I'm not getting those results, but I'm never logged into Google and have their trackers blocked. If you're logged into a Google account when searching, you get "personalized search results" based on at least your previous searches.
For the most effect go to the county library and use one of the Internet computers. Bring up Google and try a few short queries. It reflects the searches of the people who regularly use the library's Internet access computers.
The point of the article (as I understood it) was that perhaps Google could recognize queries that were hateful (it already squashes queries that are sexual) and either suppress them or proactively suggest counter narratives that the searcher might not be aware of. It is the latter which is both quite difficult and usually quite offensive to people, if you just suppress the suggestions it generates less controversy.
I don't trust google with any search results. They business is censorship business and thats my reason to use any other search engine!
Google, No thaaanks!
His spin on how Google doesn't know who exactly uses their services got me fuming. But well, it's an interview from 2008, so before Android, Google+, and so on. So, let's give him benefit of doubt. But then he answers direct question about anonymity of the collected data with this golden quote: “I've never heard of any case of anybody's search getting expose by virtue of Google Search lose”, and it left me speechless.
The results for [are jews evil] are indeed terrible. It's not just the ideology that's wrong, they're also just generically low-quality.
The only non-hateful results was RationalWiki(.com/wiki/Evil_jew), and its effectiveness is blunted by google choosing the quality warning as the only result.
Let's see if my edit sticks and indexing is fast enough to have some impact if thousands of people try the search after reading the article.
Because Google actually censored auto-complete search to be in favor of Clinton while Yahoo, Bing, and Duck Duck Go did not. I've got screenshots I can dig up and post if the HN bregade doesn't downvote me into grey.
It's simpler to explain in economic terms. You get what you incentivize. The google-dominated internet incentivizes page views for money. People are going to post what makes people click - shock, hate, gossip. Posting feel-good me-too stuff does not bring views. So you end up with an internet looking like this. Does it matter? I m not sure, the yellow press was always there and we did not end up with a new hitler every 4 years. Does google matter that much to democracy? I'm not sure its propaganda can be that effective. I think the best kind of propaganda is economy: give people the economic rewards they asked for, and they 'll quit being racist.
Another post-Trump FUD piece about tech companies. Urgent call to "fix" something that has been broken for many years and no one in the media minded one bit.
Firstly, I don't get many of the mentioned autocomplete results, so the very premise of the author (that Google actively peddles this stuff to neutral, well-meaning users) is made-up. Go into private window mode, try it for yourself. No, really.
Secondly, if you do search for "are X evil" you will get what you asked for: conspiracy websites and low-grade propaganda. This is not restricted to the few Xs cherry-picked in the article. If you query is shaped for getting biased results, why shouldn't you? That's the whole point of search engines.
This is not to say that Google is doing a great job at filtering out SEO-d garbage. It's been getting worse in this regard for the last decade. But the article's "analysis" is awful. It jumps around, it anchors, it uses name-dropping. (Ironically, they name-drop China, which indeed "solved" the problem with bad search results.)
Furthermore, the article goes on and and on, but never does it mention the little fact that Larry, Sergey and Mark are all themselves Jewish, AFAIK. I was expecting to see it mentioned at least once to highlight the inherent conundrum. It's possible that I might have missed it because, well, the article never seems to end.
I bet if Trump and Brexit lost the internet would be totally fine.
These articles are only to drum up support for FB and Google applying a Guardian-like bias. It's not an article about all the ways Google is biased, it's an article specifically about the instances in which google is biased against their POV.
Seriously, if you can't convince people that jews aren't evil, just give up your position as an editorial writer because you're totally incompetent at your job.
Of course the media's position is not that we make weak and ineffectual arguments, to people who have tuned them out long ago, it's that Google and FB are biased which is why no one believes their POV.
If you're losing a debate it's likely because you made crappy points not that the the debate stage is biased against you.
Jesus fuck Lincoln won an election 150 years ago being against racism, and you can't win against fucking Trump and Farage 150 years later, and you think it's because Google autocompletes, 'jews are' ... evil?
Is your position seriously that Google's bias has made people more hateful than when Lincoln won on a platform of starting a civil fucking war over slavery?
> And 20 years ago it didn’t even exist. When Tony Blair became prime minister, it wasn’t possible to Google him: the search engine had yet to be invented. The company was only founded in 1998 and Facebook didn’t appear until 2004. Google’s founders Sergey Brin and Larry Page are still only 43. Mark Zuckerberg of Facebook is 32. Everything they’ve done, the world they’ve remade, has been done in the blink of an eye.
Facepalm.jpg
Yeah, google's flawed and its results are an interesting sociological resource.
The fact "are Jews evil" is a top auto complete makes sense. I know large portions of the population that think that way. And Google returning an echo chamber result makes sense from a "these links may be of interest to your query" since they are. Google isn't there to police society - it's use for fake news and product pandering is bad enough - making it police ethics will only hide the ugly bits of humanity - which will then go unfixed.
Article is from 2013, and they do routinely update which words they police. They also now police searching for "Name+Crime", I'd guess to avoid lawsuits.
I think Google has opened themselves up to critique about allowing this antisemitism by policing other offensive terms. They need to either set up a forum for policing all offensive terms, or avoid policing. Otherwise articles like the OP will continue to plague them.
If you have a search engine, you need to rank results. The autocomplete feature uses those rankings to determine what is more likely for you to type next. That's all.
21 comments
[ 2.9 ms ] story [ 57.7 ms ] threadThe point of the article (as I understood it) was that perhaps Google could recognize queries that were hateful (it already squashes queries that are sexual) and either suppress them or proactively suggest counter narratives that the searcher might not be aware of. It is the latter which is both quite difficult and usually quite offensive to people, if you just suppress the suggestions it generates less controversy.
The only non-hateful results was RationalWiki(.com/wiki/Evil_jew), and its effectiveness is blunted by google choosing the quality warning as the only result.
Let's see if my edit sticks and indexing is fast enough to have some impact if thousands of people try the search after reading the article.
Because Google actually censored auto-complete search to be in favor of Clinton while Yahoo, Bing, and Duck Duck Go did not. I've got screenshots I can dig up and post if the HN bregade doesn't downvote me into grey.
Firstly, I don't get many of the mentioned autocomplete results, so the very premise of the author (that Google actively peddles this stuff to neutral, well-meaning users) is made-up. Go into private window mode, try it for yourself. No, really.
Secondly, if you do search for "are X evil" you will get what you asked for: conspiracy websites and low-grade propaganda. This is not restricted to the few Xs cherry-picked in the article. If you query is shaped for getting biased results, why shouldn't you? That's the whole point of search engines.
This is not to say that Google is doing a great job at filtering out SEO-d garbage. It's been getting worse in this regard for the last decade. But the article's "analysis" is awful. It jumps around, it anchors, it uses name-dropping. (Ironically, they name-drop China, which indeed "solved" the problem with bad search results.)
These articles are only to drum up support for FB and Google applying a Guardian-like bias. It's not an article about all the ways Google is biased, it's an article specifically about the instances in which google is biased against their POV.
Seriously, if you can't convince people that jews aren't evil, just give up your position as an editorial writer because you're totally incompetent at your job.
Of course the media's position is not that we make weak and ineffectual arguments, to people who have tuned them out long ago, it's that Google and FB are biased which is why no one believes their POV.
If you're losing a debate it's likely because you made crappy points not that the the debate stage is biased against you.
Jesus fuck Lincoln won an election 150 years ago being against racism, and you can't win against fucking Trump and Farage 150 years later, and you think it's because Google autocompletes, 'jews are' ... evil?
Is your position seriously that Google's bias has made people more hateful than when Lincoln won on a platform of starting a civil fucking war over slavery?
well....are you evil?
Facepalm.jpg
Yeah, google's flawed and its results are an interesting sociological resource.
The fact "are Jews evil" is a top auto complete makes sense. I know large portions of the population that think that way. And Google returning an echo chamber result makes sense from a "these links may be of interest to your query" since they are. Google isn't there to police society - it's use for fake news and product pandering is bad enough - making it police ethics will only hide the ugly bits of humanity - which will then go unfixed.
http://www.slate.com/articles/technology/future_tense/2013/0...
Article is from 2013, and they do routinely update which words they police. They also now police searching for "Name+Crime", I'd guess to avoid lawsuits.
http://www.politifact.com/punditfact/statements/2016/jun/23/...
I think Google has opened themselves up to critique about allowing this antisemitism by policing other offensive terms. They need to either set up a forum for policing all offensive terms, or avoid policing. Otherwise articles like the OP will continue to plague them.
talk about digging your own PC grave.