Google to start debunking

[Mackdog]

I ran across this today, and apparently Google is thinking about changing the way it filters search results, to more of a debunking type of system. I do not know if this will happen or not, but according to the article, they will be using information from web sites such as Snopes and Factcheck.org to determine if a website will get bumped up or down in the search results.

The software works by tapping into the Knowledge Vault, the vast store of facts that Google has pulled off the internet. Facts the web unanimously agrees on are considered a reasonable proxy for truth. Web pages that contain contradictory information are bumped down the rankings.

Makes me wonder if maybe they will start using Metabunk as a source for whether or not a claim or theory should be considered credible. It could happen.

http://www.newscientist.com/article...es-based-on-facts-not-links.html#.VPMUznysX-v

 

[Svartbjørn]

This is gonna go one of two ways... Ppl are gonna scream that google is now being controlled by "them" OR the sheer amount of bunk out there is going to overwhelm the system and known bunk is going to get bumped up because there are so many references all agreeing on the same thing. Will be useful for the rest of the world that knows how to think critically though.

 

[occams rusty scissor]

This is gonna go one of two ways... Ppl are gonna scream that google is now being controlled by "them" OR the sheer amount of bunk out there is going to overwhelm the system and known bunk is going to get bumped up because there are so many references all agreeing on the same thing.

Yes..with people assuming "well it got through the Google debunker so it MUST be true!"

 

[NoParty]

I LOVE the idea (what debunk-fanboy wouldn't welcome better filtering?) but

A) I think devising such an algorithm, that actually works, could be a nightmare, &

B) I wish it wasn't Google. I'm hardly the conspiracy theory type, but damn, I'm already
creeped out by the unnaturally long arms of this seemingly omnipresent information leviathan.
(every time I turn around, Google has expanded their presence...I no longer have any grasp of what all they control)

 

[Mick West]

Makes me wonder if maybe they will start using Metabunk as a source for whether or not a claim or theory should be considered credible. It could happen.

Metabunk is a lot smaller than Snopes, about 5% of the traffic.
http://www.similarweb.com/website/metabunk.org?competitors=snopes.com

We've got a number of very high quality debunks, but don't have the name-recognition that Snopes has/ Plus they have a lot more pages, and are a lot older.

The thing is, this is going to be an automated algorithm, and this means it's increasingly important to keep up the signal-to-noise ratio - which is why Rambles, Open Discussion, and Chit-chat are hidden from Google.

 

[deirdre]

Google is a joke. Google is youtube. Not only do they invade everyone's privacy by posting your backyards online (now you can case the best house to break into from your basement), but they knowingly leave videos up that they KNOW are illegal. They take down videos with the copyright/harrassment policies but when people clone the videos and put them back up with the EXACT same names, they make you go through the whole legal process again.

1. Google itself can't be trusted.
2. Algorhythms cant be trusted when the data input is coming from "the public".
3. Censorship is ridiculous and hypocritical. Computers making decisions for us is a very bad idea. We need teachers and parents to start teaching critical thinking skills. Band-aids are never a good solution for any problem.

 

[Mick West]

Censorship is ridiculous and hypocritical. Computers making decisions for us is a very bad idea. We need teachers and parents to start teaching critical thinking skills. Band-aids are never a good solution for any problem.

Even if we started now with very high quality critical thinking instruction in all schools:

A) Lots of people have finished with school
B) Many kids don't do well in school
C) It will take several years before the current crop of kids graduate

Band-aids are not ideal when you need stitches. But it's generally better than nothing. Apply it carefully, watch it.

Anyway, the Google Algorithm is going to happen, so I'm going to figure out how to work with it.

 

[deirdre]

Even if we started now with very high quality critical thinking instruction in all schools:

A) Lots of people have finished with school
B) Many kids don't do well in school
C) It will take several years before the current crop of kids graduate

Band-aids are not ideal when you need stitches. But it's generally better than nothing. Apply it carefully, watch it.

Anyway, the Google Algorithm is going to happen, so I'm going to figure out how to work with it.

true but google alone isnt going to stop the spread of bunk. esp when google generates most of it on youtube. : )

and doing well in school really is a different animal than critical thinking. lots of young adults i know who dropped out, flunked out are very level headed. it wouldnt be like a specific course you get graded in. its not what you teach, its HOW you teach. but yes, i'm talking long term goals.

 

[Mick West]

true but google alone isnt going to stop the spread of bunk. esp when google generates most of it on youtube. : )

But I also get half of my traffic from Google. If Google start ranking sites higher if they are more accurate, then that should theoretically be a good thing for Metabunk. Of course there's also the risk that we get flagged for having bunk, as we discuss bunk.

 

[Mick West]

The "knowledge base" they talk about is the database that gives you the answers to simple questions - which they are constantly expanding. If you ask Google a question, it now tries to see if there's an answer, rather than just looking for keyword matches.

For example, search for "are chemtrails real", and you get:

At least here the more reasonable gawker article is presented as the answer. But Natural News comes in above Contrailscience, and there the Kucinich article is first, which is not what people are looking for.

So there's certainly room for imporovement

 

[deirdre]

The "knowledge base" they talk about is the database that gives you the answers to simple questions - which they are constantly expanding. If you ask Google a question, it now tries to see if there's an answer, rather than just looking for keyword matches.

ah. thats how i found your answer to the bunnykin thing so fast. i forget what i typed in now as the question. metabunk came up first!

 

[qed]

Of course *some* people might even claim the CIA is behind Google and that Google is behind the NSA.

Don't be evil.

[... Strike that. Just Googled it and it's bunk...]

 

[Gridlock]

The thing is, this is going to be an automated algorithm, and this means it's increasingly important to keep up the signal-to-noise ratio - which is why Rambles, Open Discussion, and Chit-chat are hidden from Google.

Just an FYI, they aren't hidden from Google per se, it will still be taking the content of these pages into account. They are simply withheld from the index (as per your robots meta-tag).

Not wanting Google traffic to these pages is fine of course, but you should be uber-careful about trying to 'manipulate' Google (zero negative connotations intended) or to shape the algorithm's view of your site.

You also have a slightly weird robots.txt, in that it responds with 200(OK) but doesn't appear to exist?

Anyway, enough day-jobbery

 

[Mick West]

Just an FYI, they aren't hidden from Google per se, it will still be taking the content of these pages into account. They are simply withheld from the index (as per your robots meta-tag).

They are hidden, as you need to be logged in to see them. Previously they were just not indexed.

Not wanting Google traffic to these pages is fine of course, but you should be uber-careful about trying to 'manipulate' Google (zero negative connotations intended) or to shape the algorithm's view of your site.

Yeah, but all I'm doing here is try to improve the signal to noise ratio. Improved Google rankings should be a side effect of that.

You also have a slightly weird robots.txt, in that it responds with 200(OK) but doesn't appear to exist?

It exists, it's just zero bytes. I can't remember why I did that. Google does not seem to mind, but I should probably do some Xenforo specific stuff there.

 

[Mick West]

There has been a pretty significant increase in Google activity starting a few weeks ago though:

 

[Gridlock]

^ That's indicative of an algorithm change, IMO, and not in a bad way. Obviously! (Edit - duh, there I go again, that's the bot not the search traffic.) However, are there 25K pages that warrant crawling?

Apologies for not understanding the logged-in part, obvious in hindsight.

Serving 404 for a robots.txt is fine, if you don't want to block any bot from crawling anything.

 

[mrfintoil]

I'm wondering if this new mechanics of Google will impact debunkers means of finding source material to debunk..?

 

[Svartbjørn]

I'm wondering if this new mechanics of Google will impact debunkers means of finding source material to debunk..?

Nope... people are people.. someone, somewhere will pull something out of their ass and make it sound JUST real enough to be plausible.

 

[occams rusty scissor]

Nope... people are people.. someone, somewhere will pull something out of their ass and make it sound JUST real enough to be plausible.

I have been wondering exactly how google will be able to make that work, as in, get around the human factor. I admit I'm no programmer, and making an engine smart enough to debunk, or at least to partially sift bunk, seems like a tall order to me.

On the other, creepier (CT?) side of this, it does kind of sound like Google is trying to decide what form of the "truth" we will be getting.

 

[Svartbjørn]

I have been wondering exactly how google will be able to make that work, as in, get around the human factor. I admit I'm no programmer, and making an engine smart enough to debunk, or at least to partially sift bunk, seems like a tall order to me.

On the other, creepier (CT?) side of this, it does kind of sound like Google is trying to decide what form of the "truth" we will be getting.

Its all math. The algorithms aren't going to care what's what... It's going to look at ratios of what's being discussed and come up with probabilities which we will see on the end users side. The amount of micromanagement involved to filter only specific sites because that's what google wants would require a LOT of processing power. If they just programmed what sites they wanted it would show up in the code and be outed... Google wouldn't want that.. Bad for business. Tho in theory you could pay google to be topped on searches... Like advertisers do now but you'd be labled as having paid.

Kinda like this

Spoiler

Ads

• Learn SQL Online | lynda.com
  www.lynda.com/subscribe-today
  Online video tutorials for SQL. Get Started.
• Learn SQL Server - Get Trained on Microsoft SQL Server.
  www.quickstart.com/sql
  Get Trained on Microsoft SQL Server. Train from Home/Work. Since 1988.
  Microsoft Gold Certified
  
  
  SQL Server Learning Path
  
  Train From Anywhere
  
  
  Money Back Guarantee
  
  
• Enroll in a SQL Course | Phoenix.edu
  phoenix.edu/it
  University of Phoenix®. Hands-On SQL Training with Lab Simulations.
  8200 Roberts Drive, Sandy Springs, GA (678) 731-0555 Directions
  Online Classes
  
  
  Business Degrees
  
  
  Education Degrees
  
  Risk Free Trial
  
  
  Nursing Programs
  
  
  Certificate Programs [/spolier]
  

 

[occams rusty scissor]

Tho in theory you could pay google to be topped on searches... Like advertisers do now but you'd be labled as having paid.

That's in the vein of what I mean - how long before this sort of thing becomes an issue? I know that the maths behind it isn't interested in who gets paid what, but the maths are put there by the same people who have made billions out of it already.

Or maybe I've subconsciously got too tapped into what the CT's will be thinking and it's slowly taking over my psyche...

 

[deirdre]

I have been wondering exactly how google will be able to make that work, as in, get around the human factor. I admit I'm no programmer, and making an engine smart enough to debunk, or at least to partially sift bunk, seems like a tall order to me.

On the other, creepier (CT?) side of this, it does kind of sound like Google is trying to decide what form of the "truth" we will be getting.

Bunk online seems to be a hot topic lately what with FB too claiming too want to get rid of more bunk. Which, i seems to me based on what FB was saying, has more to do with dollar signs than actually caring about bunk. If people are complaining they dont 'trust' Google as a source because it leads to too many false stories, then Google will try to do something about that so they don't lose customers.

But yes, many will see it as nefarious and I also cant really 'get' how a computer can pull that off unless they use manual tags.

saw an interesting new 'study' about the news media not researching stories and/or purposefully spreading bunk. It's crazy long, so i'm going to bypass the no-click policy and not post exerpts. But it is interesting.

Join the Tow Center for Digital Journalism for the launch of Craig Silverman's report Lies, Damn Lies and Viral Content: How news websites spread (and debunk) online rumors, unverified claims, and misinformation.
http://towcenter.org/wp-content/uploads/2015/02/LiesDamnLies_Silverman_TowCenter.pdf