In the end, I think most people could use either Google, Bing or DuckDuckGo and be happy with the results. It's just that we're stuck on "Google is the best search engine", and while that may be true in some technical sense, many of the other search engines are just as good for most of us.
I don't really notice any infrastructure problems for ddg in Europe though, so Finland may be atypical?
that's not what Incognito mode does. It prevents your search from being included in the browsing history and doesn't send cookies from active sessions, but that's about it. Google still knows this is you being unauthenticated. You don't need to be logged into google to be reliably targeted with ads that fit your profile.
Another example is a query for "elm dict". DDG has little idea what you're looking for while Google links you directly to the docs of Elm's Dict data-structure.
You know you're looking for a data structure, right? Add structure to your query, and DDG will do fine. Easy fix.
The opposite case, when google thinks he knows what you're asking but it's wrong, is impossible to fix.
I don't want them to be smart, I want them to be predictable and search what I say, not trying to reinterpret the query.
Most of the time I see comments like this, people don't provide hard examples of queries they found unsatisfying with one search engine compared to another.
There are advertising companies that use fingerprinting for ad targeting, but Google doesn't.
(Disclosure: I work at Google on ads, speaking only for myself)
Seems like a pretty good reason to think it still knows who you are.
As to how that works, you tell us.
One way you could see something similar to this would be if you opened a clean session, logged into Google, logged out of Google, thought you closed the last incognito window but didn't, and then opened a new incognito window? Then the user cookie would still be in client-side storage
https://meta.stackexchange.com/questions/331960/why-is-stack...
Try visiting from tor
It’s in their best interests to also use it for ad targeting (in a plausibly deniable way so they don’t get in trouble).
We’ve seen them using dark patterns to coerce users into opting into more data collection, and another advertising company got caught using phone numbers for ad purposes even if they originally promised to only use them for 2FA, so why should we trust them this time?
If it was being used for targeting it would be practical to run an external study demonstrating that.
Also, people are more knowledgeable about the field in which they work, so it makes HN strictly worse if the environment becomes so poisoned that they're disincentivized to participate.
That's not how I see that conversation:
* reaperducer was asking why it was useful for the browser to show that the page was one that usually loaded quickly/slowly
* As someone who had worked on an effort to speed up the web I replied with why I thought it was useful
* jfoster gave a good response describing why it might not have the effect I expected, since if users know a site is usually slow that may make them more patient
* I replied that this was still good, because users were in a position to make a better decision about whether to continue waiting for the site to load.
* You responded with something completely unrelated to what we were talking about.
* I tried to be helpful anyway, even though your comment wasn't something I knew much about.
* You continued in a direction that I don't know much about (how to communicate things like whether location tracking is on) and linked to a study which I didn't have time to read.
* This wasn't a discussion I was interested in, so I didn't respond. I don't see how the study you linked contradicted anything I was saying.
Jeff, you defended Google once saying that their decisions are motivated by wanting to help users make informed choices. The study was just to show that they have a track record of doing the exact opposite. Your characterization of Google was misinformed at least on that occasion. I made an educated guess that if you were willing to defend one stance that was proven wrong (that Google has any vested interest in helping users make informed choices) then it's possible you may make the same mistake again.
But dang is right, in the spirit of collegiality I should have found a better way to point out this mistake or even not do it at all.
I would be very curious as to how you’d prove this is or isn’t happening with a reasonable degree of accuracy considering all the factors involved in ad targeting. Unless you’re willing to give us access to all your source code and SSH access to the systems running it, it’s reasonable people have their doubts.
An external study to evaluate whether Google is using fingerprinting would be some work, but pretty doable. Targeted advertising is generally very blunt: if someone thinks you're especially interested in a valuable category they'll often pay a lot to advertise to you. So you could set something up where test browsers visit pages related to high-value categories (mattresses, asbestos cancer, credit cards, ...), clear client-side data, and then visit a site that loads ad scripts only from Google (to make sure you're not getting someone else's fingerprinting) and see whether the ads differ from a control group that never visited those pages.
And surprisingly for most of HN readers, Google has been pretty transparent on the policy of its ads business. In fact, Google has pretty strong incentives for transparency in this area due to advertisers, who give all the money anyway.
IMO, ads are probably the least worrisome way the data could be used. A boring but scary example is that aol search history leak (which is still searchable today):
This person is identified by name for example: https://searchids.com/user/19431784-joann_whitman
It's also a good example of their monopoly position; Android, Chrome, Chrome OS, advertising and analytics code on almost every website, ownership of multiple of the most popular websites and services on the internet puts them in a unique position that no one could ever hope to compete with realistically. Competitors have to rely on imperfect fingerprinting whereas Google can probably detect you with more accuracy than a DNA test.
That's literally the opposite of what s/he just said. The person you're responding asked "How would it know?", implying that they (while being on the Google ads team) think there is no way to know without fingerprinting (or cookies from non-incognito mode).
Personally, Google ads give me mixed feelings. I see how personalization is useful for everyone involved and, so long as only machines look at my data, I don't have any personal issues with it. But at the same time, Google collects everything on everyone worldwide to the point where I feel like the USA would have an easy time conquering any country they please (if a nation already has live data on pretty much all its enemy's subjects, war would be exceedingly efficient for them to start and quickly win), so that kind of threatens our freedom if you see what I mean; and secondly the data is not necessarily 100% secure, so in the event of a breach it might be seen by humans, specifically people that I would not want to know what I searched for (or pages I visited that have Analytics or an embedded YouTube video or ads or a map on their contact page or ...). So it's a mixed bag of feelings and your position (job) seems like the kind that would make one think about before accepting. I'm curious to hear your thoughts on it.
I've written some about this: https://www.jefftk.com/p/value-of-working-in-ads
"Many people would put ad tracking on this list of downsides: sites pass information to data brokers that build custom profiles for each user and allow personalizing ads. From my perspective, however, while having this information collected seems a bit creepy, it allows showing ads I'm more likely to be interested in. This makes publishers more money than showing untargeted ads, and I'd much rather fund them through better ad targeting (invisibly intrusive) than through more obnoxious ads (visibly intrusive)."
I chose this team because I thought the work would be interesting and I liked the people on it, and they were interested in me because of my prior work on mod_pagespeed rewriting websites so they would load faster.
> if a nation already has live data on pretty much all its enemy's subjects, war would be exceedingly efficient for them to start and quickly win
Lots of thoughts:
* I think you're dramatically overestimating how much data Google has and how well that is mapped to the kind of identity the military would care about.
* I don't think Google would share this information unless legally required to, and I don't think such a request would be constitutional.
* Many other countries are in similar positions; for example Criteo is based in France and has a similar ad tracking reach to Google.
* I'm still not sure how this is especially useful militarily. Military targets are mostly not in the data one of these companies would have, and none of these countries would go to war targeting civilians.
> I chose this team because I thought the work would be interesting and I liked the people on it
That is fair! I guess most people would make that decision if you already know people there and you think you'll enjoy the work as well.
> none of these countries would go to war targeting civilians
Not as if people in the army are somehow exempt from tracking though?
As for whether Google would share it in the first place: I don't think the government cares much what Google thinks if they're willing to kill (us) over something. Laws can be made by the same people that decide on this. I don't mean to pose it as a simple matter, but I'm pretty sure that's how it works in principle.
Now that I think of it: aren't "national security letters" exactly this? "It has something to do with the safety of the country, just give us that data [e.g. Lavabit private key]"?
Of course, the chance is remote in the first place. Much more likely, if it is ever used for this kind of purpose in the first place, it'll just be posturing and threats, and people will protect themselves better before it ever gets to armed conflict. Just imagine, though, if you're not in the USA, China, or Russia, and one of the three (the most democratic one of the tree, it is fair to add) has the rest of the world's data. That's kind of uncomfortable when I pause to consider it.
> how well that is mapped to the kind of identity the military would care about.
While not readily available, I expect that it's not hard to find a few datapoints to filter them out. Following someone for 10 minutes as they go through traffic and matching the coordinates against location history data is probably enough to find a subset of 1-5 possible accounts. But I doubt physical following is even necessary to find enough datapoints to find them in the data.
And if he really is saying that Google doesn't track you in incognito mode, then I'm going to go ahead and assume he's either lying, or he's not in a position to know about that system. This is Google we're talking about here.
https://www.theguardian.com/technology/2017/nov/22/google-tr...
For the record, I'm not saying that I expect Google not to track me when they detect some privacy mode. It'll sure try to set cookies, and it may use my IP address and connect whatever that IP accesses as a weak indicator of interest for anyone else with that IP address (for a limited amount of time, since IPs change in many countries). What I don't think is that, when they say they don't do fingerprinting, they're lying. This person may not be privileged to know and say "I don't know", but that's different from saying "Google doesn't".
Also for the record, I didn't downvote you (and when you reply to me, I can't; I don't have an alt account with 1k rep or whatever it is one needs to downvote).
Anyhow, based on what I've seen, when you go into incognito mode it definitely doesn't use your profile when you go to Google services. Search results, suggestions, etc... are different. In fact, one of the use cases for incognito is using a different Google profile on a someone else's or a public computer. Keeping contextual search results from the 'main' profile logged into the browser would be counterproductive...
Google also censors some results in controversial categories, where Bing/Yahoo return what I am looking for. I don't trust a search engine that censors entire categories of results.
I am not an adwords user but anecdotes in another HN thread seems to affirm that Google is due to be disrupted: https://news.ycombinator.com/item?id=21667484
DDG's first page was only fake sites and wrong answers.
Google gave me the right answer right away (in the positions 2 and 3)
(for those who don't know legacy electronics: I was looking for the box with labeled pins, like on first page of https://www.turus.com.tr/class/INNOVAEditor/assets/PDF/MM584... )
"free energy" use to produce countless results that now require extra keywords.
Under "free energy suppression" you find professionally crafted hatemongering. The actual list of claimants is huge, non of it is here. https://peswiki.com/directory:suppression#Wiki_2796702
"free energy device" also produces really crappy results compared to what it was.
It only seems like things argued not to exist may or must be scrubbed from seach results. Astrology wasn't scrubed nor was any religion. Everything has its history too!
The way they "cheated" was to look at the tens, hundreds, thousands, or tens of thousands of previous queries that matched yours in substance.
If the competing websites for "web search" had the same volume of submitted queries that are substantially similar to yours, then they too would be able to give you "exactly" what you are looking for "9 times out of 10".
If all people submitting queries on the web are somehow convinced to visit only one website and submit a majority, (hypothetically let's say 93%) of their queries there, then it should be no surprise when that website "magically" starts becoming more "predictive" than other websites in returning "exactly" what searchers are looking for and becomes "by far the best at search".[1]
The question raised by the blog post is whether enabling such "magic" is worth the trade-off of also creating a "panopticon" in that single website. Giving this level of visitation and query traffic to a single website makes any competing website (working off less than 7% of people's queries) seem irrelevant, maybe even pathetic.
Without the enormous traffic, I would surmise the glory of the "empire" (to use the author's chosen term), and its appeal to "followers" (e.g., those who marvel at its "magic"), might dissipate rather quickly.
1. Especially the part of "search" that involves dealing with repeated, similar queries for popular information.
Google's advantage is simply billions of dollars and 20 years of R&D into NLP tech.
Not directly, but it does mean that, say, the second 50% of people that ask the same question will get a better answer than the first 50%.
Google has been able to build fairly accurate instant results based on which sites users were clicking on before. I'd say that a majority of simple general knowledge queries are solved by quoting the first 3 sentences of the wikipedia page that match the search query.
But let's say that there is no easy match to show a quick result for. But after 1000 queries, 98% of users clicked on one particular site on the first page and never went back to the search results. Google then A/B trials how many clicks result from showing that website at the top of the results page in an instant result window. If clicks drastically drop, that's a sign most users are satisfied with that result. They were only able to do that, because of the 1000s of times people typed that query and interacted with the site.
So I'd say that yes, having common questions asked over and over again does help you find the answer that users are looking for.
The funny / scary part here, is that this may not be the correct answer. But it's the answer that satisfies the most users, and is therefore the one that will keep the most people coming back to the Google search engine.
I doubt that users have a strict expectation of finding the correct (interpretation of "corresponding") answer. Even if the most clicked answer[s] are still not correct, they may still be the ones who suck less, which can still be acceptable/desirable.
Keep in mind that in this perspective, the concept is similar to Google Translate - Google built a translator that, at least originally, doesn't understand language, instead, it applies (applied) a static model on a large amount of documents. While they certainly poured a large amount of money on it, it's success can't be centered purely on the economical factor.
There are at least three kinds of queries that a search engine has to handle. Requests for websites e.g. "Facebook". Traditional keyword searches across the web e.g. "Twitter ban political ads", and questions "Who was the guy who voiced bender?".
Plus, you're underestimating the percentage of queries that google has never seen before. Google processes trillions of searches every year, and still, 15% of those queries have never been seen by Google before. [0]
[0] https://searchengineland.com/google-reaffirms-15-searches-ne...
No. I am highlighting that 75% are queries that they have seen before. Note also that the 15% is a figure that is declining, based on what it was in 2007.
When you see google giving you a one/two word answer as a card, that’s very likely coming from it’s knowledge graph.
What we need is more of this open knowledge graphs. Google and Wolfram Alpha are both closed sources but have deep understanding in niche domains.
For example, say the film credits for Spiderman is a "primary source", and a cast list for Spiderman derived from the credits at IMDb is a "secondary source". Google extracts the information from IMDb and substitutes itself as the secondary source.
Does this raise an issue in that ideally users should sometimes be retrieving information from (i.e., accessing) primary and secondary sources directly, whereas a third party always acting as a universal secondary source, e.g., a third party funded by advertising, might introduce (more) bias into the information retrieval process, e.g., in competition for "eyeballs".
From the OP references: https://sparktoro.com/blog/less-than-half-of-google-searches...
It's also another matter that so much of the internet is now filled with copy-pasted crap and artificially inflated content that it's actually hard to find what you are looking for on the so-called primary source webpages. If I want to find George Clooney's age, I don't want to sift through a 2000 page ad-ridden Page3-esque gossip blog about him.
At the same time, when I want detailed info about something, I will go in and try to read through the primary material.
In any event, neither a blog nor Wikipedia would likely be a "primary source". Maybe something like a driver license would be a primary source.
It could be that you have a "philosphical objection" to "copy-pasted crap and artifically inflated content". I wonder if Google could have a role in encouraging the continued existence of this stuff. The effective opacity it creates seemingly justifies having an entity like Google.
Even when I entered the actor's name plus the term "age" and was redirected to the Wikipedia search results, I could still see the actor's page as the third result and his date of birth in the summary text.
As for why anyone would want to search some things using Wikipedia versus Google, I can think of a few reasons. I cannot speak for other users however.
I have a script I wrote to search Wikipedia from the command line. For those who care, one can strip out "X-Client-IP" from the returned page html before opening the page in a browser.
If you want DDG to understand natural language queries, I think their privacy policy may need to be adjusted so that our queries can actually be used to develop that, and then they need a boatload of funding to catch up with Google's semi-ethical money-generating practices.
!bangs are the main reason why it's always going to be difficult for me to switch from DDG to anything else.
By contrast, the new privacy friendly search engine from the article, with a lot less money and users, can answer a simple question like "news guy actor in spiderman" with
Cliqz:
- And the new actor playing Spider-Man is... this guy (www.foxnews.com)
- J. Jonah Jameson (J.K. Simmons) | Spider-Man Films Wiki (spiderman-films.fandom.com)
Not bad for a small German company uh...
And even if it didn't, when I know it's Jonah Jameson, I can search for the actor that played Jonah Jameson. Search engines are not about having "the answer to life the universe and everything" but helping users refine their searches until they find what they were looking for.
But Google fails as well.
Most of the world population is not native english speaker.
Try "attore che recita la parte del giornalista in spiderman" (the actor that plays the journalist in spiderman) in Italian
- Spider-Man: Homecoming - Wikipedia (https://it.wikipedia.org)
- James Franco - Wikipedia (https://it.wikipedia.org)
- Spider-Man film: tutti gli attori | Popcorn Tv (https://popcorntv.it)
- Martin Sheen, dieci ruoli per scoprire un grande attore ... (https://www.consigli.it)
Or the simplified version "attore che fa il giornalista in spiderman" (same meaning as before, just more down to earth)
- J. Jonah Jameson - Wikipedia (https://it.wikipedia.org)
- J. Jonah Jameson - Wikipedia (https://en.wikipedia.org)
- Spider-Man: Far From Home, nel cast anche l'attore ... (https://tg24.sky.it)
Refining is still a very present need, it's simply that for common searches on common topics in english it's less so...
It's more or less the same in German, French, Spanish, Portuguese... I can't even imagine the results in languages like Arabic, Indonesian or Balinese.
Local, culture aware, search engines are the future, despite Google efforts, the generalist "one size fits all" search engine is not.
"Google's dominance is almost entirely due to the fact that its by far the best at search."
is, and I'm sorry to say, a make-believe
Google's quality is better than anyone else, that is a fact. Let's go to the point. [[I work on Cliqz, and in search]]
Do you know how much Google pays apple to be the default search engine?
According to you, nothing, because people will go to Google because it's the best.
Well, it turns out that last year was more than 9 Billion (with a B). Quite a lot of money poorly spend, someone should really get fired :-)
The other option is that Google does not pay to get the Apple users, which might come otherwise, but to prevent Apple to try something funny, either directly or indirectly.
In any case, the "build a better product and people will come" mantra is flawed when companies in the space pay each other billions for distribution.
Instead, Apple has a thing to sell -- "position of default search engine" -- and it is selling it to highest bidder.
I bet if Bing would offer 10 Billion, they'd make Bing default instead. Hey, if DDG could offer them 10 Billion, I am sure they would set it as default, easily ignoring the fact that many people say they do not like its results.
But please focus on Google, which is the one putting the money!
Why Google pays so much money if they already have the best product and what the users want? Are they just giving B10$ as charity?
There are many reasons for doing that, but none of them is aligned with the make-believe statement that "build a better product and people will come"
I would agree that is a necessary condition, but by no means sufficient, which is kind of sad.
Google pays billions to get those 90% of don’t care users. They already have the users who care, but more users = more money, so presumably it is worth it.
Cliqz, DDG, and others have no chance for those 90%, but they fight for the rest. It’s a hard fight because google is so good. If one spends a hour searching for stuff on alternative engine, then goes to google and finds the results in a few minutes, they will be unliketo visit other engines again.
Yeah, and due to economies of scale and network effects it will stay that way, unless some people are willing to suffer the minor inconvenience of using a slightly inferior competitor.
Yes, when what you want happens to be the lowest common denominator - in this case, the most recent Spiderman movie. When the next movie in the franchise comes out in a few years and you don't like it so much because the news guy is now played by someone else, you'll be complaining that your preferences have stayed the same but Google no longer handles them as well.
But is it? I find myself often having to fight it to search for what I entered into the box, rather than what it thinks I really meant. If you know what you want just not where it is, DDG and Bing are both superior.
Maybe it’s denial or weirdness on my side... after all I am also a vegan because I don’t want to hurt animals or other people just for a little bit of taste and convenience and most people seem to think that’s stupid on my side.
I know, Bing.. But I have used Windows for about 30 minutes in the last year and that was just to help my mom fix her printer. I don't think I have ever had a Microsoft related account since Hotmail. I know they track me but I don't use MicrosoftDrive or WintowsTube so I can live with it.
That said, it does't prevent the situation where searching for certain gifs or images gives you a page full of softcore porn, even with the moderate safe search enabled.
You can almost always, by phrasing more carefully, get exactly what you need with DDG.
I also think that it's better to learn to search properly than finding the best search engine.
For instance:
"the actor that plays the news guy in spiderman"
ddg:
Tom Holland (side bar)
Spider-Man Homecoming (imdb)
Tom Holland (wiki)
Spider-Man: Into the Spider-Verse (imdb)
google:
Jonathan Kimble Simmons (splash, link to wiki)
J. Jonah Jameson (wiki)
J. K. Simmons (wiki)
The google result is exactly what I want. And the results were made in incognito mode so Google wasn't able to cheat with privileged information about me as a user.
At the end of the day, most people care about the product. I'm only willing to sacrifice so much to satisfy the ideal that there should be less concentration. Make a better search engine but trying to pull at the heart-strings of users about how Google is an empire and too powerful just won't work and it undermines your product and mission.