Climate deniers beat Google and topped the page on searches for "climate change"


#1

Originally published at: https://boingboing.net/2017/12/30/liar-liar-world-on-fire.html


#2


#3

SEO, like PR, becomes even more important when the product or service or ideology one is hawking is total rubbish. It’s not like climate change deniers have a dearth of time and resources due to having to do actual science, so it makes sense that this would be the outcome.

Also, “tricking” Google’s algorithm is relatively easy, albeit a time-consuming and soul-draining grind. As I’ve discovered from working with my long-term SEO subcontractor, some of the same dodgy tricks that worked ten years ago to boost a result in the rankings still work now because some loopholes inside that black box were apparently never closed.


#4

Again, why do I use Duck Duck Go?


#5

Which makes you wonder why Google is taking ad placements for things like “climate change”.


#6

Alongside this manipulation of Google’s mysterious algorithms, there’s some really interesting research into how slightly different search terms produce wildly different results. One study looked at the different results returned when people search for information on childhood vaccination. People who were worried that their might be a risk from vaccines (there isn’t) were likely to use search terms that prioritise results from anti vaccination nut jobs that promote myths and other falsehoods:

Ruiz, Jeanette B., and Robert A. Bell. “Understanding Vaccination Resistance: Vaccine Search Term Selection Bias and the Valence of Retrieved Information.” Vaccine 32, no. 44 (October 7, 2014): 5776–80. https://doi.org/10.1016/j.vaccine.2014.08.042.

And I thoroughly recommend a very scary paper from the University of California about how search engine rankings can be used to influence political opinions. It is terrifying:

Epstein, Robert, and Ronald E. Robertson. “The Search Engine Manipulation Effect (SEME) and Its Possible Impact on the Outcomes of Elections.” Proceedings of the National Academy of Sciences 112, no. 33 (August 18, 2015): E4512–21. https://doi.org/10.1073/pnas.1419828112.


#7

How many years ago was it that Google last had any sort “top of the page” for search results?
Google has personalized every single search result query for every browser since 2008.

I know for certain they’ve been doing this to all search results since 2008 which is when they published their guide to search results personalization. This is true even when not signed in to google. They were doing it as far back as 2005 for signed in google users. So “top of the page” search results is something that has only existed in scammer seo snake oil land for a good 8 years.

While Google keeps their secret formula closely guarded and changes it regularly it has been confirmed that search personalization has factored at some point the following: browser cookie, the specific signed in user, user search history, browser search history, browser url box history, gmail content history, google docs content history, location, analytics data, ratings and reviews, etc. There are different formula for mobile and desktop.

Something being in top search results is as variable and subjective as something being above the fold…both are misapplications of concepts from older times that don’t apply.

The closest you can get to non personalized search results in private browsing on a new/wiped computer, location everything shut off, and vpn’ed into somewhere with a huge diverse population.

What the article linked to describes is simply targeted ads, climate change deniers promoted their links and sites for climate change searches. Those placements are clearly marked as ads. This is different than gaming searches and doesn’t require algorithm knowledge, when promoting your links google’s tools allow you to customize the words and tells you percentages and allows you pay more to adjust. The google ad system is intended to work this way, they didn’t beat google, they simply paid them.


#8

The point is that it shouldn’t exist at all.

That, in an information-dependent industry, this should not happen.

That … really doesn’t make Google look any better.

Want your disinformation about whatever to appear near the top—no, sorry—above legitimate search results? Just pay us.


#9

But it felt cold this morning where I happened to be, so the whole thing must be a hoax.


#10

It’s almost certainly going to happen, unless Google somehow becomes a curated search engine-- which Yahoo proved is not scalable. Sure, Google is in the information industry, in which there is both bad information as well as good information.

global warming


#11

Wait what? Why? Google has always relied on Ads for profit, how do you suggest they make money, do you want to pay for searches?

Google has always put the ads at the top and clearly marked them as ads.

This is nothing new, nor is it news, and it is done across every subject you can conceive of. Anyone familiar with google or ad placement will wonder why this is an article at all and it represents a gross misunderstanding of both.

Perhaps a refresher of Google Ads over the last few years will help illustrate how they’ve separated these from search results visually…


#12

If you’re an “industry player” in searches for information, disinformation is poison.

I see no problem with paying for searches. (But I’m fine with my taxes also funding libraries, mind.)

But let’s say that is a problem. Is there any rule which would apply to Google that says they have to run false ads? If not, then this isn’t a problem: don’t accept money from folk who threaten the core validity your industry. Pay someone to vet ads for certain keyword groups.

It doesn’t matter if it’s marked as “this is an ad”—that doesn’t magically grant any user with a skeptical toolbox to assess and accept/reject disinformation—even assuming that ads were all false (and not all of them may be—I can easily conceive of a situation where legitimate groups might pay for ads to do some science outreach).

Have you ever argued with a denialist? Those things Google did to how ads are displayed do not work that way. Either by design or incompetence, the disinformation was still effectively delivered. Via Google.

The point is Google plays a role in being disinformative, not that they didn’t do their best to visually set apart and mark things as ads.

I think it’s already happened. But I don’t know what you mean by “scalable”—are we talking about algorithms? If the algorithms are that broken, don’t use them. Use a curated list, employ experts to maintain it, put it at the top, put it first. If Google must accept money from denialists, their ads shouldn’t appear alongside those who do the actual work on climate science.

If they have to, they can make a specially composed page which the user is sent to once certain search terms are invoked. We need to stop pretending algorithms are objective—they are arbitrary and can apparently be gamed.


#13

if you know the history of search engines and search based ads, you’ll know that they used to be seeded into the search results as results. Google was one of the key players that championed moving away from poisoned search results and keeping ads in their own area and clearly marking them as ads not search results. Poisoned search results is an industry term for a thing that google helped end.

That’s cool, but the industry has always run on ad revenue since its inception, since long before google was a player. I hope they consider offering a pay for an ad free experience at some point.

the core of their industry is to sell ads, providing us with search results provides them with human eyes which is their product. we are not their customers, we are their product. google doesn’t vet any ads for validity of claim, that would be an impossible endless task.

most people who use the internet regularly have learned to reflexively completely ignore the ads. no regular user would be fooled into thinking they are actual results. sure, some people will be taken in by the ads, which is the entire point of ads. people are allowed to advertise for all sorts of things i don’t agree with, that is part of being in a free country with free speech.

but that has much greater issues with equality, bias, and is an impossible task with the nature and size of the internet.

algorithms are not arbitrary nor objective, rather they are crafted with a specific objective and can be evaluated and refined against said objective. Almost all systems are susceptible to being gamed, even the human ones you are proposing. Algorithms are at least a known, quantifiable, examinable, autiable, refinable thing…as well as essential for every single aspect of modern life, none of which could exist without them.


#14

“End”?

This is not entirely wrong, but it’s it is wrong to say we’re not also the customers.

Yes, yes, yes, the laundry or doing the dishes never ends, right? But are you going to go through life only wearing dirty clothes merely because you only see the futility in cleaning things? Or only wash that shirt if enough people point out the stains in public?

Like the price of freedom, vigilance in information stains is likely eternal.

You use that word a lot, but while tech players may have changed online aspects of information, that doesn’t mean their choices and applications are the only means of doing things.

In my experience, expert opinion offers a great deal of value in what information is useful to sift/distribute.

So who told you it was “impossible”?

There’s no reason to pretend something is working when evidence shows the opposite. And if the algorithm, “known, quantifiable, examinable, autiable [uh? audit-able, maybe?], refinable” appears to have failed, then relying on it less makes more sense than defending it.

I admit I don’t see algorithms as magic (or even worthy of the lofty status assigned to them). They’re just sometimes crappy technology/sophisticated technology breakable by some savvy but otherwise unsophisticated folk. Also, I find it weird that you appear to believe issues like “equality and bias”—legit issues—are things which can’t be addressed without resorting to game-able algorithms. Diversity can probably help overcome these issues. But I don’t know, I don’t work for Google, I just have to argue with their results, often by proxy.

Meanwhile, in the real world, the algorithms Google is using to serve disinformation impact policy for dealing with serious issues. Call that “modern” if you like, but if we can’t have experts reliably informing Congress, members of congress and their constituents will turn to other means available.

Maybe you write some algorithms to detect when algorithms are broken, but it seems pretty clear some form of manual intervention is occasionally, absolutely needed, and I tend to trust experts more than I trust tech businessfolk.


#15

Yes, it is a practice that is no longer found in any of the major search engines.

Do we pay for search? nope. we aren’t the customers. the product they are selling is advertising.

Sorry I should have explained why it was an impossible task, it isn’t possible through manual human effort, period, because all the human effort from all the humans that ever existed and ever will exist isn’t enough to accomplish computational tasks regularly performed. While some computational tasks are manually impossible because of their scope to execute some tasks are impossible simple because they require feats beyond human ability. The content of the internet changes faster and is greater then humans could manually index, it is a task of impossible scope.

what? which specific algorithm failed? In which way?

they are just math

https://en.wikipedia.org/wiki/Algorithm

Human systems like you propose are still systems and are more gamable and less auditable. Humans created machines to accomplish these tasks more accurately for a reason. Human systems still use algorithms even if not formally defined. Maybe you don’t understand what an algorithm actually is.

wait what? if “experts” are quoting Ads from search results to congress in isn’t the fault of algorithms, you are maligning math for human idiocy that is completely unrelated to it.


#16

I’m not sure if it’s useful to fight falsehood with falsehood. Accuracy should be favored.


#17

This topic was automatically closed after 5 days. New replies are no longer allowed.