, Whatsapp chainmail or whatever....
Yes! It's foolish to simply trust what you read. And that's especially so for contentious issues. Every source is biased, intentionally or not. And expecting Google to do the heavy lifting, and just trust what they give you, is just as foolish.
> I only once in a while use the !g bang and almost always come back disappointed.
Ha! My exact experience shows the same. I'm not finding something using DDG and think "well google will find it, surely", use !g and bam! I've got the same irrelevant results. The correct approach is to adjust the query but it's hard to fight the habit of trusting that Google knows better...
it's not just you, I have been having a similar experience. In fact I see goggle often turning up results in those answer boxes that come from unqualified sources. What's worse, when you click through them you get video autoplay ads.
I use ddg, and I dont like the results for anything not one dimensional. Now I can agree that google results are less appealing to me but still way above the pack.
I've used DDG since 2012 and I rarely go back to Google for text searches. The event that made me search (no pun intended) for an alternative was the key word limit in place to filter out the result. It seems that Google ranking engine doesn't promote links based on certain type of content, and that makes it difficult to use in one of my previous work environment.
> Not sure if this is just me, but Google Search results are getting less and less relevant for my queries and at the same time, provide more and more spammy sites as results.
This had been going on for over 15 years now, I remember a time where you could have over a dozen pages of relevant results then focus was put on first page and it was over.
Google always provides the "best results for you", this is called the filter bubble.
DDG just provides the best results.
I'm not interested the best results for you, I'm not even interested in the best results for me. I have no use for dictionary that removes the words I don't like, I want both sides of the coin.
I have no idea what you mean by this. I type something into google and I get a relevant result, exactly what I was looking for. I search DDG and... Sometimes I get what I'm looking for, other times I'm hunting through pages of results and eventually give up hope.
In what way is DDG better here exactly? I'm not buying it.
It's due to alphabet altering their "algorithm" to favor established companies and sources ( aka fight "fake news" and clean up the "interwebs" ). Google stopped being a faithful reflection of the internet. It's why now, any coooking and recipe searches now brings up nytimes links first rather than those that people clicked/linked/sourced.
It's not just google search. It's the entire alphabet ecosystem. Google news, youtube, etc all are getting worse. Google news was "cleansed" and now just peppers everyone with nytimes and wapo articles. If you have the time, waybackmachine and see how different googlenews is now compared to just a couple of years ago. And youtube trending is just awful as well. It doesn't reflect what people are watching. It's just a advertising platform for established news channels and late night shows.
It's the direction the internet is going. From user directed (bottom-up ) system to a moderator driven ( top-down ) system.
The government and establishment media has put so much pressure on social media/internet companies that they finally have given in.
Essentially, the age-old tactic of blaming someone of aiding the enemy ( aka russians/fake news/propaganda ) and "we must protect the children" was used to intimidate and cow tech companies.
I'm pretty sure Bing still makes up a large portion of their results. I remember reading some official stats at some point that their own crawler was such a miniscule part of the results that in effect you were just basically using a more limited version of Bing.
I've been using duckduckgo for several years and unless Bing had radically changed in recent times, I would consider Bing to be the more limited of the two.
I wanted to look up a translation and tried !wr word on bing and it did not load the wordreference.com page for word
I wanted to find a pizza for tonight and I'm not familiar with my current location, so I typed pizza near me in bing and it returned results for pizza chain store in US, Canada and India while I'm in western Europe.
Same search in duckduckgo gives me an openstreetmap based map centered on my location with addresses of places to get pizza and yelp reviews.
I could go on for a while with other examples showing that duckduckgo is definitely not a limited version of Bing.
Thanks. I would like to see numbers on how many of the search results actually come from their own crawler though. I may be wrong but it feels like most of them come from Bing or Yahoo.
> Why would DDG be any better? It's just Bing wrapped in incognito.
Turns out DDG is not "just Bing wrapped in incognito"[1] so you actually answered your own question here. But even if it were, the incognito part is enough to make much much better.
> DuckDuckGo emphasizes returning the best results, rather than the most results, and generates those results from over 400 individual sources, including key crowdsourced sites such as Wikipedia, and other search engines like Bing, Yahoo!, Yandex, and Yummly.
So basically they don't have their own crawler but aggregate results, mostly from Bing which is (or was) clear from comparing search results. However, a sibling comment claims they run their own spider now so my info may be outdated (quote from Wikipedia).
I find Google annoying to use ever since they removed the AND (+) operator (i.e. that search term MUST be present on the page).
Whenever I need a query that involves the AND, I always have to resort to Bing to get a result. Putting things inside quotation marks in Google doesn't help one bit - they're not a replacement.
Removing features and not replacing them causing irrelevant results is what's degrading Google for me.
This stopped working like this a long time ago; if you put something in quotes, Google will still sometimes return pages that don't include it, or that contain "synonyms" for it. You need to turn on the verbatim option for the old behaviour, I think.
This is something else that used to work - but for the past few years it only seems to make it more likely that the search term is in the page. Very annoying :/
"like" "so", or "like so"? Because the latter means that it should be present on the page, as well as in that order. I'm not sure about the former though.
The former forces both "like" and "so" to be present in the result without any further relation between them. It's awkward and unintuitive, but it was specifically introduced when they remapped +.
I miss the old cluster graph that AltaVista was rolling out right before they died.
I could search something like "trek", see that there were several very different clusters (one big one for science fiction; one big one for outdoors; and a smattering of smaller ones), and select the cluster I wanted to drill down into.
That should be even easier to do nowadays. Sadly, that seems to get in the way of irrelevant ad clicks, so nobody seems to have picked it up.
We used to rely on gossip and hearsay to determine what was true. Now we rely on social media and search engines. Which are basically just gossip and hearsay amplified a million-fold.
One of the strengths that's getting overlooked here is that crowd-sourced news seems to be more reliable than news from any one source. Everyone is biased, but if enough people are talking about something the bias tends to cancel out.
I do like the fact that people are finally starting to take a closer look at the hold algorithms hold over our lives. Maybe tech companies are going to start addressing some of their flaws.
> One of the strengths that's getting overlooked here is that crowd-sourced news seems to be more reliable than news from any one source. Everyone is biased, but if enough people are talking about something the bias tends to cancel out.
The problem with this is that people only have time to check a limited number of sources. Furthermore, many orders of magnitude more sources means that there are many orders of sources that are dangerously wrong as well. Thus, if someone sees a number articles/tweets/posts/whatever from a variety of people espousing that the Texas church shooter was an antifa commie starting the revolution, they are much more likely to believe it.
On top this, with so many sources, people need methods of filtering, and, of course, one of our primary methods of filtering is to find communities/people we tend to agree with. Thus, the now cliche echo chambers are born. These communities are furthermore highly susceptible to manipulation (see the russian facebook ads).
Furthermore, this makes the system fairly easy to take advantage of. If a number of people that seem unrelated online engage in a coordinated effort to spread a particular message, they can do it with ease, since when you see the same message from multiple, seemingly unrelated sources, you're much more likely to believe it.
So while you're right that we used to rely on gossip and hearsay, simply amplifying the number of sources in no way implies that we're getting a less biased message.
>So while you're right that we used to rely on gossip and hearsay, simply amplifying the number of sources in no way implies that we're getting a less biased message.
That's sort of what I was getting at. At heart, these new technologies aren't necessarily more reliable than what we had before. But on the other hand, I think that at the very least they're not any less reliable either. Simply by having more people reaching a consensus on something, you're probably going to have some great diversity of opinion.
In his book The Wisdom of Crowds, James Surowiecki posits that there are four factors that determine whether or not a crowd will be able to make intelligent decisions: diversity of opinion, independence, decentralization, and mechanisms for aggregation of collective judgement. Not to be self-congratulatory, but I think that's part of the reason communities like HN produce more intelligent and well-reasoned content than, say, an NRA fan club on Facebook.
> One of the strengths that's getting overlooked here is that crowd-sourced news seems to be more reliable than news from any one source. Everyone is biased, but if enough people are talking about something the bias tends to cancel out.
I don't think that is true at all in general. If the opinions are divided like 50/50, perhaps yes, but what if it's 75/25 or 1-versus-the-rest?
Also, most people find themselves effectively in massive echo-chambers, and the amplification factor of resonant opinions may not be very predictable.
Not sure who you are referring to when you say "we" relied on gossip and hearsay to determine what's true.
I clocked my share of years of life before google was even born and I have no recollection of the times you are referring to, very rarely gossip/hearsay were hold as truth. Maybe you are refferring to times before reading and writing was a thing ?
Nowadays I do not rely on social medie as I do not even have "social media" (care to explain what this buzzword means? I still have found a sense to this word). I do use the web and search engine to search for information but I do not hold what I found as truth and I can usually recognize bunk when I see some.
There are some people with an ability to think and some even have access to critical thinking.
> crowd-sourced news seems to be more reliable than news from any one source
this may be true in the sense that if you have a lot of news reports it is more likely that one of them might be accurate but in terms of the average accuracy of a single report I am sure it is much lower.
the current internet information free-for-all seems to me to mostly be pushing people toward a position of not believing in or trusting in anything.
the effect of this is not some kind of enlightened skepticism but that people will believe and accept anything, no matter how bat-shit insane, because they have no criterion for judging the accuracy of information.
> pushing people toward a position of not believing in or trusting in anything.
cause
> people will believe and accept anything, no matter how bat-shit insane
?
I'd say that both of those are possible extremes of not being able to tell which sources are trustworthy, i.e. some will distrust all sources equally, and some will trust all sources equally.
Which extreme someone tends to likely depends on their subject-matter experience. E.g. when I was a child, my parents bought me a time-travel comic (the German series Mosaik) which always had some background information on the epoch. I always enjoyed reading that and trusted the accuracy of their description. When they had an issue on April Fool's, it took my parents a while to convince me that giant tunnel-boring ladybirds on the moon were not real. Nowadays I'm obviously better able to distinguish truth from falsehood, but if I'm just learning about something for the first time, I might easily be taken for a fool again.
I wasn't aware that there was a perception that Google's results had any indication of the truth of the content it returned. I thought everyone just knew that anything on the web is just as likely to be false, spam, shills, or just bad marketing as it is to be an accurate worthwhile piece of content.
So while Google may have its flaws... I wouldn't go to them to solve this problem. I'd teach the people more critical thinking and encourage them to apply it to whatever they read.
Maybe there wasn't one. But Google has moved from a "top 10 results" strategy to a "we'll give you the best answer for your question" search strategy. They did this for one because that's what kind of service they'd ideally like to be for their users, but also because that's the only way voice search can work - by giving you just one answer.
The problem with this is that while before the #1 result on their search engine didn't actually have to be the best most relevant result that there is on the internet, but just one of the best answers, now it has to be the absolute best and most accurate result, because of this shift in strategy of only showing you one answer. They can't afford any mistake with this. And that's way harder to solve as a problem than the problems they had before with the top 10 results.
In this case, it apparently wasn't an answer, though. It was a box saying "popular on Twitter". That's not much of an endorsement.
It's a cliché to say that people should think more critically. In this case it's just a matter of knowing what "popular on Twitter" means when it comes to breaking news. It's not really about whether you trust Google and more about knowing that Twitter rumors are often inaccurate, however you run across them.
Here's a graph depicting Google's quarterly revenue, in billions, versus total number of internet users, also in billions. https://imgur.com/a/eBB9E
The gap that opens up between the two roughly shows how much more advertising they are pumping at us. Some of it understandable, like their acquisition of YouTube. A lot of it, though, is just turning search results into nothing but ads above the fold.
Investors are expecting to see that gap grow at the same, or higher pace. From my perspective, that means anything that doesn't advance that goal is a low priority. Whether that's accurate news, quality of the organic search results, etc.
> Investors are expecting to see that gap grow at the same, or higher pace.
It's sad to see Google obey to these Wall Street expectations. I remember reading about the stock structure Larry and Sergey set up before their IPO. They wanted to have control about Google and do good (Remember Don't be evil?). Nowadays they seem to just cater to the shareholders. Products get less great by the day and it really looks like they stopped listening to the users.
I remember quite correctly the "don't be evil" motto and I remember it having a very different meaning than "do good". I also remember that it was among the most misunderstood things at google and that it served google well.
I do not remember a time when google listened to the users but I do remember when google went from being a search engine to an advertising network. This is the point in time where the downhill slope started.
Part of the gap can be explained by: (1) Increasing income of internet users, (2) The average user spending more time on the internet, (3) Google getting more effective at matching advertisers with Google users
If anything, the average income of internet users is decreasing. The majority of people in the rich countries have been online since the 90s and early 2000s. Everyone who's been joining after that is most likely from an underdeveloped or developing country.
Your comment reminds me of a story [1] about how a Google engineer once significantly reduced the size of the YouTube video page but average load times got worse. Turns out they were getting lots of new users from third world countries with much worse connection speeds.
I also wonder to what degree this effect counters the argument that average incomes in Australia and US have not risen for decades [2]. Perhaps native-born people see their incomes rise, but the average is dragged down by immigrants who typically have lower incomes.
I believe that, but nobody other Google really knows how much of the gap is things that are either good/neutral for consumers, and how much of it isn't. Certainly some part of it is pushing the organics down, long forced ads in front of videos, etc.
However, investors are expecting that continued ~20% YoY growth...which will further widen the gap. I don't see how that's possible without doing more things that are bad for consumers.
I'm actually a Google shareholder and yes I do expect continued 20% growth, at least for the next decade. But increasing the ad/content ratio is only a small component of that growth.
Here are my estimates for the major components of Google's revenue growth over the next 10 years:
1. Number of Google users * avg income per user grows 3% per year
2. Time spent on internet by average user grows 5% per year
Do some compound interest calculation on some of those and you'll see the problem.
Assume they've been doing it for years and years already.
There is a hard cap on some of those. Especially ad/content ratio and CTR. At some point, you're done. For many queries, ad/content ratio is already maxed.
Like, how do you optimize this further? It's a 2 page ad already, with links out to other ads that pay google. https://imgur.com/a/hAFas
Also, there is a point where that sours. Some consumers will bail once the ad overhead is too much. Either via ad blockers, or the next "Google".
I agree Google's revenue will not grow 20% annually forever, but I don't see any problem with my 10-year projection. When I watch YouTube I see about 30 seconds of ads for every 10 minutes of video content. I could easily see this rising to 60 seconds over the next decade.
And the ads I do see are not very well tailored to me: I see cat food ads even though I don't have a cat. I see gambling ads even though I've never gambled in my life. I see ads for a sit-stand desk even though I already bought it. I never see MacBook Pro ads even though I put one in my shopping cart on Apple's website 6 months ago but did not purchase it. (Google has all my photos, all my emails and my location history for past 4 years, so Google should know all this).
Google has a long way to go to maximize the effectiveness of their ads.
I think your (1) is mistaken, it should be something like google expanding its domination over the web by multiplying the places to show adwords and adsense (gmail, youtube, analytics, maps, android, play store, earth, docs, drive, etc.)
Having the results you are most interested in highlighted and the bad ones removed seriously upgrades the search experience. Would be nice if Google had this functionality built in. I know they had the remove results thing builtin back in the day, but I find highlighting trusted sites even more important.
The way this journalist is using Google is so different from the way I use Google that I find it hard to understand what the problem is.
Trying to think what I googled recently. Let’s see...
- what new pages have appeared about ReasonML over the past week
- ditto for rxjs
- any code samples of doing server-side rendering when using redux-observable?
- any examples of applying functional javascript patterns to drawing on canvas?
- what javascript conferences are planned for the coming year?
- how is the phrase "the good wife" used in the books (searchable via Google Books)? is the title of the eponymous series an obvious allusion to some literary work? (haven’t fugured that one out, by the way)
For all these searches, I was very happy with what the relevance of what Google found.
You use Google for very specific niche queries where you already know enough about the topic to judge the veracity and usefulness of the results. If someone makes a site that associates ReasonML with ISIS and Google surfaces that as the #2 result, you'll just laugh it off.
The article is talking about people using Google for news and contested political topics. The extent of damage resulting from poor/misleading/fake results for the search "Texas church shooter" is much greater than for "ReasonML".
> You use Google for very specific niche queries where you already know enough about the topic to judge the veracity and usefulness of the results. <...> The article is talking about people using Google for news and contested political topics.
Yes indeed. Could it be that search engines are in principle ill suited for searching for news and contested political topics, especially if one has no way of judging whether a particular result is correct or not? Or, more generally, that searching without knowing how then to verify the quality of the results is a pretty silly undertaking?
In other words, shouldn't all responsibility be on the user of a search engine, and not on the search engine itself?
I switched to DDG on mobile at the recommendation of others on HN. I've been on it for a few months and it's great. 90% of the time it gets me what I need. When it doesn't I just add a !g and get the google results. Most of the time when I'm adding the !g is because I need a specific google feature like knowledge box or maps.
I haven't enabled it on my laptop because Google and I have been finding answers to my programming problems for 10 years. We have a nice relationship.
Rehashes a few talking points from media about Google having bad results in searches, throws in the election stealing with fake news conspiracy theory and presents the whole steaming pile with hot pink highlights as something worth reading. Fails to explain why I should not trust Google search (other than some bad results may pop up... a fact I'd previously believed to be well known about the Internet in general).
Why is the verge so negative on Google? Stop trusting them for search? Maybe if you are searching for simple things, but for more advanced searches Google is, by far, the best.
Google still almost always gives me what I want but when they "improve" my results by excluding words for me it ALWAYS fails. This morning I went looking for an Ansible role and the first few results it "helpfully" excluded the word ANSIBLE for me. I see this behavior quite often and I really do not understand it.
The way I see it is that Google is getting closer and closer to showing you only what they want... They tell you that they are guessing what you meant so that if it is completely unrelated, you don't get too suspicious.
I switched to Startpage last quarter as my default search engine. I did this mostly to make myself consciously aware of every time that I actually use Google.
My experience thus far. Google beats out Startpage when it comes to queries related to programing issues. If I have an error, or some issue Google seems to do a decent job of sending me to the right documentation or relevant Stackoverflow question.
Google also seems to do a decent job if I'm searching for particular commercial services if those services have physical location that it can point me to (Restaurants, brick-and-mortor stores, etc). All search engines seem to fail horribly if I'm looking for contractors/service providers in my area. Try searching for who actually provides internet coverage for your street address and you'll end up with every provider under the sun listed in the results. Try searching for a plumber, electrician, chimney sweep, roofer, etc. and you'll get SEO crap for people not even in your state. I've found going directly to government registries (broadband.gov, Arizona Registrar of Contractors [https://roc.az.gov/]) to be the only way to get around this.
What I've really found disappointing is how search results for humanities related topics have really gone down hill in the last decade. For example, I just searched for "The Rebel Camus." The results are always Wikipedia, Amazon, and then after that an assortment of either encyclopedias, links to Youtube, Goodreads (or similar sites). Used to be, that I would get taken to the personal website of a scholar, hosted at a University, whose life studies revolve around that topic -- links to scholarly papers, coursework, suggested readings on the topic.
> Google beats out Startpage when it comes to queries related to programing issues.
I have found that Duck Duck Go has excellent results for programming issues. You should try it!
I can echo what you said about services with a physical location. Typically when I'm looking for something nearby, I will search in DDG with the g! modifier to take me straight to Google. The place that I want almost always appears on the right, with map, phone number, hours of operation, etc.
With the growth of 'category kings' (Amazon, Stackoverflow, Wikipedia etc) does anybody else feel that google's search becomes less and less useful?
I never understood the 90's obsession with domain names until I skimmed several news articles from that era. The main logic was:
1. People will go to a couple of websites for most their needs, the ones with memorable names will be easier to go to.
2. Things like search engines (Google) weren't as hot as curation directories (yahoo) because nobody expected so much fragmentation on the web.
3. The more fragmented the market in an industry, higher the value of search ads. If you know the most options and high quality reviews are available on Amazon, you don't really bother with as many reviews.
While this analysis failed for the past few years, with the amount of consolidation happening in the industry recently its possible search engines loose their main source of income (e-commerce) and start being used more from ideleness/curiosity perspective.
Wonder if google should pre-emptively start penalizing the aggregation marketplaces and promote direct providers (essentially try to be the Ebay for everything) to prevent such an outcome.
At the risk of being presumptious, I would classify searches into broadly two categories:
1) Finding good providers/discussion places regarding something you need, say news, goods, experiences etc
2) Finding more info about a specific piece of content/goods . Eg: Product reviews, solutions to some technical doubts etc
For most people I know, 1) is pretty much sorted. They have a list of places they check for clothes, groceries, jobs etc. News is read from specific websites/apps. Reddit/HN/Fb/Twitter provide most the social interaction and discovery aspects.
2) is where I see Google providing much more value. But the way things are proceeding, most of the top links relevant to a google search are generally from aggregators themselves. Amazon for most ecommerce, Tripadvisor for travel, Stackoverflow for tech, Practo (in India) for doctors etc.
Let's assume Stackoverflow introduces points for using (and help 'improving') their own search. Most people will open stackoverflow first before searching on google. Amazon has already achieved this level for most things related to e-comm (specially less expensive products).
There will still be a market for Google ofcourse, but it will be a much smaller chunk than currently being predicted.
I actually stumbled across this line of reasoning when somebody asked me if its a good idea for individual service providers (docs/merchants etc) to open their own websites. For 99% of them it isn't.
Maybe we shouldn't be thinking of Google as a source of truth, and instead educate people on the differences between popular/relevant and "true". I don't trust Google or progressives or the alt-right to arbitrate truth; epistemological fractures are already the cause of much of our political dysfunction, no need to bring that into search as well.
It's interesting that newspapers used (before they all went effectively bust or got bought by monsters) to make a big deal about editorial independence and journalistic integrity. This caused people to trust them (in time). How come Google doesn't appoint a board of editors and grant them independence ?
Google is not an oracle. Trusting search results is the problem especially, when the algorithm reutrns popular/relevant results, is the problem. Using Facebook as your primary news source is the problem.
It's become fashionable for media outlets lately to beat up on tech companies and blame them for the general public's lack of critical reasoning and analytical skills. Many decry that democracy is under threat. However, it is often said that an educated and well-informed citizenry is necessary for a viable democracy.
A lot of people discuss about Google's faults and improvements. If there's anything I want from Google is cheap API access to search for deep searches (much more than the first 1000 results). There's no systematic way to get deep into results, and running even a few thousand queries breaks the bank. It's useless for automated search agents.
I second DuckDuckGo, especially because of the !bang modifiers. Makes it easy to also get to Google — just incase the query is too vauge, which DuckDuckGo still needs work on.
Use Google search all day and just find it amazing at getting what I am looking for. Never seen Google filter in a way that helps them. I will get negative Google articles and they would be the first Google filtered it there was an issue.
tl;dr: there needs to be a tl;dr on every web page so People don’t have to read content they don’t care about to get an answer to something they googled.
Yeah? Maybe. I don’t know. I don’t google much.
I go straight to BBC or the Guardian for news.
I get my tech news from email lists or HN.
When I’m curious about random stuff, I’ll end up on Wikipedia.
What i use google for is getting to a deep link in a site i know i want to go to:
* shortcuts to amazon
* shortcuts to stackoverflow
* probably specific porn term
When I use google I’m usually looking for a specific answer to something that I’m not sure where to find it.
The two most recent example are:
- safety rating and electric range of the Volvo T8
- a recipe for chicken tortilla soup
And here is where i get to “maybe”.
I want an answer to those questions. What do i get? 40 paragraphs of SEO unique content around some experience someone had as a child with their grandma and then a recipe for chicken tortilla soup buried in there somewhere.
I’ve started and worked for a few big content sites and i know the SEO game well.
Websites that want to show up in google can’t just focus on what their core competency is, they also have to be content site and have writers.
It kinda sucks because I just hit the back button when I’m expecting a succinct answer and I’m confronted with a novel (and i think that’s still a big penalty to them) and try to find the quickest answer without reading a life story or opinions on a car.
Finding a useful site on page 2 and beyond? Good luck!
I've switched to DDG half a year ago or so and haven't looked back much. I only once in a while use the !g bang and almost always come back disappointed.
Sure, there are also some spam results on DDG (and probably there are going to be many more once it gains more relevance to be spammed) but for now and my searches i exclusively use DDG on my private computers as well as at work...
So, to summarize, i think the relevancy of Google as a search engine will, imho, continue to drop and also the perceived "truthiness" of statements found thereon as well.
To add, i also don't necessarily think of this as a "Google" problem - the same people that will believe in a statement gathered from a quick search will also believe in the same thing when they see it in a Facebook post, Whatsapp chainmail or whatever....