The Whole YouTube Radicalizes People Story Doesn't Seem To Have Much Evidence To Back It Up

from the myths,-myths-all-around-me dept

There seem to be a lot of "myths" about big internet companies that don't stand up to that much scrutiny, even as they're often accepted as common knowledge. There's the idea that Facebook's algorithm remains in place only because it makes Facebook more money (Facebook's own internal research suggests otherwise), or that disinformation goes viral on social media first (a detailed study showed cable news is a much bigger vector of virality).

Another big one is that YouTube "radicalizes" people via its algorithm. There are lots of stories about how someone went on to YouTube to watch, like, video game clips, and within a week had become an alt-right edge lord troll shouting Trump slogans or whatever. Hell, this was a key plot point in the Social Dilemma, in which the young boy in the fictionalized sitcom family starts watching some videos on his phone, and a week later is participating in an extremist political rally that turns into a riot.

However, a very thorough recent study (first highlighted by Ars Technica) found that there's really not much evidence to support any of this narrative. From the abstract:

Recently, YouTube’s scale has fueled concerns that YouTube users are being radicalized via a combination of biased recommendations and ostensibly apolitical “anti-woke” channels, both of which have been claimed to direct attention to radical political content. Here we test this hypothesis using a representative panel of more than 300,000 Americans and their individual-level browsing behavior, on and off YouTube, from January 2016 through December 2019. Using a labeled set of political news channels, we find that news consumption on YouTube is dominated by mainstream and largely centrist sources. Consumers of far-right content, while more engaged than average, represent a small and stable percentage of news consumers. However, consumption of “anti-woke” content, defined in terms of its opposition to progressive intellectual and political agendas, grew steadily in popularity and is correlated with consumption of far-right content off-platform. We find no evidence that engagement with far-right content is caused by YouTube recommendations systematically, nor do we find clear evidence that anti-woke channels serve as a gateway to the far right. Rather, consumption of political content on YouTube appears to reflect individual preferences that extend across the web as a whole.

Of course, this isn't the first study to find the same thing. A similar study that was released last year came to the same basic conclusion:

In conclusion, our study shows that one cannot proclaim that YouTube’s algorithm, at the current state, is leading users towards more radical content. There is clearly plenty of content on YouTube that one might view as radicalizing or inflammatory. However, the responsibility of that content is with the content creator and the consumers themselves. Shifting the responsibility for radicalization from users and content creators to YouTube is not supported by our data.

A study from two years ago... also found the same thing:

In short, the best quantitative evidence available demonstrates that any “radicalization” that occurs on YouTube happens according to the standard model of persuasion: people adopt new beliefs about the world by combining their prior beliefs with new information (Guess and Coppock, 2018). People select information about topics that interest them; if political, they prefer information that is at least some what congenial to their prior beliefs (Stroud, 2017). Persuasion happens at the margins when it does happen.

Indeed, that study showed that the classic story of someone watching a Trump-leaning "alt-lite" video and getting sucked down into alt-right extremism doesn't seem likely to happen that often.

A random walk algorithm beginning at an Alt-Lite video and taking 5 steps randomly selecting one of the ten recommended videos will only be recommended a video from the Alt-Right approximately one out every 1,700 trips. For a random walker beginning at a “control” video from the mainstream media, the probability is so small that it is difficult to see on the graph, but it is certainly no more common than one out of every 10,000 trips.

And, not that you would necessarily trust research coming directly from YouTube itself, but the company recently released some information on this question as well. Contrary to the "common knowledge" the company hasn't seen "more engagement" on extremist content:

Actually, through surveys and feedback, we’ve found that most viewers do not want to be recommended borderline content, and many find it upsetting and off-putting. In fact, when we demoted salacious or tabloid-type content we saw that watchtime actually increased by 0.5% percent over the course of 2.5 months, relative to when we didn’t place any limits.

Also, we haven’t seen evidence that borderline content is on average more engaging than other types of content. Consider content from flat earthers. While there are far more videos uploaded that say the Earth is flat than those that say it’s round, on average, flat earth videos get far fewer views. Surveys show that borderline content is satisfying to only a very small portion of viewers on YouTube. We’ve invested significant time and money toward making sure it doesn’t find its way to broader audiences through our recommendations system. Today, borderline content gets most of its views from sources other than non-subscribed recommendations.

Now, I will note that YouTube recently changed that final line. I found it last month when Evelyn Douek tweeted about it -- but she highlighted that final sentence which said something different at the time:

In the version she tweeted, the final line reads: "Today, borderline content gets more of its views from other platforms that link to YouTube." But, sometime after she tweeted that, YouTube changed it to read: "Today, borderline content gets most of its views from sources other than non-subscribed recommendations." That's... a little different. And at the very least, it makes me wonder why the change -- and at least highlights that there may be a difference between "subscribed recommendations" and "non-subscribed recommendations."

Still, there are questions about how people are finding YouTube videos -- and they do come from all over. As Douek highlights in a follow up tweet, the largest number of external views on Facebook are to YouTube. This doesn't absolve YouTube and its algorithm, but once again highlights how (1) all of this is a lot more complicated than people make it out to be, (2) the internet is an interconnected ecosystem, not just a few giant sites, and (3) the common wisdom you may have heard... might not be supported by the data.

Filed Under: algorithms, myth debunking, outside links, radicalization
Companies: youtube

New Study Suggests That YouTube's Recommendation Algorithm Isn't The Tool Of Radicalization Many People Believe (At Least Not Any More)

from the well,-look-at-that dept

It's become almost "common knowledge" that various social media recommendation engines "lead to radicalization." Just recently in giving a talk to telecom execs, I was told, point blank, that social media was clearly evil and clearly driving people into radicalization because "that's how you sell more ads" and that nothing I could say could convince them otherwise. Thankfully, though, there's a new study that throws some cold water on those claims, by showing that YouTube's algorithm -- at least in late 2019 -- appears to be doing the opposite.

To the contrary, these data suggest that YouTube's recommendation algorithm actively discourages viewers from visiting radicalizing or extremist content. Instead, the algorithm is shown to favor mainstream media and cable news content over independent YouTube channels....

Indeed, as you read through the report, it suggests that YouTube's algorithm if it has any bias at all, it's one towards bland centrism.

The recommendations algorithm advantages several groups to a significant extent. For example, we can see that when one watches a video that belongs to the Partisan Left category, the algorithm will present an estimated 3.4M impressions to the Center/Left MSM category more than it does the other way. On the contrary, we can see that the channels that suffer the most substantial disadvantages are again channels that fall outside mainstream media. Both right-wing and left-wing YouTuber channels are disadvantaged, with White Identitarian and Conspiracy channels being the least advantaged by the algorithm. For viewers of conspiracy channel videos, there are 5.5 million more recommendations to Partisan Right videos than vice versa.

We should also note that right-wing videos are not the only disadvantaged groups. Channels discussing topics such as social justice or socialist view are disadvantaged by the recommendations algorithm as well. The common feature of disadvantages channels is that their content creators are seldom broadcasting networks or mainstream journals. These channels are independent content creators.

Basically, YouTube is pushing people towards mainstream media sources. Whether or not you think that's a good thing is up to you. But at the very least, it doesn't appear to default to extremism as many people note. Of course, that doesn't mean that it's that way for everyone. Indeed, there are some people criticizing this study because it only studies non-logged in user recommendations. Nor does it mean that it wasn't like that in the past. This study was done recently, and it's been said that YouTube has been trying to adjust its algorithms quite a bit over the past few years in response to some of these criticisms.

However, this actually highlights some key points. Given enough public outcry, the big social media platforms have taken claims of "promoting extremism" seriously, and have taken efforts to deal with it (though, I'll also make a side prediction that some aggrieved conspiracy theorists will try to use this as evidence of "anti-conservative bias" despite it not showing that at all). Companies are still figuring much of this stuff out and insisting that because of some anecdotes of radicalization that it must always be so, is obviously jumping the gun quite a bit.

In a separate Medium blog post by one of the authors of the paper, Mark Ledwich, it's noted that the "these algorithms are radicalizing everyone" narrative also is grossly insulting to people's ability to think for themselves:

Penn State political scientists Joseph Philips and Kevin Munger describe this as the “Zombie Bite” model of YouTube radicalization, which treats users who watch radical content as “infected,” and that this infection spreads. As they see it, the only reason this theory has any weight is that “it implies an obvious policy solution, one which is flattering to the journalists and academics studying the phenomenon.” Rather than look for faults in the algorithm, Philips and Munger propose a “supply and demand” model of YouTube radicalization. If there is a demand for radical right-wing or left-wing content, the demand will be met with supply, regardless of what the algorithm suggests. YouTube, with its low barrier to entry and reliance on video, provides radical political communities with the perfect platform to meet a pre-existing demand.

Writers in old media frequently misrepresent YouTube’s algorithm and fail to acknowledge that recommendations are only one of many factors determining what people watch and how they wrestle with the new information they consume.

Is it true that some people may have had their views changed over time by watching a bunch of gradually more extreme videos? Sure. How many people did that actually happen to? We have little evidence to show that it's a lot. And, now, there is some real evidence suggesting that YouTube is less and less likely to push people in that direction if they're among those who might be susceptible to such a thing in the first place.

For what it's worth, the authors of the study have also created an interesting site, Recfluence.net where you can explore the recommendation path of various types of YouTube videos.

Filed Under: algorithms, engagement, radicalization, recommendation algorithm, recommendations
Companies: youtube

UK Goes Full Orwell: Government To Take Children Away From Parents If They Might Become Radicalized

from the ministry-of-love dept

If there are two edicts I try to follow whenever I'm writing, they are, first, write what is true and, second, avoid cliche at all costs. I bring that up only as a preface before saying the following: the UK is walking down an Orwellian path. It's nearly the cliche of cliches to say something like this, and yet it happens that the cliche is true. While there is most certainly a real thing known as a threat from Islamic terrorism, there is also such a thing as overreaction. What started as the British government's attempt to ban extremist thought from social media and television (under the notion that some thoughts are too dangerous to enjoy the freedom that other thoughts deserve) then devolved into the conscripting of teachers that were to be on the lookout for children that might become radicalized. To assist them with this, the government helpfully provided spy-software to use against students. Spy-software which itself was found to be exploitable in the most laughably easy of ways. This employed two of the most horrifying aspects of Orwell's Oceania: the concept of thought-crime and the employ of citizens to fearfully surveil one another.

And now it seems the UK is going even further, adopting Oceania's reputation for the swallowing up of citizens should they be found suspect of thought-crime by those watchful citizens. Specifically, the Family Division of the Judiciary has put out a memo declaring exactly how it will remove children from the homes of anyone it suspects might radicalize those children. Here's a snippet.

Recent months have seen increasing numbers of children cases coming before the Family Division and the Family Court where there are allegations or suspicions: that children, with their parents or on their own, are planning or attempting or being groomed with a view to travel to parts of Syria controlled by the so-called Islamic State; that children have been or are at risk of being radicalised; or that children have been or at are at risk of being involved in terrorist activities either in this country or abroad.

Only a local authority can start care proceedings (see section 31(1) of the Children Act 1989 – the police powers are set out in section 46). However, any person with a proper interest in the welfare of a child can start proceedings under the inherent jurisdiction or apply to make a child a ward of court.2 Usually, in cases falling within the description in paragraph 1 above, it will be the local authority which starts proceedings under the inherent jurisdiction or applies to make a child a ward of court, and the court would not expect the police (who have other priorities and responsibilities) to do so. There is, however, no reason why in a case where it seems to the police to be necessary to do so, the police should not start such proceedings for the purposes, for example, of making a child a ward of court, obtaining an injunction to prevent the child travelling abroad, obtaining a passport order, or obtaining a Tipstaff location or collection order. Given the complexities of these cases, I have decided that, for the time being at least, all cases falling within the description in paragraph 1 above are to be heard by High Court Judges of the Family Division.

In other words, the High Court Judges within the Family Division are now tasked with determining whether children will be made wards of the state based solely on suspicions of possible radicalization. Children torn from mothers and fathers in Muslim homes will be subject to the whims and inherently flawed watch of the larger citizenry. A citizenry, mind you, that has had its vigilance unduly ramped up by the government's past actions and requests. It's hard to imagine a better recipe for the unfair targeting of Muslim families than this. Unfortunately for all concerned, this same memo imagined just such a recipe, making things even worse.

Judges hearing cases falling within the description in paragraph 1 above will wish to be alert to: (a) the need to protect the Article 6 rights of all the parties;4 (b) the fact that much of the information gathered by the police and other agencies will not be relevant to the issues before the court; (c) the fact that some of the information gathered by the police and other agencies is highly sensitive and such that its disclosure may damage the public interest or even put lives at risk; (d) the need to avoid inappropriately wide or inadequately defined requests for disclosure of information or documents by the police or other agencies; (e) the need to avoid seeking disclosure from the police or other agencies of information or material which may be subject to PII, or the disclosure of which might compromise ongoing investigations, damage the public interest or put lives at risk, unless the judge is satisfied that such disclosure is “necessary to enable the court to resolve the proceedings justly” within the meaning given to those words when used in, for example, sections 32(5) and 38(7A) of the Children Act 1989 and section 13(6) of the Children and Families Act 2014; (f) the need to safeguard the custody of, and in appropriate cases limit access to, any sensitive materials provided to the court5 by the police or other agencies;6...

It goes on from there, essentially giving courts and law enforcement an absolute free pass to deny the court open access and review of the very intelligence that landed the case before it in the first place. This is a memo designed to create a court system by which Muslim parents will lose their children and won't even be told why, or have the opportunity to rebut evidence against them, as no evidence need be presented. This isn't just overreaction, it's terrifyingly provocative action designed with one target in mind and built on the back of a process designed to be flawed in favor of a government that apparently can't get its head on straight.

Nobody means to suggest that there is no threat that the UK might face from Islamic terrorists and/or extremists. But you simply don't adopt the tactics of Orwell to combat that threat. Not if you want to claim your own people remain free, that is.

Filed Under: children, courts, families, family services, radicalization, thought police, uk