It is not. There was no human on Mars (yet) and that's the point of the question, proving that google's ML approach will recorrect and reinterpret the question and give a wrong answer.
Although it's interesting to note that two presumably human commenters in this thread got that wrong. Maybe the Google results are more human than we would have guessed.
it's more showing that the search engine isn't much better than bag-of-words model, and doesn't seem to "know" enough logic/reasoning to parse a sentence and determine that it's false because it says mars instead of the moon.
BTW, a big reason for this is the search quality folks at Google left the building and got replaced with growth marketers.
This narrative doesn't really ring true to me. In any case, if you're saying that this result indicates that the system is no better than a bag of words model, doesn't that indicate that humans are equally no better, considering the errors made by the two commenters in this thread?
To me it just seems like a query that is likely to trip any imperfect entity up, since a "when" question usually implies that the event in question is known to have happened.
Google search doesn’t know the correct answer, but it does know what people usually search for and returns an answer to that. It’s silently adjusting Mars to Moon before returning an answer.
"Broken" is a stretch. It gives the only result that could be reasonably expected by anyone making that query in good faith. If you're genuinely expecting a result like "never," then I wonder, would you also expect a result of "never" for the query "When did Neal Armstrong set foot on the Moon?" (notice the misspelling of his name)
For instance I saw that many Russian generals had been killed during the war. I wondered how many Ukrainian generals had been killed over the same period.
Still can't figure it out. Lots and lots of articles about dead Russian generals.
There's got to be some way where the search engine doesn't second guess you. You can never find the answer to something adjacent to a popular question with the current state of things.
From google's point of view, it's not second guessing you. It just associates words it decides are similar. You'd probably accept results for "dead generals", "killed generals", or "killed commanders". To google, dead = killed, generals = commanders, Russian = Ukranian, and moon = mars. It will expand the meanings of words until it has a page full of results and ads to show you.
Ukraine has repeatedly asserted that the HQs of all their major units are intact and as far as I know there have been no announcements of any Ukrainian generals killed and as far as I see even Russian media have not contested that. With some searching on the Russian internet, I find some assertions in Russian social media from early March that Lt Gen Юрій Содоль commanding the Ukraine Marines was killed, but not anything conclusive. So the number seems to be 0, or perhaps 1. And IMHO that isn't surprising, it's quite normal to fight major wars without many generals KIA, it is very surprising that Russia lost so many of them.
Even on colonel level the losses seem to be small - I think that this https://www.republicworld.com/world-news/russia-ukraine-cris... this was the first Ukrainian ground force colonel KIA that I recall, there was at least one air force colonel before that.
Ukraine has better signals security so we don’t know.
The other thing is that Russia is operating WW2 style. It’s a top down system where the lower level people have no autonomy. American colonels and generals die in helicopter shoot downs and accidents. Russian generals get assassinated on the front screaming at soldiers to move trucks, etc.
But "Mars" is not simply a misspelling of the correct object, it confuses two distinct objects. A better comparison would be "When did Louis Armstrong set foot on the Moon?"
Indeed, it's a confusion between two distinct objects, and the intended object is overwhelmingly clear, which is why it's appropriate to correct the confusion and give the intended response.
Amusingly, "what date did Louis Armstrong land on the Moon" displays the same behavior of showing the date "July 20, 1969" and no indication that it answered a different question than the one you asked.
I'd consider it broken. In many other cases, if I make a typo in my search, Google will (rather prominently, IMO) at the top of the search display something like:
> Showing results for "When did Neil Armstrong set foot on the Moon?"
> Show instead for "When did Neil Armwrong set foot on the Moon?"
However, it seems to only work for actual spelling errors (e.g. if I typed Armstrong incorrectly) vs. context errors.
So to answer your question, I'd expect Google calling out its interpretation, but not necessarily a "never".
What is “good faith” in that case though ? Some people will really not know (think kids for instance, but not only)
The interesting part to me is that Google will show redirected queries from typos (“Did you mean “When did *Neil* Armstrong set foot on Mars” ?”) but not disclose the term it completely ignored.
I makes it look like it parsed and validated all the search terms before coming up with the prominently displayed date, which makes it worse.
On your point, it’s broken because it shouldn’t display “Never”, and instead skip the date widget and show that the page results are for the moon, like it does for typos and other kind of corrected search terms.
Or alternatively _always_ show which part of the query the results are based on. Failure to be transparent creates the problem.
> Some people will really not know (think kids for instance, but not only)
It seems very improbable that a kid would ask that question without having any preconceived notion whatsoever of a person with that name famously stepping foot on an extraterrestrial body. It's overwhelmingly likely that they have the famous 1969 event in mind, and that's why Google's response is appropriate.
Of course I'm not disputing that it wouldn't be even better for the Google response to explicitly correct the mistake in the query. I'm only disputing the fact that it's "broken." If you were asking this question of a human as some sort of test (but the human didn't know they were being tested), it would clearly be a "gotcha," and you could likely "fool" highly educated people who know full well that it was the Moon and not Mars.
The answer Google gives to the question is wrong. End of story. Google aims to be a question-answering machine with the Info box, so giving the wrong answer is a broken state. Not sure you can really reason you way into how it's kinda-understandable-and-gotcha! to be 'not broken'.
Aside: Going to get real confusing when another Neil Armstrong born in the 2010s does set foot on Mars in the 2030s. I wonder if we will still get the first response from Google, 1969, and if you will still consider it to be 'not broken' then too.
It’s one of the core assumption that is correct for us to have in our day to day interactions with other real people, and we also are able to adjust the probability levels looking at the person or the context, and follow up depending on the reaction to our answer.
None of that applies to Google Search [0], and we are fed with “most probable” results without qualifiers, little to no sensible adjustments, and very little control or opt out options.
Billions of people use Search every day, at these scales what is “improbable” actually happens millions of times, and I feel too many people are willing to throw the odd ones out under the bus, even if the current situation isn’t perfect either for the 90 part of the 10/90 split.
[0] search personalization based on logged in profile could be used, but in practice they only apply that to very crude adjustments like country, language and frequent searches.
There is something to be said about how failure cases are handled even for systems that are almost always (say, 99.9%) successful. This is particularly true for systems where failure cases cause harm to people. A common example might be assumptions about people's names that are implemented in some form on a government website. Perhaps it is the case that 99.9% of people in this jurisdiction do have a first name and a last name each of which are representable in Unicode with a character count less than 50. But that's still 0.1% of people who now can't get a driver's license, and that's unacceptable.
However, in this particular case, I think the very low probability of someone genuinely asking when Neil Armstrong set foot on Mars, combined with the very low probability of any measurable harm being done to those people by Google's response, makes me conclude that this is reasonable expected behavior and not something I would call "broken."
On the Google side, I think it’s an issue that has more consequences than just the moon landing.
For instance, for me “When is the francis election” (where I would have men “french or france instead of francis”) gives me a big and bold “Mar 13” with smaller below “Anniversary of the election of Pope Francis Observances” and an long anniversary Year/Week/Date table taking 80% of the widget display.
And as with the other examples, there is nothing showing the word approximations that has happened regarding to the original query.
There must thousands of other instances where a search result will come up with a big widget, an answer in big and bold font, except it will be completely wrong and have a direct impact on the user missing a deadline or taking the wrong action.
Of course users are supposed to know better and check the full result, but as you point out, if it’s almost always what they expected, they’ll learn to rely on it and be more complacent.
How is this different from answering the query "2+22=?" with "4"? There's a factual question that was asked and the answer provided is objectively wrong. Whether the user meant to ask a different question is irrelevant. Either say "2+2=4" or say "24", but under no circumstances say "4" without noting that you've answered a different question than what was asked
It's different because nobody is actually going to ask this silly question in good faith. It does not matter in the slightest bit to anyone whether Google can produce a correct answer to nonsense questions that nobody is ever going to actually ask with the intention of getting a correct answer. Google also can't produce correct answers to queries written in ancient undeciphered languages. Oh well.
It also seems rather unlikely that someone would search for "2+22", but not obviously any more or less unlikely than "2+2", and building a calculator into your search engine is trivially easy compared to handling natural language queries, so it's not a useful comparison.
> Whether the user meant to ask a different question is irrelevant.
Of course it's relevant. If Google can correctly determine what question the user meant to ask, then obviously Google should provide an answer that provides value to the user instead of trolling the user with nitpicking about spelling or what have you. If you ask Google for the "capitol of India" you'll get New Delhi, even though that is the "capital" of India, and arguably the "capitol" of India is actually the Chandigarh Capitol Complex, so the result is "objectively wrong".
Google can interpret the intended meaning of arithmetic expressions like that with even higher confidence than this Neil Armstrong query. But the confidence with which Google can interpret the intended meaning of this Neil Armstrong query is well above any reasonable threshold one might propose.
I would argue that it's roughly just as clear that the query intends to refer to the Moon as it is that the query intends to refer to the Canadian Neil Armstrong who was killed in a plane crash in the Antarctic in 1994.
At the very least I would expect a "Did you mean..." message at the top. In fact, it correctly does this for the "Neal" query. The lack of one indicates a failure on google's part to understand the answer it gives.
Yes, people probably don't make this exact query in good faith very often, but it raises significant concerns for other queries where google tries to supply an answer.
I would say that I have the exact opposite experience.
My company is moving from IE to Edge as the default browser and I search for code issues on Edge (Bing). I usually do not like the results and then have to manually type in Google.com > search for code issues as that gives me better results.
I like Edge, but the setting for the search engine is quite difficult to find despite being (I assume) one of the most common settings to change. You have to click on "Privacy, search, and services", then go the "Services" heading and click on the tiny "Address bar and search" box, and then you finally come to the page where it will let you change the default browser. Said "Address bar and search" box is also the very last item on the "Privacy, search, and services" page with no highlighting or emphasis whatsoever.
Meanwhile for Firefox, I just have to go to "Search" and the setting for default search engine is immediately obvious. Chrome seems to have a similar layout as well despite being produced by a search engine company. Edge is my daily driver, but it still took me way longer to find the default search engine setting in Edge compared to Chrome and Firefox.
Also, I have to compliment Firefox for making it really easy to search with a non-default search engine. I generally use Google, but I use Bing when searching for internal work stuff since it is integrated with O365 and SharePoint.
> setting for the search engine is quite difficult to find
Didn't seem all that difficult to me.
I went to settings, put "search" in the search box. That highlighted in yellow "address bar and search" click "manage search engines" click, 3 dots next to google, click, make default, click. 4 clicks with guiding highlight throughout. Didn't even need to google how to google with edge.
I'm still using FF. However, switching search engines in Edge seems about the same difficulty as doing the same in FF or chrome.
I kinda hate searching for recipes. It's always the same 3-5 sites who optimized for seo and , no idea what words to use here, for non European influenced cooking the recipes that surface at the top are often by people that really have no clue what authentic is.
not a search example but it's like Jamie Oliver's fried rice
we added recipe search to Breeze last month -- search for recipe name, ingredients, etc., click recipe tab in results -- would be curious if same / better / worse, etc.
An interesting observation is that over the years, I've sort of found that if there are not many good results, maybe I'm not asking the right questions, or I'm not thinking about the problem the right way. For example, you come up with some weird programming idea to solve a problem and no results are found -- I'm almost always headed down the wrong path. This has been proved to me over and over again as sort of a canary in the search coal mine.
Google still solves all my questions, if I'm asking the right questions, I guess is what I'm getting at.
>"if there are not many good results, maybe I'm not asking the right questions"
I used to think this way too, but I've come to realize that Google has turned from a search engine that returns results based on my input to an answer engine that actively tries to reframe whatever I enter into some other more generic query - usually with the intent of selling me something or returning SEO spam. I've also found that the old google-fu techniques are now so unreliable that they must have been deprecated. I can't tell you how often I use quotes in my query and see results that don't contain that text at all.
Nowadays you should google using what in 2005 would be considered "the wrong way to search". Back then it was all about keywords, quotes and so on. For example if you wanted to know if dogs could eat apples you'd search:
> dog apple dangerous
Today you should search:
> can dogs eat apples
And you get way better results with the second form. I've noticed that people who are stuck thinking that the right way to google is still the former overlap a lot with the crowd that keeps complaining google is worse now than it was before.
"you're thinking about it wrong" isn't really that great a defense of Google, you know? I'm not being flippant. The two examples you gave are different searches, with different goals.
I'm in the awkward middle area I think where I'm almost a digital native, and what you said maps perfectly to my experience. In middle school and high school I got pretty good at the first type of search you mentioned. I was young enough that I was almost learning how to "speak" google and it came easily, I was fluent. Throughout the end of high school and into college I had more and more issues finding what I wanted and one day after I had failed to search several times I thought to myself "Fuck it. I'm just going to type in exactly what I want in plain English, as if I was a moron, and see what happens." And it worked, and that day a lightbulb went off in my head.
I’d argue that it’s not a “better way”, but rather a regression: I assume Google genuinely wants to understand my query, and I don’t see any reason why it shouldn’t still be obvious to Google what kind of results it should return.
Otherwise, it there truly is a “right way of asking Google questions”, why doesn’t Google release and promote a guide about it so people can be more successful in their search?
Okay, so answer me this... I'm trying to find out about the extra controls in the old space suits. I believe they used a chin switch or chin toggle (resistive). If I google "did astronauts control things with their chin" I see: 'It's All Different for Women in Space | Marie Claire' and 'Pillsbury Space Food Sticks were sweet Slim Jims in space' as the first 2 answers... nothing about switches. If I google 'space suit chin switch' I get links to resistivity toggle switches.
heck I was putting in exact model numbers today like MS3015S20-27SZ and google was returning me OTHER models such as MS3015S20-27PZ which is the male or female version of a part (PZ vs SZ), or changing the20-27 to a 29-4 which are all completely different parts. This is with verbatim on.
Also worth comparing (seems to focus more on diabetes than cyanoglycosides. Diabetes is the bigger chronic problem, and cyanoglycosides are the bigger acute problem, so which is a bigger danger largely depends on how disciplined the owner is.):
Right - I think this is part of the "scaling" of the internet. Those of us (relatively small set of earlier heavy internet users) are forced to change as the product is built for the wider audience.
Back when I started at Google, they at least said they actively targeted "long-tail" queries to gain market share. The thinking was that people would use a search engine until they hit some query that their usual search engine failed on. Then they'd give another search engine (perhaps Google) a try, so if you really targeted those tough queries and were at least decent at the common queries, you'd gain market share.
Though, even when I was doing indexing changes at Google, the common practice was to do A/B testing with both the most common queries and a uniformly random sample (see reservoir sampling) of queries in order to justify a go-live of indexing changes. The former explicitly over-weights common queries, and the latter still optimizes for the common case. (In case you're wondering, the worst query I had to manually check in A/B testing was [flesh hook suspension].)
Google used to turn off some of the query re-writing logic (that tries to fix your query) if you used a query operator. (It has been a while, but I think maybe even their "Kansas" user info database kept track of the last time you used an operator, and would turn off some of the cleverness if you had recently used a search operator, as it was a good signal that you were a power user capable of optimizing your own queries.) My understanding is that they don't disable any of the too-clever bits for power users any more, and that everything uses all of the cleverness of learn-to-rank all the time.
I suspect it has gotten even worse with learn-to-rank, as it must be incredibly difficult to intentionally under-weight the uncommon/difficult queries.
They did keep track of when users re-issued similar queries in a short period, as a signal that the ranking algorithm wasn't doing well. I think an optimal system would use learn-to-rank for the first query in a related sequence of queries, and then switch to turning some of the smarts off, and finally switching to a learn-to-rank algorithm trained only on later queries in these related query sequences. That way, they can avoid the secondary learn-to-rank instance from over-fitting the median/easy queries.
They taught us the search engine operators in elementary school computer class. It’s a shame that they’ve gone away now, and the internet is so much worse than it was then.
Be careful. Maybe it's a useful rule of thumb, but be careful of the "searching for your keys under the light pole" problem.
I worked at Google on rich content indexing for 4 years, more than a decade ago now. Google is pretty good. It used to really cater to "long tail" searches, but a combination of SEOs getting better (and specifically targeting Google) and "learn to rank" over-optimizing for the median query, means that lots of long-tail queries don't do very well on Google.
This is a great way to describe it. Google still gives me good results for the common stuff, but really fails me when I'm looking for something in the "long tail".
There is, use a different search engine. I've even found something as new as Andisearch good for finding research papers and Qwant good for general searching. There are many more and they keep up with Google pretty well now without the trash.
More than a decade ago, Google had some special-case logic for these "navigational queries". Lee Kai-Fu (head of Google China at the time) said that navigational queries were particularly prevalent in Chinese search traffic.
Though, there's also a use case for people trying to find third-party information about various websites, particularly in trying to figure out if the website itself is a scam. You really want to have the navigational result up at the top unless you have a very high degree of confidence that the site is a scam, but in all cases you want high quality reviews of the site to follow up the navigational link.
I wish I could find forums with my answer instead of a multitude of websites and blogs that do not answer my query. I guess I could add 'forum' or something but when I am already searching a specific query and particular piece of hardware I don't need the first page to show me retailers of said hardware.
Tried it, blog spam mentions finding or not finding stuff in forums way too much and it basically didn’t help unless I started getting into weird Google fu like including strings from the copyright footer of various forums in an advanced search using the OR operator and multiple quoted sections … at which point I was like “why am I working this hard on making Google better when they don’t fucking pay me”
I googled a phone number recently. It was the first result, got my answer. The rest of the results were websites for physical locations I had been recently, including a hospital, none of which had any relationship to the phone number.
Google is _not_ a good guideline for asking the right questions. It is trying to make as much money as possible and it will scramble anything it can think of to do that. Don't use that as a guideline for how you are thinking about questions.
oh Haha, funny old meme. Step 2 though is very clearly to show more ads and more content, to show more 'full results pages', etc, rather than just show me the 1 thing I needed and be on my way.
I should include that Google said next to each result ('"555-555-555" is not on this page') and then show the page. Totally knows that the results are unrelated to my query, but shows them anyway. Why?
So was there even a single ad on that search result page? From your original description it certainly didn't sound so. If not, how did it contribute to you being shown more ads?
> For example, you come up with some (...) idea (...) and no results are found -- I'm almost always headed down the wrong path.
And by that you mean, Google is telling you how to think, and that you should think in some way and not in some other way. I'm quite sure I am not comfortable with that.
This resonated a lot with me. Often if I'm learning a language and I see no one is doing what I'm querying for, I take a step back and try to ask a more basic question
The question is whether any of those other sites could maintain quality while also being Google, that is, while being target #1 for every SEO bad actor, spammer, and anti-trust regulator on the planet.
The people at W3schools are a combination of all of the above: filthy SEO spammers who would readily make an antitrust stink if Google started just ripping off their content, but who in all likelihood don't care if Neeva does that, if they've ever even heard of Neeva.
Google spies on us so they can better serve us. Maybe it’s time that they let us directly tell them what we want. Let me easily blacklist sites from the search results.
They could even take the blacklists from all people with a high reputation and use that as a signal when they are building search results for others.
If Google can't tell the difference between a click farm and real people then how can they hope to have advertisers trust that their ad performance is accurate?
it's not really that Google lags, but rather SEOers have optimized for Google. The problem is intractable. When people talk about the 'good ol days' or times when Google was better, it was simply because there was less SEO, less spam and generally fewer pages on the internet.
Google could be better than it is now, but there's no incentive to do so, unfortunately. Say Google allowed you to blacklist entire sites from the results - inevitably those sites that have the most ads would be the most likely to be blocked, resulting in lower revenue for Google.
The fact that recipes on on this list suggests this is the case.
Recipe sites are notorious for SEO tactics. They all follow the same highly optimized format with the stupid story about the author's grandma and how they just couldn't get enough of these cookies, and how the recipe was lost for 90 years until recently their great great uncle Lou found a copy of the recipe in an old donut.
Google has all of the tools to solve recipes. Make Google Recipe with a standard template and a way to link in and out of YouTube. People who contribute popular recipes get ad revenue. People with recipes and YT videos get even more. Adding ways to find similar recipes would be a killer feature. Who hasn't found a recipe that was almost what they were looking for, but was missing that je ne sais quoi.
The stories also serve the purpose of providing copyright. A recipe alone doesn't have copyright, but the story mixed throughout the recipe does have copyright.
Nah, the stories are there so you have to scroll 6000 pages to finally find the actual recipe. Why do they want you to scroll? Because every 5th word is an advertisement. More scrolling, more ads, more revenue.
All of these recipe sites (that I've seen) will drop the ingredients + directions on the bottom of the page.
I thought it was so the site has longer to mine monero using your CPU while you scroll?
Ever notice how incredibly slow any device gets when looking at recipes?
It's possible that the scrolling is giving the remote server time to use your CPU/GPU to mine some space cash, but it's also possible that attaching listeners to events that happen many times per second (such as scrolling, mouse movement) is common, especially if 3rd party ads are allowed to interact with the page, making it nearly impossible to know what those ads' scripts are doing.
No, it's so the pages follow thr Google blessed "ideal" page format. X number of words, 2-3 pictures, iser stays on page for longer than Y time.
Last one is the killer, a site that quickly gives you the info you are looking (or quickly shows you it doesn't have the info) for is punished by google search ranking.
"Recipe sites are notorious for SEO tactics. They all follow the same highly optimized format with the stupid story about the author's grandma and how they just couldn't get enough of these cookies, and how the recipe was lost for 90 years until recently their great great uncle Lou found a copy of the recipe in an old donut."
Isn't that similar to what they tried for things like shopping for example, and they got sued for? The problem is that if they become an aggregator for a specific type of content, like let's say travel or lyrics, then other aggregator websites start suing.
You don't even have to blacklist... just look at user behaviour. Did the user come back to google and select the next result after two seconds on that site? Did many users do that? Something has to be wrong with that site, rank it down.
I'm pretty sure i'm not the only one who clicks on pinterest results in eg. image search, and immediately click back to find another image somewhere else.
No, that SEO shit works. The whole adbased internet is optimized to waste your time.
Did I spend a minute or two trying to read a 1000 word article, because I was looking for how many pixels there is in a 4k monitor? Or did I spend 2 seconds visiting a chart?
> Did the user come back to google and select the next result after two seconds on that site?
It usually takes me at least a minute to recognize a search-optimized text if it's a topic I am unfamiliar with. Let's say I am googling something about windproofing underfloor insulation. The article starts with some basics about underfloor insulation in general, so I skim through the introduction, start hunting for the part of the article where it actually starts talking about windproofing and realize it's been cobbled together out of six random introductions or generated by GPT-3.
It has been a constant battle of SEO improving and Google improving the page ranking algorithm. For a long time Google was winning that battle, but more recently they haven't seen as interested. The economic argument you're making may be the reason, but it's short termist if people slowly get driven away from Google by poor quality search results.
Five to ten years ago no other engine was close - the fact that duckduckgo was even try was comical. Google had an effectively monopoly. That might not be true at all five to ten years from now if the trend continues.
Google also isn't great at surfacing genuine product reviews that aren't just people reiterating amazon reviews with referral links and SEO. Like even stuff reviewed by reputable websites somehow loses out regularly.
If I want someone's actual opinion and people I know don't have one, I have to check reddit or metafilter.
I used to add inurl:forum or inurl:thread to my searches to find actual opinions from people instead of ads disguised as articles. Now I just search reddit.
Google made an algorithm change a few years back that knocked ask.metafilter.com answers way down in search results. The resulting drop in ad revenue was disastrous for metafilter, which runs with a small staff on a tight budget. Kind of bitter to remember considering the dismal quality of search results today.
Question regarding the recipe example. Does Google deduct quality points the more ads that are put on the website? A long time ago, I recall hearing how Google was able to beat Yahoo by focusing on quality of ads & a higher click rate. Has Google defeated it's rivals & switched to their tactics?
Side note, I prefer DDG as my search but only because of the bang operators. For recipes !b added to the search lets me use Bing. As the article points out, Bing is really awesome for searching for recipes.
Looks like I need to start trying out Neeva & You.com. They had some nice features in this article.
Neeva seems to be stealing all the important content from the recipe website which is a highly discussed issue. Bing tries to walk this line by making you still go to the website to read the instructions. Obviously people have trashed Google for doing this same thing on other kinds of websites. Though blogger recipe websites have somewhat encouraged this behavior due to their insane amount of ads & life stories they're well known for.
A very high percentage of ads on sites now use the Google ad network. So it may actually be in their interest to promote sites with more ads. They certainly promote a lot of garbage new sites on their Google news feed app and the product that they include on the pixel phones.
When it comes to programming it is not really due to superior work by the competition, but by so many people gaming the Google PR flooding the front pages with content-farm level crap.
Generally for any search string producing 3-4 pages of spam, the competition would give a better results, without necessarily having any technological superiority. I cannot believe Google have not recognised this as a key risk.
Also noticed a recent change that Google did, if I search for e.g:- Volvo XC 60, it does not show me the Volvo XC 60 page on Volvo.com, it instead shows the ad that Volvo paid for and a bunch of other ads.
So they essentially don’t want you to click on the organic link that points to Volvo.com.
They want Volvo to know that all the traffic to them is being sent due to the Ad and not from any organic links.
This is what the not so evil company is doing, imagine if they are actually evil..
If I search for "Volvo XC 60" then there is absolutely no doubt I want to see information about that vehicle. Any information about Subaru is just irrelevant noise.
And google hasn't handled complex queries for quite a while. A long time ago, you could get sensible answers to boolean queries - not so much now. A useful search engine would understand meaning, rather than just an index lookup.
anyone else find it so hard to find basic apparel and shoes? All my search results return me known brands with high ad spend. Recently started to go to reddit male fashion advice to look for crowdsourced recommendations but that's fairly time consuming.
do you have some specific examples? if don't want to post here, can dm dotdotjames on twitter -- we're adding product search to breezethat.com sooooooooon
> I already know what an Exception is, so I don't want to scroll halfway down the page to find what I'm looking for.
I can’t relate to this. That example, of the first result being the canonical documentation on the subject, is a search engine working exactly the way I want.
I thought the same thing but in the past I've also found the official python documentation to be somewhat verbose. Maybe these days I'd appreciate it more.
I always found the MDN documentation striking a better balance of being in-depth but not an essay.
I know google search is probably not the same as google flights, but I spent a few hours recently (over a few days) using google flights and it was a delight to use compared to the Expedia children.
We had a bereavement in the family and had to book multiple independent tickets because we could not travel together. This was just after the Ukraine war started so prices had gone through the roof. Exact number of days was not that important as compared to the price and the duration and using google flights UI to slice and dice the data was such a joy. Want to freeze the airline and look at the alternatives - which include from and to dates, number of days of trips, or freeze any other parameter and analyze others, the response was sub-second. Did not eventually book through them since I did not want to get into the google payment system (they offer booking through others that I did not explore).
On the opposite side was Expedia children where they would show a price of 2800 and when you click the price invariably it has gone up to 3600. Again. And again. And again. Not sure if that problem existed with google although I paid the exact same price as the airline as shown by google, it could just be a coincidence.
Google bought the main provider of flight information in 2010 (ITA). ITA had such a monopoly that Google was required to continue licensing their software out for years afterwards, though I think it expired in ~2016.
I wonder how much of the old Lisp codebase is left.
Also, it looks like the Matrix might go extinct soon:
> This interface runs on a deprecated web platform. At some point in the near future, we will be forced to shut it down. We unfortunately do not have a timeline on when this will happen. We welcome feedback about features missing from the new interface, we read all feedback and open bugs accordingly.
That's on the LCC, not the aggregator. They don't want to pay commission, and that's fine. If you want to fly on them you have to actually go on their website.
Well, you're getting it free: if you don't get any use from it, don't use it. Try doing an iterated search of every possible route on every possible airline website and let us know how you get on!
At least on my searches with Google Flights, Wizzair shows up as a carrier for EU flights.
Good search engine for low cost EU carriers is https://www.azair.eu/
I personally like to use Momondo. Have had pretty good experience with Momondo over the years so far. I haven’t tried Expedia, so I can’t directly compare it. But I recommend giving Momondo a try anyways.
I will also plug Kayak as being the only other decent flight search site.
The difference between Kayak/Google and the others is that Expedia and friends are online travel agencies. Kayak and Google are really just search engines. It makes a world of difference.
On top of querying the travel agency APIs, Kayak will also query each airline directly. So you also get fares and airlines that aren’t published to agencies
Google flight's true genius is being able to search multiple city-pairs.
Let's say you want to get from the West Coast to Europe on a business class flight, but want to save some money.
Realistically, it's cheap and easy to get from West Coast airport to another, and similarly cheap and easy to get form one European airport to another.
Google flights will let you search for the best combination of flights that depart from any combination of up to 6 airports, and arrive at up to 6 airports. The technology they purchased (ITA) will let you do this as well, but limits you to a single country. Google flights? No problem with destination airports in multiple countries.
So, you could search for a flight that originates in some combination of, say Seattle, Portland, San Francisco, Los Angeles and Vancouver BC, and lands in some combination of London, Amsterdam, Paris, Madrid, Milan or Frankfurt. (Or any other 6 large European airports).
Google Flights will also show fares for a specific trip length (say, 14-days) over an entire month.
Want to filter by a maximum flight duration? No problem. Number of connections? Done. Specific airline or alliance? Easy.
In a traditional search engine, I'd have to run close to 400 individual searches to get the data that Google Flights gives me in a single screen.
It works well for domestic flights as well. I recently helped someone who had to attend a wedding across the country find a flight that was less than $180/person, when they thought they were going to have to pay over $800/person. Just by using Google Flight's tools for about 10 minutes.
There's plenty to complain about in the Google ecosystem, but Google Flights is amazing.
Same! Their “agony” sorting (basically sorted flights by a combination of # of stops, total hours, and price) was a breakthrough at the time. It seems to be factored into most flight searches today, even though it’s not labeled as such.
The one thing I wish that Google offered would be the ability to blacklist sites for a period (coud be fixed - say 6 months).
So damn annoying when the top search results all lead to shitty SEO-optimised sites that use a whole page to blather on and on, leading to a tiny information nugget at the end. No value, just excellent SEO scamming.
As these scam artists get better and better at this, Google gets less and less useful.
When I see a site like that, I can be quite sure there is no value to me from that site. I want to blacklist it - not forever (though I'd settle for that) but so I don't see it in search results again.
The crazy thing is that this could even be a benefit for Google themselves. They could aggregate these signals and use them to identify SEO scammers, since their algorithms clearly can't. I'm sure that Google aren't happy with the lacklustre performance of their search in modern times.
This used to be a feature of Google Search. It then moved from the Google Account to Chrome. I'm not sure if it still exists in the Chrome browser. It can be done with addons but of course that doesn't feed back to Google.
It does surprise me that Google wouldn't want to capture this signal. Maybe it is too susceptible to abuse?
It surprises me too. Perhaps they are snow blind from having had something like it in the past and then removing it.
But in the modern era, I feel that being able to use human signals like that - on top of the fancy algorithms - could well be a killer network effect for them.
The SEO dreck drives more ad impressions than real content. Why would they improve the results when they have a captive audience who won't switch to other search providers.
That's a weird thought. Medium might be irritating as an interface or as a business but there are still some really interesting things written on it, no?
This is completely anecdotal but most medium (and similar platforms) that I end up on from HN are basically an A4 page of what is essentially nothingness.
This is not to say it’s not interesting, because it is, but what typically happens is that I end up in the “want to know more” state, but then don’t actually get to know more. This is not really on the medium authors, it’s more just my social media consumption taking me back to HN and then never looking into it again, but that also means that my time on the medium article was sort of wasted doesn’t it?
I can’t for the life of me remember a single medium article that had any sort of impact on me or anything I do tech wise.
There may be good content on medium, but more often than not I'm frustrated by long-winded, less accurate retellings of official documentation. I think it might be because of the monetization. People have an incentive to write shallow blog posts/tutorials on things like "throwing an exception with Python." If I'm out of options I will skim through a Medium post, but usually I avoid them at first because they tend to be less helpful than non-Medium blogs, official docs, etc.
I've found the Medium posts that make it to Hacker News to be drivel: Uninteresting and inaccurate. The titles themselves are interesting, but the content itself isn't worth reading.
It's happened so much that I just won't click on anything that's on medium.com.
I've noticed the same thing, but (like n=2, so take it with a grain of salt) even for authors whose other writings I enjoy; I wonder if there's something weird with how they structure the incentives for writing there such that this is what you get
There is a programmable search feature [0] that lets you limit search to a defined list of sites. Someone did a ShowHN a few months ago where they had built a programmable search with 200ish common sites that a stereotype HN reader might like (software documentation, wikipedia, reddit, some news and other media, etc), and it was actually pretty good.
I've said before, google is now basically what I'd call a "smart" portal site. For most stuff, you already know the handful of sites you might want to look at, and google just sort of brings you there from a relatively clean interface, as opposed to a traditional portal that would have lots of categorized nested links to traverse. In most cases you're not searching for a random site that you wouldn't know existed if it wasn't indexed, like in 1998. So the whitelist approach actually works pretty well.
Current Google search doesn't add features, it removes them (see end of + sign and other stuff). This is sadly the logical behavior of a "mature product". When you own the market, your rational question is "how much can I extract from the customers who already must come to me" and that means becoming more and more directive towards the customer. Letting the customers customize only stands against this.
The problem, as always, is credibility. If they're not careful about the credibility of the aggregate blocking signal, this would become another tool in the SEO scammer toolbox as they use sockpuppet accounts to downmoderate their competitors.
This is a great idea. On YouTube I use the 'Don't recommend channel' button a lot and consequently the algorithm has learned to filter out a lot of the nonsense.
This button doesn’t really work for me. Often I need to click it multiple times before a video dissapears from my recommendations. Same with the ‘Not interested’ button.
I’m guessing YT works with a system that is ‘eventually consistent’, but having to click the button 2-3 times before a video disappears doesn’t seem particularly consistent to me.
I don’t care, they are a blight on the internet that deserves to end. I have NEVER wanted to see a Pinterest board in my Google search results image or otherwise and their SEO optimisation serves only to parasitically drive traffic in order to feed their conversion funnel.
If your Pinterest board is public and you let others see it that’s nice, but the Pinterest developers should never have let that show up in my search results, your board is not and will never be more interesting than the original content you have pinned from other places on the internet, shitty aggregate pin pages hurt original content producers by ranking better and stealing page views and preventing users from navigating to the original content without jumping through hoops in the Pinterest user interface.
> Blocks sites you specify from appearing in Google search results
This extension prevents the sites you specify from appearing in Google search results.
I use uBlacklist in Safari - highly recommended. In Chrome I have a custom search string that adds "-pinterest -quora" because result on those spammy sites are behind a login wall
Ironically I tried to find a blacklist extension or similar through Google search past week, and couldn't find it due to blogspam talking about how it used to be. So added reddit and clicked in past year, but reddit now doesn't show accurate dates on posts so that didn't work.
> So damn annoying when the top search results all lead to shitty SEO-optimised sites ...
The problem is, since these sites dominate the SERPS, there is little incentive for anyone to offer the result you want. As a consequence, the web page that would satisfy your request probably doesn't exist.
This is not only a problem with Google search. DDG is now also flooded with auto (AI?) generated content. For example there are now loads of websites about programming topics. They look real but always lack the main point of the article which makes you navigate to other articles on the same website.
kagi.com allows you to adjust from the results page how often you want to see a page in the results or just block it entirely. I don't think there is a way to temporarily ban though.
That was a weird example to me. Google, although it ends up returning seo spam like w3schools, usually surfaces good SO answers, which are imo (and I think generally recognized as) the best answers that also contain discussion and caveats, if warranted. It's considered bad practice already to just cut/paste the SO answer unthinkingly. The only thing worse would be to do the same thing with whatever this search engine dragged out of w3schools.
There are lots of criticisms of google search, but in this case, I think the Google result is better - anyone looking regularly for answers is conditioned to tune out the Geeksforgeeks and w3schools and other spam, so as long as good SO answers are from and center, I think google wins
It's a middle man is all imo. I have actually used it for css stuff, but for python (as a python programmer), I want SO or the actual documentation , so its annoying when what I consider ad-driven content gets in the way of my search.
I can tell you why I hate them... explanations are usually incomplete or wrong in some cases. Examples often use antiquated browser code when the answer should just be vanilla JS. They don't use best practices... like ever. They don't generally use modern code (don't use arrow functions or continute to use var). MDN is pretty much better in every way.
For CSS and Javascript I've found MDN to be more comprehensive.
I do use w3schools for SQL syntax though - I haven't found any better resource for this, and generally Google gives me MSDN articles about TSQL which are usually of no use to me.
Google lags behind in Image search quality too, surprisingly. Bing consistently does better for me at least.
Google also lags behind searching for torrent content, not surprisingly.
In fact, I'm going to say, I use Google knowing that it sucks in many areas, just because it's hassle to use multiple search engines, and the quality was acceptable enough that it got the job done.
But now, I do more searches in both Google and Duck.
Their Youtube search engine is starting to suck too, because it's deliberately mixing completely unrelated items in the result.
Google sucks at image search but I just wanted to add, Bing either intentionally or unintentionally broke one of my porn searches. I'd search for "my-fetish-keyword-of-choice" and then pick to only show results for the last month. Since like January of February the results have been zero which I know, based on all the other places I look for content, is false. Content from sites that otherwise show up in Bing results. And no, I'm not looking for anything remotely illegal.
One Google feature that I miss terribly is hard filtering.
Used to be if you included "term" or -"term" you'd only get results that did/n't include those terms. But it seems Google has gone all in on the "I don't think you really meant that" approach [], and the hard filters have become suggestions at best.
--
[] Ok, I know it's probably because they're switching more and more to semantic search and ML, but they could retain the hard filters on top.
>[] Ok, I know it's probably because they're switching more and more to semantic search and ML, but they could retain the hard filters on top
The issue is that "power users" that are even aware of quotes and and hard filtering are now the long tail that google is no longer optimizing for. They'd much rather focus on the 99% of searches by, for lack of a better term, normies, and as a consequence, rather than expecting users to learn to think, their search features and performance are regressing toward a totally dumbed down mean. And I think society is worse for it.
Do you have any examples of searches that don't respect hard filters? The only time I've seen it ignoring quotes is when there are 0 results and it puts up a little notice at the top "No results found for "…". Results for … (without quotes):"
I mean, when you think about it, it's kind of obvious that an entity extracting revenue from selling advertisement impressions is not going to produce a very useful search engine — it's going to produce a search engine that maximises advertisement impressions. Either by encouraging the advertisers to game the search engine, or by simply preferring the paying advertisers' results, or both, on some level.
So you end up with a Google that prioritises 2-3 promoted results above actual search results (some of which represent the opposite of what you're looking for!) and everything beneath is either a massive mainstream content factory, a Reddit/SO/Quora thread, or any one of a billion terrible blogs/news hosts that contain no content or simply regurgitate someone else's content with adverts, modals, etc galore.
In fact, the only reason why Google's search engine is fairly safe in its product space — the considerable head start on potential disruptors notwithstanding — is that there apparently exists no comparably successful method to extract revenue from running a search engine.
Not that many people are realistically going to pay a monthly fee to have a Google without the noise, despite the amount of noise there is (I probably would). And let's face it, if there was a market for that, Google Premium would probably exist already. Talk about vertical integration if they did though. Help pollute the ocean of the internet then sell people premium membership to sail across it rather than swim in it.
One of the biggest things that Google did wrong was try to act as curator. The moment they start screening results by compliance with Google standards, introducing stuff like AMPHTML etc, anything like that is the moment they make themselves no different to Facebook, Twitter and every other walled garden community.
The internet is supposed to be about broadcasting information, about exploration and chasing the horizon — not locking information behind forced memberships of social networks and paywalls, tardis-like megastructures where you're encouraged to become locked in yourself.
All that should matter is that a webpage's content matches a query. Not whether their website matches some random person's idea of good UX etc. Does the content match the query. That's it. Refined obviously to assess whether a page's content is too insanely well matched (old SEO bullshit) and maybe include that domain rank stuff (if it was based on conversions so it's self-moderating rather than Google saying "[majorpublication].com is better than [randomblog].com because big business website > random person's website")
Google could have been worse, of course. AMPHTML is pretty much over, right? They track your data, yes — but show me the company that doesn't do that. Apple? Apple probably do, they just track less or whatever. Just because it's in the marketing doesn't necessarily mean that they don't do it, it just means that they know people will buy their shit if they say they don't, or make a point of doing it less.
I don't blame Google for the way their search engine has turned out. As others have said, a big part of it is the sheer amount of noise out there now. But more importantly, the way capitalism works causes most companies to produce increasingly shitty products over time. This eventually creates the opportunity for someone new to release a great product, and eventually their great product becomes a good product, and eventually that will become a shitty dividend-paying product, and the cycle will continue.
(to clarify, the opportunity isn't just there for competitors, the opportunity also exists for Google to sort their shit out too).
For Python programming I would say 100% of the time you should look the answer up in the official manual for a well-defined problem (delete a file) because the manual is correct, well-written, etc. It's astonishing how often Google and Bing snatch defeat from the jaws of victory on queries like this.
If you go looking in splogs, spam overflow and other spam sites at best you are going to get wrong answers, at worse you will get answers that "aren't even wrong".
Of the three code related searches in the original blog post, only one Google example listed the official Python documentation as the first link. In the other two examples, the sites shown where:
- W3 Schools
- StackOverflow
I will definitely agree with you that the author seems to want very different things from their search results than we might though. I will always always prefer the official docs (which I can pick through) when I make as vague a search as "<language> throw exception". If I wanted to know "how do I <verb> <noun> in <language>", then that's what I'd Google.
>because the manual is correct, well-written, etc.
It's also verbose and requires effort and working memory to parse, when I could get a trivial one line answer or code snippet from a stackoverflow post specific to my question. Yes, in an ideal world we would all read the manual, but unfortunately manuals are inconvenient. And in my experience the vast majority of stackoverflow answers are correct.
I get so annoyed that geeks for geeks is the default thing for so many things I search for in python. Really, it would be far more useful to just bring up the python docs most of the time.
I haven't coded python in a while, but does the python manual follow the man pages style of listing every possible parameter including the weird deprecated options that haven't been used since 1998 before actually providing information and examples of typical usage and instructions on how to handle the common use-cases?
Because that's the software equivalent to recipe spam.
Top 4 result is "Miscellaneous operating system interfaces", which does hold the answer, but it is not obvious, and browsing through that page is quite a chore before you finally get to `os.remove`, which says that it deletes "a path", which even I, a seasoned developer need to look twice to make sure that path removing a path and a file is the same thing.
https://docs.python.org/3/library/os.html?highlight=delete%2...
I don't feel like ranking results according to a query makes a lot of sense nowadays. Google works pretty well that it is almost always able to give you results that are relevant to your query.
But where Google fails and other search engine is to show results that are relevant to the user.
As some comments have said here for the query "python throw exception" some people want to get the official python doc when others want a snippet and when other want a tutorial, and some want other things. The fact that there is only one first result and it is the same for everyone IS the issue.
In my personal experience Google works very well for code queries much better than any queries on SEO topics
You're exactly right. Google tends to regress to the mean - what do most people want to see as their first result? but You.com allows you to prefer sources that you'd personally like to see first
361 comments
[ 6.5 ms ] story [ 407 ms ] thread"When did Neil Armstrong set foot on Mars?"
https://news.ycombinator.com/item?id=28224730
Edit, yes I'm wrong, did a mental s/mars/moon/ without noticing.
BTW, a big reason for this is the search quality folks at Google left the building and got replaced with growth marketers.
Is this the transition they made to AI powered rankings a couple of years ago?
To me it just seems like a query that is likely to trip any imperfect entity up, since a "when" question usually implies that the event in question is known to have happened.
Still can't figure it out. Lots and lots of articles about dead Russian generals.
There's got to be some way where the search engine doesn't second guess you. You can never find the answer to something adjacent to a popular question with the current state of things.
Even on colonel level the losses seem to be small - I think that this https://www.republicworld.com/world-news/russia-ukraine-cris... this was the first Ukrainian ground force colonel KIA that I recall, there was at least one air force colonel before that.
The other thing is that Russia is operating WW2 style. It’s a top down system where the lower level people have no autonomy. American colonels and generals die in helicopter shoot downs and accidents. Russian generals get assassinated on the front screaming at soldiers to move trucks, etc.
> Showing results for "When did Neil Armstrong set foot on the Moon?"
> Show instead for "When did Neil Armwrong set foot on the Moon?"
However, it seems to only work for actual spelling errors (e.g. if I typed Armstrong incorrectly) vs. context errors.
So to answer your question, I'd expect Google calling out its interpretation, but not necessarily a "never".
The interesting part to me is that Google will show redirected queries from typos (“Did you mean “When did *Neil* Armstrong set foot on Mars” ?”) but not disclose the term it completely ignored.
I makes it look like it parsed and validated all the search terms before coming up with the prominently displayed date, which makes it worse.
On your point, it’s broken because it shouldn’t display “Never”, and instead skip the date widget and show that the page results are for the moon, like it does for typos and other kind of corrected search terms.
Or alternatively _always_ show which part of the query the results are based on. Failure to be transparent creates the problem.
It seems very improbable that a kid would ask that question without having any preconceived notion whatsoever of a person with that name famously stepping foot on an extraterrestrial body. It's overwhelmingly likely that they have the famous 1969 event in mind, and that's why Google's response is appropriate.
Of course I'm not disputing that it wouldn't be even better for the Google response to explicitly correct the mistake in the query. I'm only disputing the fact that it's "broken." If you were asking this question of a human as some sort of test (but the human didn't know they were being tested), it would clearly be a "gotcha," and you could likely "fool" highly educated people who know full well that it was the Moon and not Mars.
Aside: Going to get real confusing when another Neil Armstrong born in the 2010s does set foot on Mars in the 2030s. I wonder if we will still get the first response from Google, 1969, and if you will still consider it to be 'not broken' then too.
This is I think the critical point.
It’s one of the core assumption that is correct for us to have in our day to day interactions with other real people, and we also are able to adjust the probability levels looking at the person or the context, and follow up depending on the reaction to our answer.
None of that applies to Google Search [0], and we are fed with “most probable” results without qualifiers, little to no sensible adjustments, and very little control or opt out options.
Billions of people use Search every day, at these scales what is “improbable” actually happens millions of times, and I feel too many people are willing to throw the odd ones out under the bus, even if the current situation isn’t perfect either for the 90 part of the 10/90 split.
[0] search personalization based on logged in profile could be used, but in practice they only apply that to very crude adjustments like country, language and frequent searches.
However, in this particular case, I think the very low probability of someone genuinely asking when Neil Armstrong set foot on Mars, combined with the very low probability of any measurable harm being done to those people by Google's response, makes me conclude that this is reasonable expected behavior and not something I would call "broken."
On the Google side, I think it’s an issue that has more consequences than just the moon landing.
For instance, for me “When is the francis election” (where I would have men “french or france instead of francis”) gives me a big and bold “Mar 13” with smaller below “Anniversary of the election of Pope Francis Observances” and an long anniversary Year/Week/Date table taking 80% of the widget display.
And as with the other examples, there is nothing showing the word approximations that has happened regarding to the original query.
There must thousands of other instances where a search result will come up with a big widget, an answer in big and bold font, except it will be completely wrong and have a direct impact on the user missing a deadline or taking the wrong action.
Of course users are supposed to know better and check the full result, but as you point out, if it’s almost always what they expected, they’ll learn to rely on it and be more complacent.
It also seems rather unlikely that someone would search for "2+22", but not obviously any more or less unlikely than "2+2", and building a calculator into your search engine is trivially easy compared to handling natural language queries, so it's not a useful comparison.
> Whether the user meant to ask a different question is irrelevant.
Of course it's relevant. If Google can correctly determine what question the user meant to ask, then obviously Google should provide an answer that provides value to the user instead of trolling the user with nitpicking about spelling or what have you. If you ask Google for the "capitol of India" you'll get New Delhi, even though that is the "capital" of India, and arguably the "capitol" of India is actually the Chandigarh Capitol Complex, so the result is "objectively wrong".
I would argue that it's roughly just as clear that the query intends to refer to the Moon as it is that the query intends to refer to the Canadian Neil Armstrong who was killed in a plane crash in the Antarctic in 1994.
Yes, people probably don't make this exact query in good faith very often, but it raises significant concerns for other queries where google tries to supply an answer.
See for example: https://gizmodo.com/googles-algorithm-is-lying-to-you-about-...
Meanwhile for Firefox, I just have to go to "Search" and the setting for default search engine is immediately obvious. Chrome seems to have a similar layout as well despite being produced by a search engine company. Edge is my daily driver, but it still took me way longer to find the default search engine setting in Edge compared to Chrome and Firefox.
Also, I have to compliment Firefox for making it really easy to search with a non-default search engine. I generally use Google, but I use Bing when searching for internal work stuff since it is integrated with O365 and SharePoint.
Didn't seem all that difficult to me.
I went to settings, put "search" in the search box. That highlighted in yellow "address bar and search" click "manage search engines" click, 3 dots next to google, click, make default, click. 4 clicks with guiding highlight throughout. Didn't even need to google how to google with edge.
I'm still using FF. However, switching search engines in Edge seems about the same difficulty as doing the same in FF or chrome.
not a search example but it's like Jamie Oliver's fried rice
example -- mango avocado -> click recipe at https://breezethat.com/?q=mango+avocado
Google still solves all my questions, if I'm asking the right questions, I guess is what I'm getting at.
I used to think this way too, but I've come to realize that Google has turned from a search engine that returns results based on my input to an answer engine that actively tries to reframe whatever I enter into some other more generic query - usually with the intent of selling me something or returning SEO spam. I've also found that the old google-fu techniques are now so unreliable that they must have been deprecated. I can't tell you how often I use quotes in my query and see results that don't contain that text at all.
> dog apple dangerous
Today you should search:
> can dogs eat apples
And you get way better results with the second form. I've noticed that people who are stuck thinking that the right way to google is still the former overlap a lot with the crowd that keeps complaining google is worse now than it was before.
Otherwise, it there truly is a “right way of asking Google questions”, why doesn’t Google release and promote a guide about it so people can be more successful in their search?
https://www.google.com/search?hl=en&q=dog%20apple%20dangerou...
https://www.google.com/search?hl=en&q=can%20dogs%20eat%20app...
Also worth comparing (seems to focus more on diabetes than cyanoglycosides. Diabetes is the bigger chronic problem, and cyanoglycosides are the bigger acute problem, so which is a bigger danger largely depends on how disciplined the owner is.):
https://www.google.com/search?q=what+is+dangerous+about+feed...
Though, even when I was doing indexing changes at Google, the common practice was to do A/B testing with both the most common queries and a uniformly random sample (see reservoir sampling) of queries in order to justify a go-live of indexing changes. The former explicitly over-weights common queries, and the latter still optimizes for the common case. (In case you're wondering, the worst query I had to manually check in A/B testing was [flesh hook suspension].)
Google used to turn off some of the query re-writing logic (that tries to fix your query) if you used a query operator. (It has been a while, but I think maybe even their "Kansas" user info database kept track of the last time you used an operator, and would turn off some of the cleverness if you had recently used a search operator, as it was a good signal that you were a power user capable of optimizing your own queries.) My understanding is that they don't disable any of the too-clever bits for power users any more, and that everything uses all of the cleverness of learn-to-rank all the time.
I suspect it has gotten even worse with learn-to-rank, as it must be incredibly difficult to intentionally under-weight the uncommon/difficult queries.
They did keep track of when users re-issued similar queries in a short period, as a signal that the ranking algorithm wasn't doing well. I think an optimal system would use learn-to-rank for the first query in a related sequence of queries, and then switch to turning some of the smarts off, and finally switching to a learn-to-rank algorithm trained only on later queries in these related query sequences. That way, they can avoid the secondary learn-to-rank instance from over-fitting the median/easy queries.
Google has come full circle and turned into Ask Jeeves.
I worked at Google on rich content indexing for 4 years, more than a decade ago now. Google is pretty good. It used to really cater to "long tail" searches, but a combination of SEOs getting better (and specifically targeting Google) and "learn to rank" over-optimizing for the median query, means that lots of long-tail queries don't do very well on Google.
I just wish I could find obscure things again.
Though, there's also a use case for people trying to find third-party information about various websites, particularly in trying to figure out if the website itself is a scam. You really want to have the navigational result up at the top unless you have a very high degree of confidence that the site is a scam, but in all cases you want high quality reviews of the site to follow up the navigational link.
Google is _not_ a good guideline for asking the right questions. It is trying to make as much money as possible and it will scramble anything it can think of to do that. Don't use that as a guideline for how you are thinking about questions.
2. ???
3. Make as much money as possible
oh Haha, funny old meme. Step 2 though is very clearly to show more ads and more content, to show more 'full results pages', etc, rather than just show me the 1 thing I needed and be on my way.
I should include that Google said next to each result ('"555-555-555" is not on this page') and then show the page. Totally knows that the results are unrelated to my query, but shows them anyway. Why?
And by that you mean, Google is telling you how to think, and that you should think in some way and not in some other way. I'm quite sure I am not comfortable with that.
The people at W3schools are a combination of all of the above: filthy SEO spammers who would readily make an antitrust stink if Google started just ripping off their content, but who in all likelihood don't care if Neeva does that, if they've ever even heard of Neeva.
They could even take the blacklists from all people with a high reputation and use that as a signal when they are building search results for others.
I'd prefer a browser extension that removes results locally. Actually all I need is probably just a rule in uBO.
Google could be better than it is now, but there's no incentive to do so, unfortunately. Say Google allowed you to blacklist entire sites from the results - inevitably those sites that have the most ads would be the most likely to be blocked, resulting in lower revenue for Google.
Recipe sites are notorious for SEO tactics. They all follow the same highly optimized format with the stupid story about the author's grandma and how they just couldn't get enough of these cookies, and how the recipe was lost for 90 years until recently their great great uncle Lou found a copy of the recipe in an old donut.
Google has all of the tools to solve recipes. Make Google Recipe with a standard template and a way to link in and out of YouTube. People who contribute popular recipes get ad revenue. People with recipes and YT videos get even more. Adding ways to find similar recipes would be a killer feature. Who hasn't found a recipe that was almost what they were looking for, but was missing that je ne sais quoi.
All of these recipe sites (that I've seen) will drop the ingredients + directions on the bottom of the page.
Last one is the killer, a site that quickly gives you the info you are looking (or quickly shows you it doesn't have the info) for is punished by google search ranking.
I LOL'ed. Thank you for that.
I'm pretty sure i'm not the only one who clicks on pinterest results in eg. image search, and immediately click back to find another image somewhere else.
Did I spend a minute or two trying to read a 1000 word article, because I was looking for how many pixels there is in a 4k monitor? Or did I spend 2 seconds visiting a chart?
It usually takes me at least a minute to recognize a search-optimized text if it's a topic I am unfamiliar with. Let's say I am googling something about windproofing underfloor insulation. The article starts with some basics about underfloor insulation in general, so I skim through the introduction, start hunting for the part of the article where it actually starts talking about windproofing and realize it's been cobbled together out of six random introductions or generated by GPT-3.
Five to ten years ago no other engine was close - the fact that duckduckgo was even try was comical. Google had an effectively monopoly. That might not be true at all five to ten years from now if the trend continues.
If I want someone's actual opinion and people I know don't have one, I have to check reddit or metafilter.
https://blog.google/products/search/more-helpful-product-rev...
Side note, I prefer DDG as my search but only because of the bang operators. For recipes !b added to the search lets me use Bing. As the article points out, Bing is really awesome for searching for recipes.
Looks like I need to start trying out Neeva & You.com. They had some nice features in this article.
Neeva seems to be stealing all the important content from the recipe website which is a highly discussed issue. Bing tries to walk this line by making you still go to the website to read the instructions. Obviously people have trashed Google for doing this same thing on other kinds of websites. Though blogger recipe websites have somewhat encouraged this behavior due to their insane amount of ads & life stories they're well known for.
https://old.reddit.com/r/linux/comments/qqpggk/new_youcom_pr...
You’re holding it wrong :-)
So they essentially don’t want you to click on the organic link that points to Volvo.com.
They want Volvo to know that all the traffic to them is being sent due to the Ad and not from any organic links.
This is what the not so evil company is doing, imagine if they are actually evil..
I liken it to a defense strategy.
Imagine if they didn't, and a search for Volvo XC 60 showed an ad for a Subaru!
https://twitter.com/jasonfried/status/1168986962704982016
Anyone else has other ways?
I can’t relate to this. That example, of the first result being the canonical documentation on the subject, is a search engine working exactly the way I want.
I always found the MDN documentation striking a better balance of being in-depth but not an essay.
We had a bereavement in the family and had to book multiple independent tickets because we could not travel together. This was just after the Ukraine war started so prices had gone through the roof. Exact number of days was not that important as compared to the price and the duration and using google flights UI to slice and dice the data was such a joy. Want to freeze the airline and look at the alternatives - which include from and to dates, number of days of trips, or freeze any other parameter and analyze others, the response was sub-second. Did not eventually book through them since I did not want to get into the google payment system (they offer booking through others that I did not explore).
On the opposite side was Expedia children where they would show a price of 2800 and when you click the price invariably it has gone up to 3600. Again. And again. And again. Not sure if that problem existed with google although I paid the exact same price as the airline as shown by google, it could just be a coincidence.
[1] https://googleblog.blogspot.com/2011/04/ita-software-acquisi...
https://oldmatrix.itasoftware.com/
Also, it looks like the Matrix might go extinct soon:
> This interface runs on a deprecated web platform. At some point in the near future, we will be forced to shut it down. We unfortunately do not have a timeline on when this will happen. We welcome feedback about features missing from the new interface, we read all feedback and open bugs accordingly.
2. I don't use it, I point out for others facts about service OP recommends so they don't waste their time or you prefer everyone wasting their time?
The difference between Kayak/Google and the others is that Expedia and friends are online travel agencies. Kayak and Google are really just search engines. It makes a world of difference.
Let's say you want to get from the West Coast to Europe on a business class flight, but want to save some money.
Realistically, it's cheap and easy to get from West Coast airport to another, and similarly cheap and easy to get form one European airport to another.
Google flights will let you search for the best combination of flights that depart from any combination of up to 6 airports, and arrive at up to 6 airports. The technology they purchased (ITA) will let you do this as well, but limits you to a single country. Google flights? No problem with destination airports in multiple countries.
So, you could search for a flight that originates in some combination of, say Seattle, Portland, San Francisco, Los Angeles and Vancouver BC, and lands in some combination of London, Amsterdam, Paris, Madrid, Milan or Frankfurt. (Or any other 6 large European airports).
Google Flights will also show fares for a specific trip length (say, 14-days) over an entire month.
Want to filter by a maximum flight duration? No problem. Number of connections? Done. Specific airline or alliance? Easy.
In a traditional search engine, I'd have to run close to 400 individual searches to get the data that Google Flights gives me in a single screen.
It works well for domestic flights as well. I recently helped someone who had to attend a wedding across the country find a flight that was less than $180/person, when they thought they were going to have to pay over $800/person. Just by using Google Flight's tools for about 10 minutes.
There's plenty to complain about in the Google ecosystem, but Google Flights is amazing.
So damn annoying when the top search results all lead to shitty SEO-optimised sites that use a whole page to blather on and on, leading to a tiny information nugget at the end. No value, just excellent SEO scamming.
As these scam artists get better and better at this, Google gets less and less useful.
When I see a site like that, I can be quite sure there is no value to me from that site. I want to blacklist it - not forever (though I'd settle for that) but so I don't see it in search results again.
The crazy thing is that this could even be a benefit for Google themselves. They could aggregate these signals and use them to identify SEO scammers, since their algorithms clearly can't. I'm sure that Google aren't happy with the lacklustre performance of their search in modern times.
There are numerous extensions to block domains, like uBlacklist: https://chrome.google.com/webstore/detail/ublacklist/pncfbmi... (not affiliated)
It does surprise me that Google wouldn't want to capture this signal. Maybe it is too susceptible to abuse?
But in the modern era, I feel that being able to use human signals like that - on top of the fancy algorithms - could well be a killer network effect for them.
The incentive to block one's competitors from 10,000 different accounts is likely why they no longer offer this function?
I wish I could do that at Hacker News too.
I really just don't want anything from medium.com.
This is not to say it’s not interesting, because it is, but what typically happens is that I end up in the “want to know more” state, but then don’t actually get to know more. This is not really on the medium authors, it’s more just my social media consumption taking me back to HN and then never looking into it again, but that also means that my time on the medium article was sort of wasted doesn’t it?
I can’t for the life of me remember a single medium article that had any sort of impact on me or anything I do tech wise.
It's happened so much that I just won't click on anything that's on medium.com.
I've said before, google is now basically what I'd call a "smart" portal site. For most stuff, you already know the handful of sites you might want to look at, and google just sort of brings you there from a relatively clean interface, as opposed to a traditional portal that would have lots of categorized nested links to traverse. In most cases you're not searching for a random site that you wouldn't know existed if it wasn't indexed, like in 1998. So the whitelist approach actually works pretty well.
[0] https://developers.google.com/custom-search/
HN comments: [Show HN: No Trash Search](https://news.ycombinator.com/item?id=29774456)
It is absolutely impossible to find anything anymore, I gave up on google for all intents and purposes and use it exclusively as reddit indexer.
They get crowds of people and/or scripts to blacklist all other sites.
I’m guessing YT works with a system that is ‘eventually consistent’, but having to click the button 2-3 times before a video disappears doesn’t seem particularly consistent to me.
I can’t think of any reason why a spam Pinterest link has any value, yet it’s ranked high.
There are a lot of real world people I meet who speak highly of all the great creative ideas they get from there.
If your Pinterest board is public and you let others see it that’s nice, but the Pinterest developers should never have let that show up in my search results, your board is not and will never be more interesting than the original content you have pinned from other places on the internet, shitty aggregate pin pages hurt original content producers by ranking better and stealing page views and preventing users from navigating to the original content without jumping through hoops in the Pinterest user interface.
Any ranking manoeuvrability that Google offers can and will be used against them.
The SEO scammers out there would just automate millions of proxy IP addresses to blacklist all of their competitors sites.
https://github.com/iorate/ublacklist
https://chrome.google.com/webstore/detail/ublacklist/pncfbmi...
Thanks, will add this!
The problem is, since these sites dominate the SERPS, there is little incentive for anyone to offer the result you want. As a consequence, the web page that would satisfy your request probably doesn't exist.
There are lots of criticisms of google search, but in this case, I think the Google result is better - anyone looking regularly for answers is conditioned to tune out the Geeksforgeeks and w3schools and other spam, so as long as good SO answers are from and center, I think google wins
I do use w3schools for SQL syntax though - I haven't found any better resource for this, and generally Google gives me MSDN articles about TSQL which are usually of no use to me.
They’ve since changed their stance, but I think the website has not substantially improved since w3fools’ inception.
Google also lags behind searching for torrent content, not surprisingly.
In fact, I'm going to say, I use Google knowing that it sucks in many areas, just because it's hassle to use multiple search engines, and the quality was acceptable enough that it got the job done.
But now, I do more searches in both Google and Duck.
Their Youtube search engine is starting to suck too, because it's deliberately mixing completely unrelated items in the result.
I use Yandex for image search, much more variety in results and the interface is also quite good to quickly go through a lot of images at once.
Used to be if you included "term" or -"term" you'd only get results that did/n't include those terms. But it seems Google has gone all in on the "I don't think you really meant that" approach [], and the hard filters have become suggestions at best.
--
[] Ok, I know it's probably because they're switching more and more to semantic search and ML, but they could retain the hard filters on top.
As an example, if I search for 'gilbert gottfried' and set the date filter 1 apr -> 3 apr, I still see stories about his death.
The issue is that "power users" that are even aware of quotes and and hard filtering are now the long tail that google is no longer optimizing for. They'd much rather focus on the 99% of searches by, for lack of a better term, normies, and as a consequence, rather than expecting users to learn to think, their search features and performance are regressing toward a totally dumbed down mean. And I think society is worse for it.
It's extremely frustrating how much "content" is displayed on the page versus how much is hidden in the page source.
So you end up with a Google that prioritises 2-3 promoted results above actual search results (some of which represent the opposite of what you're looking for!) and everything beneath is either a massive mainstream content factory, a Reddit/SO/Quora thread, or any one of a billion terrible blogs/news hosts that contain no content or simply regurgitate someone else's content with adverts, modals, etc galore.
In fact, the only reason why Google's search engine is fairly safe in its product space — the considerable head start on potential disruptors notwithstanding — is that there apparently exists no comparably successful method to extract revenue from running a search engine.
Not that many people are realistically going to pay a monthly fee to have a Google without the noise, despite the amount of noise there is (I probably would). And let's face it, if there was a market for that, Google Premium would probably exist already. Talk about vertical integration if they did though. Help pollute the ocean of the internet then sell people premium membership to sail across it rather than swim in it.
One of the biggest things that Google did wrong was try to act as curator. The moment they start screening results by compliance with Google standards, introducing stuff like AMPHTML etc, anything like that is the moment they make themselves no different to Facebook, Twitter and every other walled garden community.
The internet is supposed to be about broadcasting information, about exploration and chasing the horizon — not locking information behind forced memberships of social networks and paywalls, tardis-like megastructures where you're encouraged to become locked in yourself.
All that should matter is that a webpage's content matches a query. Not whether their website matches some random person's idea of good UX etc. Does the content match the query. That's it. Refined obviously to assess whether a page's content is too insanely well matched (old SEO bullshit) and maybe include that domain rank stuff (if it was based on conversions so it's self-moderating rather than Google saying "[majorpublication].com is better than [randomblog].com because big business website > random person's website")
Google could have been worse, of course. AMPHTML is pretty much over, right? They track your data, yes — but show me the company that doesn't do that. Apple? Apple probably do, they just track less or whatever. Just because it's in the marketing doesn't necessarily mean that they don't do it, it just means that they know people will buy their shit if they say they don't, or make a point of doing it less.
I don't blame Google for the way their search engine has turned out. As others have said, a big part of it is the sheer amount of noise out there now. But more importantly, the way capitalism works causes most companies to produce increasingly shitty products over time. This eventually creates the opportunity for someone new to release a great product, and eventually their great product becomes a good product, and eventually that will become a shitty dividend-paying product, and the cycle will continue.
(to clarify, the opportunity isn't just there for competitors, the opportunity also exists for Google to sort their shit out too).
If you go looking in splogs, spam overflow and other spam sites at best you are going to get wrong answers, at worse you will get answers that "aren't even wrong".
I guess I just want different results to the same query than the author
- W3 Schools
- StackOverflow
I will definitely agree with you that the author seems to want very different things from their search results than we might though. I will always always prefer the official docs (which I can pick through) when I make as vague a search as "<language> throw exception". If I wanted to know "how do I <verb> <noun> in <language>", then that's what I'd Google.
It's also verbose and requires effort and working memory to parse, when I could get a trivial one line answer or code snippet from a stackoverflow post specific to my question. Yes, in an ideal world we would all read the manual, but unfortunately manuals are inconvenient. And in my experience the vast majority of stackoverflow answers are correct.
If it was was Clojure or some other language that has an awful online manual it is one thing but the Python manual is good.
And... I'm guessing w3school?
Because that's the software equivalent to recipe spam.
Mostly yes, you just search efficient documentation.
https://docs.python.org/3/search.html?q=delete+a+file&check_...
Top result: https://docs.python.org/3/library/configparser.html?highligh...
Top 4 result is "Miscellaneous operating system interfaces", which does hold the answer, but it is not obvious, and browsing through that page is quite a chore before you finally get to `os.remove`, which says that it deletes "a path", which even I, a seasoned developer need to look twice to make sure that path removing a path and a file is the same thing. https://docs.python.org/3/library/os.html?highlight=delete%2...
But where Google fails and other search engine is to show results that are relevant to the user.
As some comments have said here for the query "python throw exception" some people want to get the official python doc when others want a snippet and when other want a tutorial, and some want other things. The fact that there is only one first result and it is the same for everyone IS the issue.
In my personal experience Google works very well for code queries much better than any queries on SEO topics
This is going to sound tongue-in-cheek but I mean it with all sincerity: isn’t it wild given the data they hoover up?