YouTube’s Product Chief on Online Radicalization and Algorithmic Rabbit Holes
nytimes.com
nytimes.com
* Let me turn it off entirely.
* Let me adjust the relative weight that the current video gets vs my past views (i.e. do I want to see related videos to this one, or ones that YouTube generally thinks I'll click on)
* Let me adjust the weighting of thumbs up vs. watch time.
* Let me configure the homepage
* Let me make the entire recommender panel be just subscriptions or items from a particular list
* Let me replace the recommender entirely with something of my own devising, called back through a webhook
These things would make YouTube much more useful for me. I'm not going to YouTube just to kill time, and I pay them $13/mo. They're not getting any ad revenue from me, so why am I stuck with their recommender that only cares what it thinks will cause me to spend the most time on YouTube. I am not interested in spending the most time on YouTube. I'm interested in getting the most out of it in the minimum time.
... give users so much control over their experience is counter to having algorithms that suck them into spending more time.
their optimization makes more sense when you consider what metrics generates them money.
the one which I only partially understand is Netflix, but I think it's because of the show pricing hidden variable.
At a certain level of volume, more transparency should be LEGALLY required for multiple companies (auto personalization / recommendation engines / etc.). At least the ability to opt-out. Just hard to define. Could you legally define “bubbling” or super duper smart recommendation algorithms?
Big tech knows this is the direction IMO.
It’s my guess why they randomly started pressing so hard for screen time usage reminders.
They’re selling addiction to kids (and adults).
YouTube, FB, Instagram, Twitter Moments all have liability when it starts to click with everyone.
It’s a serious issue that is totally unregulated and not understood by most.
To play the devil's advocate, isn't technological "addiction" the inevitable future, so why even fight it? The more engaging technology becomes, the more people will naturally(and perhaps rightfully) choose it over reality. Isn't trying to curb the addiction merely an attempt to postpone the future?
In the UI I eschew every traditional user manipulation technique I could think of: there are no notifications, there are no unread message counts. There even isn’t read/unread message state. It's launching now, so finally I get to find out: can a platform that doesn't vie for attention, that lets users create their own unguided, unprompted experience succeed?
[0]: https://www.pony.gg
- Weekly/Fortnightly/monthly delivery. I want a way to say to an old friend "hey, lets keep up with each other every fortnight.
- Offline mode. I want to be able to take 5 letters to a park near my house, turn off wifi, and just write with a clear head.
> - Weekly/Fortnightly/monthly delivery. I want a way to say to an old friend "hey, lets keep up with each other every fortnight.
This is probably the most requested feature right now. I want this too but decided I had to keep it as simple as possible for the launch. But this will definitely be fleshed out as there becomes room to evolve the concept. Turning those delivery time radio buttons into check boxes will be exciting. I, for one, would like a delivery mode that syncs up with sunrise/midday/sunset at your geolocation. This will be fun to explore.
> - Offline mode. I want to be able to take 5 letters to a park near my house, turn off wifi, and just write with a clear head.
This is already in the works. I want this too. I even have a ServiceWorker installed for another reason (https://stackoverflow.com/a/55245528/741970) but will leverage it for proper offline functionality as well.
Suppose we want to do a movie recommendation system, and we have a bunch of data consisting of anonymous user identifiers and for each user a list of movies they have seen and their rating of that movie.
Imagine that we had a list of movie attributes that we thought might be relevant to whether or not someone would like the movie. This might be things like running time, if it is funny, if it has car chases, if it has romance, if there are horses in it, if there is profanity, and so on. Let's say we've got 50 of these attributes.
Now suppose we had for each movie a vector of length 50 whose components were how well the movie fit each of those attributes, on a -1 to 1 scale.
If we had that movie attribute data, then what we could try is modeling the users as if each user had a 50 component vector telling how important each attribute is for that user, and for each user we could go over the list of movie ratings and try to find a set of weights for the components of that user's vector such that the dot product of their vector with a given movie's attribute vector correlates with that user's rating.
That works pretty well, but there are two practical problems. First, we need to come up with that attribute list for the movies. Second, once we've got our attributes someone has to go through each movie and figure out the vector.
So forget about that approach for a moment. Suppose instead we came up with a list of attributes, somehow, but instead of figuring them out for the movies and then inferring the user weights, suppose we told the users the attributes, and asked them how important each was, and assumed that the users are actually right about what is important to them?
Then maybe we could take all the movies, and try to assign attribute weights to them that lead to consistent predictions of user scores!
It turns out that works, but we still have the problem of guessing what attributes matter, and depend on the users actually somewhat knowing what makes them like a movie.
So...if we knew the movie attribute weights we could infer the user preference weights, and if we knew the user preference weights we could infer the movie attribute weights.
The brilliant solution is to get rid of the attribute labels. All we decide on is the number of attributes! So we might decide that there are going to be 50 attributes, and we can assign the movies random weights for all those attributes. We can assign the users random preferences for the attributes.
Then you can go through an iterative processes where you tune the movie attribute weights to better predict user scores, and you tune the user preferences to better match the movies. This ends up converging to a set of movie attribute weights and user preference weights that do a good job of reproducing the user's scores for the movies, and makes good recommendations...
...and you have no idea when it is done what the attributes mean!
None of this is exactly deliberate, it's just that the recommender will figure out that if you can get someone angry about something, or entranced by some conspiracy, they're going to be very engaged.
The problem is, going to a social media site with a specific intention, it's an attentional battle. They are going to show you the most entrancing thing they can come up with above the fold. For example, on youtube.com, you can choose to some degree what sections are shown, but the unconfigurable recommender is always at the top and cannot be moved. "Saved for later" is always a couple of screenfuls down no matter what you do.
This is hostile technology. YouTube is not made to be a tool for you to use. It's made so that you're a tool for it to use. It pokes and prods at you with its recommendations to serve its own interests. It does not care at all about yours as long as it doesn't piss you off so much that you stop going there.
I can think of two reasons.
1. If YT recommends good content to people, and make them watch YT more, they are more likely to keep paying the monthly subscription fee. So I'm not sure that the recommender is completely useless for their subscription business. I do agree that for subscribers, the weight of engagement metrics should probably be lowered in favor of user satisfaction or some other metrics.
2. YouTube is an extremely complex system developed over many years, and it's very expensive and error-prone to try and change anything. For many years, it was monetized through ads; so it was designed around that goal. The monthly subscription is relatively new, and it's still probably a tiny fraction of the ad revenues, so many features of YouTube are just left untouched even when they don't make much sense for monthly subscribers.
I do agree that 1-2 or the stuff mentioned above would be nice, but most of it comes off as a programmer trying to tweak the hell out of a system, ignoring that 99.99% of users actually won't benefit from ANY of that.
The search on youtube is also completely broken, I have not monetised any of my videos and the penalties for that makes it so that if I search for the exact title + my channel name the video will show up on something like page 5.
Also, complex != confusing. You can have very simple UIs that are confusing as hell if you're not used to them or you need a specific issue. I usually have trouble helping my mother on her iphone because I'm not used to it and it has absolutely no guides or help functions if you can't find what you need. Usually it's a "magical gesture" that is documented on some obscure web page.
Spotify is just as bad. I once made the mistake of liking an instrumental track on one of the radio stations I listen to. Now a good 25-30% of the songs that come up are just the instrumental versions, not the songs themselves. No matter how many of them I dislike they keep appearing. The same goes for remixes.
There's no notion of "nuance" in these systems and it just makes me think that they were only ever evaluated internally. The models might perform well when they're tested on all 1 billion YouTube users, but on an individual level they're really awful.
As an aside, I've never given them any of my data so this is purely anecdotal, but I frequently hear (in both positive and negative contexts) about how great Facebook's recommendation systems are on both Facebook and Instagram. I think that's what investor confidence in them as an ad company has stayed so high even through their controversies. I honestly don't think I've ever heard anyone praise Google's recommendation systems, by contrast.
Well, that is all that really matters to them in the end, isn't it? A more accurate recommendation system might be better for many individuals but not drive engagement, whereas what they are going after may be lousy for the individual but drive more engagement overall. It's the classic playing to the lowest common denominator. Getting enough people to keep viewing ads is all that matters.
Moreover, Youtube became increasingly resistant to user configuration. And the discussion doesn't touch on user configuration. I'd say the reason is that there's been uniform switch across multiple platform to auto-recommender model. Why? Part may be bureaucracies assuming they need control, part is a "the average user is a total moron who can't configure and will just accept defaults anyway" consensus (not entirely false but not entirely true), part of it is the rise of machine learning makes having supposedly strong recommendation engine be equated with having value.
And the situation with a completely unconfigurable recommendation is that they can't show people what they want. They can only show people "what people can be conditioned to like, what grabs some average more." Which is going to be polarizing content.
- Let me configure the order of posts based on voting patterns (hot, newest, controversial)
- Let me adjust the weighted score of posts based on the age of the account.
- Let me configure the home page.
> Let me replace the recommender entirely with something of my own devising, called back through a webhook.
Is there any popular consumer product that allows custom recommendation based on a webhook? Opening up a core part of your product to an untrusted and probably unreliable third-party is not a good idea.
> why am I stuck with their recommender that only cares what it thinks will cause me to spend the most time on YouTube.
This is addressed by the article:
> The first is this notion that it’s somehow in our interests for the recommendations to shift people in this direction because it boosts watch time or what have you. I can say categorically that’s not the way that our recommendation systems are designed.
This is actually better in business sense, since it saves so much bandwidth:
A. Watch N videos that you like or are neutral about. B. Watch N/4 videos that you really like.
The latter is decreased server load and less bandwidth wasted on content that users may not like at all.
Youtube is currently wasting Terabytes of bandwidth on such silly ideas as autoplaying somewhat relevant videos in a chain(similar to TV).
There is no additional cost to Google for playing additional videos. They have interconnects and they aren't full, so there is no reason not to add more content.
It's only the last mile that has bandwidth constraints really.
Occasionally I would try dismissing one of the suggestions in the hope I wouldn’t see similar again, but then I’d still see it’s kind repeated and was just disappointed it wasn’t taking my request into account.
Suggestions are hard though. No problem.
One day, though, I made the mistake of following a link to YouTube which turned out to be to a video of an American political figure (not even sure if he is a politician) whose views were of the sort that court controversy - on purpose, I guessed.
What I didn’t expect was that YouTube would then offer me videos featuring the same person - or people with similar ultra-polarising messages - every time I returned to YouTube.
I hadn’t realised that not only does YT take into account what you have watched a _lot_ of, it also seems to massively favour what you have watched once - if that is something especially controversial or perhaps otherwise ‘special’.
While I don’t mind pointless suggestions I will ignore, it’s very concerning that without ‘liking’ or repeatedly watching such material, people are being suggested it time after time, even when they ask to have it taken away.
This is my experience as well. YT thinks I love Jordan Peterson. Four or so months ago, I didn't know who he was, and clicked one clip, found it ridiculous and never clicked another one. But my recommendations are lousy with the guy.
This is kind of my nightmare! Condolences.
YT, wtf.
Imagine this translated to print media, imagine YouPaper: a publisher of magazines. Going by these same engagement metric driven algorithmic based systems such a platform would rapidly spiral toward pushing tabloid content to the detriment of everything else.
> I’m trying to describe to you the nature of the problem. So what I’m saying is that when a video is watched, you will see a number of videos that are then recommended. Some of those videos might have the perception of skewing in one direction or, you know, call it more extreme. There are other videos that skew in the opposite direction.
I assisted with data research on YouTube recommended videos (https://www.buzzfeednews.com/article/carolineodonovan/down-y...), and this claim is misleading at best. There may be some videos that aren't extreme, but that doesn't matter if they are a) in the extreme minority and/or b) appear very far down in the queue of 20+ recommended videos when users typically either do "Up Next" or look at/action only the top few recommendations.
This isn't a "both sides" issue.
https://www.washingtonpost.com/opinions/the-southern-poverty...
This is simplistic, but say all videos are on a 1-10 integer scale in terms of politics. YouTube always recommends 80% videos that are on the same value as the currently viewed video and 10% are one value higher and 10% are on one value lower. It doesn't push anyone in either direction. Randomly clicking on recommended videos will randomly move you up or down (and usually not move at all). But, no matter where you begin, you can always reach either end in 4 clicks - cue people talking about the YouTube rabbit hole.
But hey, maybe it's not enough to merely prefer content that isn't extreme, what if we wanted to actively push people to the center? We could recommend 80% videos that are one value higher or lower towards the center spectrum (or the same is the viewed video is at 5 or 6), and 10% videos that are at the same value, and 10% towards one value away from the center. Randomly clicking on videos _will_ bring most viewers to the center in this system.
But we can still reach the extremes in just 4 clicks. This is why complaining that users can go down a rabbit hole feels disingenuous to me. They always will be _able_ to do so - unless YouTube never recommends videos that aren't more centrist than the video being viewed.
I suppose YouTube could just ban all videos that are at 1 and 10 in our spectrum, but that's basically just calling for censorship. And not to mention, 2 and 9 will just become the new definition of extreme.
Edit: many of the comment lead me to think that I haven't made what is and isn't in scope of my comment clear. Allow me to point out two things in particular.
1. YouTube already does prohibit hate speech. How well it enforces its rules is a valid point of discussion, but not what I'm talking about here. I'm pointing out that the ability to get to the extremes of what _is_ allowed on the platform through recommendations should not in and of itself be a point of concern.
2. I'm also aware that sometimes the recommendation are very messed up in context of the current video - e.g. shock videos getting recommended to kids. But this is more about trolls deliberately gaming the recommendation algorithm, which I see as a distinct issue from the alleged rabbit hole pattern of recommendations.
There's no law that says they have to recommend videos. There is however sufficient evidence at this time around the rabbit-hole effect causing real-world harm. This isn't 2005. It's about time Youtube, and social media properties adopt the medical principle: "first, do no harm". Perhaps they will do this willingly, but somehow I doubt it.
To wit: https://www.bellingcat.com/news/americas/2018/10/11/memes-in...
As far as "do no harm" goes, implementing safeguard against the alleged rabbit hole has a very strong potential to do harm by making YouTube a partisan platform. Bear in mind that concern over the alleged rabbit hole is not equal between liberals, centrists, and conservatives. Many of the recommended videos I've seen people reference as evidence of the alleged rabbit hole are videos I'd considered mainstream. What it takes to placate the critics has very strong potential to violate the principle of "first, do no harm".
Regarding kids, this has been very well documented: https://www.ted.com/talks/james_bridle_the_nightmare_videos_...
https://medium.com/@jamesbridle/something-is-wrong-on-the-in...
The point I am making is that pointing out the fact that it's possible to get to the extremes of YouTube by clicking through on recommended videos shouldn't be surprising. In fact, of you couldn't so that it would be evidence of deliberately trying to keep certain content from being popular.
The question of how well YouTube polices it's terms of service (which already does prohibit hate speech) is related to this question about the alleged rabbit hole because such content presumably exists at the edge of my hypothetical political spectrum values. But it's not directly related to the fact that clicking through on recommended links can bring people to 1 and 10 even if they start at 5 should not be surprising.
Allowing monetization (where much of YouTube's controversies have been) is a completely different story since it also involves the demands of advertisers. The advertisers do have a right to avoid paying money to extremist content, just like you have a right not to donate to causes you don't believe in.
> first, do no harm
Easier said than done.
Challenge one is to define "harm" without classifying all sorts of alternative political ideas as "harmful".
Challenge two is once you've defined "harm", build a system that works at YouTube scale that can actually filter it. Also consider that it's a moving target - people are able to figure out how to get around YouTube's filtering algorithm, so YouTube needs to constantly update it to keep ahead of these people.
Sure, defining harm is not straightforward. But it's possible, and frankly as a private platform, it is their choice.
One could, for example, take a broader definition of harm, and use that to limit recommendations, but still host the video and allow it to return in search results. Is this the right answer? I don't know. But it's silly to just throw up your hands and somehow take recommendations as some sacrosanct part of the system, when it was clearly a choice to have them to begin with.
And the crux seems to be that YouTube has a single primary interest in generating suggestions: to offer the viewer a video that they probably want to see next. Netflix recs are no different.
So, the impolite or gruesome content becomes clickbait for an ad revenue machine.
I think it's simpler than what you're suggesting. They want to make money, and the indirect way they do that is to get people to watch more videos. If most people who watch Jordan Peterson tend to also watch extreme right wing content, and you watch his video, then you will get recommended right wing content, because they think that's what you will want to watch.
I work for Google but not on Youtube, opinions are my own.
From my understanding, JP is actually pretty connected with conservative/right wing politics. Not so much that he is necessarily right wing himself, I wouldn't know, but that he has a large right wing following.
I think they like him for his opposition to political correctness/gender pronouns. His philosophy on masculinity also really appeals to conservatives rebelling against feminism.
See: https://en.wikipedia.org/wiki/Jordan_Peterson#Academia_and_p... https://en.wikipedia.org/wiki/Jordan_Peterson#Gender_relatio...
And almost anyone who likes any alt-right video also likes JP, because he provides a veil of respectability for anti-feminist, anti-multicultural views.
This theme has been explored in quite many articles on news and advertisement. It is not that extreme stories get people to click more on ads, but it keep people engaged for longer and looking at more content. More views, more engagement with the publishing site, and more ad impressions.
I recall reading that there exist an interesting link between moral outrage and dopamine highs. It basically becomes a drug which publishers see as increased user engagement.
I generally agree, but maybe an important detail orbits around the "can" (of This is why complaining that users can go down a rabbit hole...): are Youtube recommendations a healthy mix that give users a choice or does the majority of recommendations push the user down the rabbit hole?
I'm actually thinking of making a script or an extension that automates, say 10, 25, or 50 random clicks in the top 10 recommendations and seeing where it ends up. Maybe this system exists already.
I have to admit that the exact definition of the problem is not easy to define:
e.g. if I have on my screen a video about "people-that-hate-bananas", would it be ok to have recommendations that show only e.g. "why-bananas-are-bad" or should I (ethically) get as well videos of the type "are-bananas-really-bad?" respectively "these-are-good-bananas!"?
Using the above example I guess that if Youtube (& similar sites) show only recommendations based on case #1 (which is eaaasy for IT) then yes here I go down the rabbit hole (reinforces/confirms the concept "bananas-are-bad") even if the recommendations are not worse than the original one => therefore even recommendations on "equal" level might have to be evaluated at least partially negatively? (as in this example the recommendations don't promote alternative point-of-views / alternative thinking / search for a flow or compromise, if you understand what I mean)
In any case the evaluation of the results you get might be very hard to do, if you want to do it automatically to understand the meaning (e.g. very hard to understand if "good-bananas-are-bad" is negative or positive towards bananas)... :P (maybe if you manage to you might become a Nobel candidate, hehe, or maybe I'm underestimating the current power of AI, which is definitely possible)
It makes sense to show more Lego content on the sidebar of a Lego set review. It's less clear that I should be chased by Lego videos across the site after I clicked on a single one of them.
Also, considering that various extremes tend to attract attention, it would make sense to use an "anti-viral" algorithm that de-preoritizes videos that get too many views at the moment. This would have an additional effect of giving better exposure to less popular channels, promoting different content creators.
Realistically, I don't think they will do either, even if they know for sure that it will help, because it will reduce addict... I mean, engagement of their user-base.
Jaron Lanier talked about these issues a lot recently. He makes some interesting points regarding incentives and feedback loops.
Key findings:
* Centrist content links to liberal content at over twice the rate it links to conservative content.
* Proportionally more conservative content links to centrist or liberal content than either liberal or centrist content links to conservative content.
* Liberal content links to itself at the highest proportional rate.
Granted, who knows what portion of liberal or conservative content counts as extreme. But concerns over YouTube extremism almost always has to do with conservative extremism and the data does not suggest conservative content is being promoted by YouTube.
When I say "extreme" I mean "extreme" in all senses, not just politically extremist. Could be just an extremely one-sided view of something ("why X sucks") or even a "top 10" or "worst fail" video.
I don't consider your opinion invalid, but I do get the sense that the majority of concern over the YouTube "rabbit hole" is about promoting politically extreme content. My comments, since the beginning, have been specific to the allegation of pushing users to the extremes politically. I tried had to be explicit about this, and even edited my comment (prior to your comment, mind you) to more explicitly spell out the fact that I'm not talking about extreme things that aren't political (e.g. the "Elsagate" issues of shock content getting recommended to kids).
If this is the issue you're taking, then bear in mind that it is not in scope of what I am pointing out, and I don't think it's on scope for the majority of people's discussions about the "YouTube rabbit hole".
If Youtube truly pulls you towards extreme videos, then you would always end up at those videos after a long run.
> ...It’s equally — depending on a user’s behavior — likely that you could have started on a more extreme video and actually moved in the other direction.
> That’s what our research showed when we were looking at this more closely..
Now this is also a bit of a weasely claim. First because of "depending on user's behaviour" which isn't very clear (but for now I'll give the benefit of the doubt and assume they are trying to say "unless the user is actively seeking radical content"). Second because this claim could be true and YT would still be radicalizing (i.e. it could go 1->3, 3->2, 2->4, 4->3, 3->5, 5->4 in terms of extremeism giving a gradual slope towards radical while still maintaing that you are equally likely to navigate to more extreme content as you are to navigate to less extreme content).
OTOH, if I understand the article you linked to correctly, all it says it that if you keep clicking "Up Next" you'll eventually get from mainstream content to extreme content, which doesn't invalidate the above YT claim if their research shows that users aren't actually doing that. Given that you assisted with that research maybe you could make it more clear what you tested and how it shows the "rabbit hole" effect?
In a lot of these experiments though, people explicitly pick the more extreme of the recommendations repeatedly, which of course will lead you down the rabbit hole.
I just did an experiment where I clicked one of these and clicked "next" a bunch going through their recommended videos. It immediately dove into alt-right stuff, and eventually led to videos of antifa protestors getting beaten up.
This is normal on YouTube.
I'm not sure if Ben Shapiro is being directly funded but it looks like PragerU is mostly funded either directly by billionairs or by think tanks (which are often also funded by billionaires but I don't have time to dig into all of the different ones mentioned here)
I looked up the hip-hop artist Mac Miller when he died of a drug overdose and the all of the recommended videos were of his pop star ex-girlfriend being the one responsible for his death. Clearly the engine has learned to go towards the extreme.
My opinion is to push for education in all aspects of these algorithms and keep your kids off Youtube! It is so far from network or cable TV it's a little scary. That said, it's an amazing resource but it's geared toward growth and profitability at any cost (most likely lives).
Might be the case that Chrome users in your area are going Incognito to watch Ben Shapiro videos. Given your reaction, that might be exactly the case.
Or it might be Ben is a B-level celebrity that might attract a click to display an expensive ad.
Maybe it's true people matching some of my info watch are likely to watch Shapiro videos, but by optimizing for engagement they basically are trying to herd similar people towards whatever version of that person watches the most YouTube videos.
Indirectly YouTube wants everyone to be a right-wing fanatic, because the algorithm has determined that that sort of person watches a lot of videos.
YouTube shows videos that are likely to get clicks, but fanatics are more likely to click and watch videos. So things fanatics care about are promoted. YouTube seems to take this stance that it's just natural social dynamics at play, excusing themselves from addressing what they're enabling.
If car modding videos led you down a rabbit hole towards "watch how fast I can go if I cut the brake lines with these $3 wire cutters! Those braking requirements are just a conspiracy between big auto and the shadow government to keep you from reaching car enlightenment, so make sure to drive through the front window of your nearest car dealership"... and then people started doing that, we would have a different attitude about them.
Why is that different than politics?
For me the line-in-the-sand is political views that advocate destroying our entire way of life, especially through violence.
In my view you don’t get to claim the benefits of free speech and tolerance if your manifesto is based on intolerance and marginalisation. If you’re intolerant, don’t expect to be tolerated.
How is this not a clear admission they are a publisher and not just a platform?
Now this does work in the other direction, where someone with more normal viewing habits watching a video will steer the conspiracy theorists toward the types of videos they watch, but the size of the pools means it's the polarizing direction that would have a greater effect.
But what I'd point out is that I see YouTube respecting _all_ the content I see, rather than just the most recent one or just right-wing content, as some commenters here are alleging. For example my feed usually includes videos from Vox as well as Fox on politics, alongside video game reviews, alongside a NASA stream, alongside woodworking, and so on.
Never seen it happen as strong with any other topic.
I don't know about increasing levels of extreme, but I can definitely see it possible that suggestions could get people to a place where they are feeling like an idea or a thing is more popular than it really is.
He talks about it briefly here (nsfw)
My beef with all this is our current Overton window has been compressed to right of center economically.
American opinion and general support for solid, center left economic policy is high and growing.
Medicare For All, and friends.
Our mainstream media simply does not report on, nor offer much opinion favorable to labor, and or framed from a labor point of view.
It used to. When I was a kid, we would compare the various points of view, learn to identify them, and discuss what having the means to society.
This lack of control is a good thing. I think we've got problems it's a struggle with, but I don't think they're as bad as what's being represented, nor do I think a nicely centered window makes any sense. I think the people should decide that.
That difference is what has people worried because the political and social endgame is not yet clear.
I also don't doubt that you can (sometimes) get to really extreme content within four clicks but we have to keep in mind that the recommendation lists can be 40 videos long. So just roughly estimating it: if 1% of videos was extreme content because 1% of the population are extremist you would expect, on average, to see roughly one and a half extreme video if you sampled 4 * 40 random videos.
In addition, just doing a quick test with Jordan Peterson in incognito mode I get the following very telling result.
If I click a video called "Jordan Peterson EDUCATES College Professors In An Epic Q & A", I get mostly other Jordan Peterson videos with some Ben Shapiro sprinkled in.
If I click on "Jordan Peterson Destroys Q&A | 25 February 2019" I suddenly get a wide range of people suggested with the common thread being that the title also contains "DESTROY", "SNAP", "OWNS" and so on. This makes me think the 'problem' might just be that people who keep clicking on the most outrageous titles will eventually see the most outrageous videos. From a technical perspective I would say that is working as intended.
By sheer quantity mainstream media leans center left and online indie media leans center right. You could say that taking the whole landscape into account that constitutes some kind of tenuous balance overall - but it would be nice if both mediums were more balanced.
Interesting aside: conservative talk radio has always been more popular than liberal talk radio - is this just the online expression of the same phenomenon? People on one side of the aisle tending to enjoy listening to and watching political content at greater length than their counterparts?
Advertisers, who got it, would support the station. We had a couple who paid well, and got results.
Dirty secret in talk radio: most owners are conservative. Deffo bias there. In my market, the progressive talk station got next to nothing from the owners. Rightie talk got tons of cross promotion.
Where a leftie talk station was not owned by a major cluster, it also would do just fine.
In mainstream media, there is a basic conflict of interest. Flat out, big business does not air progressives, because progressives would cost them and regulate them.
Online, the majors are making a push to get relevant with younger people, who now call them legacy media. (Gotta hate that if you are CNN)
Access journalism, actually a basic fail to do journalism, for fear of losing access, coupled with the same big business bias, means we do not produce economic left content in major media in the US.
Some of this "extreme" discussion centers on the fact that big business simply does not want that on the air, or in people's feeds.
Progressives struggle with this constantly.
Labor just is not in most big platform interests, and many younger people simply ignore the major media for this reason.
I, at a much older age, get pushed to watch legacy media online every day, multiple times per day.
I do not want to watch them. I want more about ordinary people, labor and the political movements seeking better.
The important thing for Lefty's, who are interested in resolving this problem, is to realize they need to support those voices. And we need to do it rather directly, because of the big systems, and the large money, really won't do it.
Can anyone say that they have distinctly not encountered this effect? I can recall having watched maybe a couple of Joe Rogan clips, a Jordan Peterson interview, and even some stuff I would label alt-right content. I'm personally fairly left-ish/liberal, so I find these videos boring/offensive and eventually will go back to my normal youtube consumption - music, sports, some tech videos.
My recommendations are all of the latter categories, not of extreme/political right wing category. I guess due to selection bias, most people who don't have a problem with their recommendations won't report anything, while most people who do will leave comments that they too experienced the rabbit hole effect. I'm wondering if that leads to the problem described being overstated. Or am I really the only person who's managed to watch some fairly extreme right wing content, and had my recommended videos stay intact?
I built Pony [0], an email platform that delivers once a day, to see if this is possible. In the UI I eschewed every traditional user manipulation technique I could think of: there are no notifications, there are no unread message counts. There even isn’t read/unread message state.
I truly wonder whether people can adapt to a totally unstructured online platform, an unguided, unprompted experience that they create for themselves. My bet is they can.
[0]: https://www.pony.gg
The postal service would like a word with you.
Turn youtube into a type of podcast db, its just a media storage site. Then you could create your own portal, parse the views/ratings and provide real statistics.
I don't watch youtube for youtube, I watch youtube for user-created content. And some of that content has already moved off onto 3rd party sites. Floatplane anyone? Patreon anyone? list goes on.
I just wish finding video content was as easy as searching for podcasts. Youtube provides RSS links, I use them to add to my podcast player.
However, if I watch a video from a conservative angle, even just one or two videos, I almost immediately get extreme right wing content in my recommendations. Stuff like PragerU, ReasonTV, NRATV, etc. Even watching videos that I wouldn't consider right wing, just critical of certain left wing sects, like h3h3 for example, tend to almost immediately lead me into videos like "DUMB FEMINISTS GET OWNED - COMPILATION".
It's strange the rabbit holes almost always take me deep into right wing territory, but never really into left wing territory.
As for your experience with recommendations - I am not sure why you would not see recommendations similar to the videos you like. I see videos from multiple ideologies regularly and get content recommended from all of them. I haven't had my feed dominated by one side. What I think is happening is that people _notice_ videos that incense them, and incorrectly perceive it as being dominant when it truly isn't.
https://www.youtube.com/watch?v=EM7BgrddY18
"Why you should love fossil fuels", "It is easy for feminists to forget that men gave women the right to vote, gave up their monopoly on power, and invented birth control." -- even-keeled?
> it is easy to find flaws with a wide range of content producers on the left as well.
More importantly irrevelant, since the criticism isn't that imperfect right wing videos are presented instead of perfect left wing ones.
> I haven't had my feed dominated by one side.
So? It's perfectly possible for you to be in a different testing bucket. Don't assume what you see is what everybody sees, and that they are just interpreting it differently than you do. I might as well say "nah, you only think you see videos from multiple ideologies".
https://www.youtube.com/watch?v=GK72U02TUsI
I was getting JP and right-leaning politics recommended to me for a month afterward, even though I never use Youtube to watch JP or political content of any kind. But maybe I was getting that content recommended because I had no prior indications of interest in that conceptual space. The recommender could be exploring previously untouched territory in hopes of finding something with high exploitation value.
A value-blind system trying to maximize engagement is going to find that some niche content performs well on some users, and will try to explore that space for other users to see if it has high exploitation value for them too. So start out watching videos about health and it's natural -- from a data perspective -- to surface some conceptually adjacent videos that include anti-vaccination conspiracies. If you watch one or two of those -- even without saving or liking -- then the exploration may expand into conceptually adjacent videos about other Things "They" Don't Want You To Know, and so on.
The Youtube recommender works pretty well for discovering new music, in my experience. The problem is that the process that gradually takes you from a Top 40 hit to a less popular musician to somebody who recorded their first song last month isn't always benign when "music" is replaced by "health" or "world news."
[1] https://towardsdatascience.com/reinforcement-learning-demyst...
For example I clicked a few months ago in a link in HN to a video of a moron that claims that can cure type I diabetes with a diet that avoid certain kind of food [1]. Now I get from time to time recommendations to see the videos of this moron and it's totally annoying, because it's clearly wrong.
I also get recommendations for music I don't like, food channels I don't like, news channel that I don't like, channels that are somewhat related to what I like but I don't like, ... Perhaps some new video appears in the recommendations for a few days. Anyway, they are not annoying enough to be remembered.
[1] It's not sugar, or HFCS or other food additive that is somewhat related to diabetes, or can cause type II diabetes. It just doesn't make sense.
If this is a ploy to push more mainstream narratives through YouTube that is akin to watching CNN/CBS or ABC, then I'll be looking for other platforms.
I'll keep you company here.
>videos are very engaging, fun and you learn something from it.
This is not an either/or situation. Extremist content can be engaging, fun, and appear to be informative.
However, JRE hosted Alex Jones (conspiracy theorist "mass shooting victoms are crisis actors"), Gavin McInnes (white supremacist), and Milo yiannopolous (conservative provocateur who among other things supports pedophilia in the gay community.
Joe himself I can't comment on. He chooses to promote controversial and actively extreme people though. Sometimes he disagrees with those people, sometimes he doesn't.
One is about extremist content. Extremist content can be interesting. Calling content interesting does not excuse its extremist nature.
There is a parallel conversation specifically about whether JRE is "extremist". The show is certainly interesting. It may also be extremist, but it being interesting doesn't make it not extremist. Because, as I previously stated, calling content interesting does not excuse its extremist nature.
I am not making any conclusions about JRE. I'm only saying that his show being interesting doesn't make it "not extremist". If you're trying to prove that he's not, you've got to do something else.
I also provided some things he has had on his show that probably are extremist content.
>Joe Rogan went on Infowars today, the same day a man who lost his child in the Sandy Hook shooting committed suicide. The grieving father was among the plaintiffs suing Alex Jones for defamation after smearing the murdered children and their parents as part of a hoax.
>In the same Infowars episode that Joe Rogan appeared on, Alex Jones pushed a conspiracy theory that the father of a child killed at Sandy Hook didn't kill himself, but may have been murdered to silence Jones and end the first amendment.
That Twitter thread is interesting, because it contains a link to a 'This American Life' program that gives a platform to Alex Jones to explain himself. I doubt anyone would call 'This American Life' a platform for spreading conspiracy theories or hate.
In this light, is Joe Rogan, who also allows people like Jones to explain themselves, a moderate or an extremist?
Alex Jones is a wacko. Talking to Alex Jones does not necessarily mean you are also a wacko. This new concept of moral badness transitively propagating through conversation is nonsensical and counterproductive.
This is no myth bro, go to youtube pick a political video leave on autoplay, watch it go to shit.
-----------
1. “of course we take this seriously”
- Mohan: "Having said that, we do take this notion of dissemination of harmful misinformation, hate-filled content, content that in some cases is inciting violence, extremely seriously."
- Zuckerberg 2016: "we take misinformation seriously. We’ve been working on this problem for a long time and we take this responsibility seriously."
2. “we’re simply trying to help people get accurate information”
- Mohan: "when users are looking for information, YouTube is putting its best foot forward in terms of serving that information to them."
- Zuckerberg 2016: “Our goal is to connect people with the stories they find most meaningful, and we know people want accurate information.”
3. “fear not, it’s our users who are in power/in control”
- Mohan: "But YouTube is also still keeping users in power, in terms of their intent and the information that they’re looking for."
- Zuckerberg 2004: “People have very good control over who can see their information.”
- Zuckerberg 2017: “Our full mission statement is: give people the power to build community and bring the world closer together. That reflects that we can’t do this ourselves, but only by empowering people to build communities and bring people together.”
And last but definitely not least,
4. “we’re proud of our progress, but there’s more work to do”
- Mohan: "It’s an ongoing effort. I think we’ve made great strides here. But clearly there’s more work to be done."
- Zuckerberg 2016: “We’ve made significant progress, but there is more work to be done.”
- Zuckerberg 2018: “I’ve learned a lot from focusing on these issues and we still have a lot of work ahead…I’m proud of the progress we’ve made in 2018… I’m committed to continuing to make progress on these important issues as we enter the new year.”
Why anyone would copy Zuckerberg’s script at this point is beyond me.
-----------
Sources:
- Zuck 2004: https://www.thecrimson.com/article/2004/2/9/hundreds-registe...
- Zuck 2016: https://www.facebook.com/zuck/posts/10103269806149061
- Zuck 2017: https://www.facebook.com/notes/mark-zuckerberg/bringing-the-...
- Zuck 2018: https://www.facebook.com/4/posts/10105865715850211/
I use YouTube, I don't browse it. I search for what I want, and then I watch it. Yet YouTube seems absolutely determined to send me down rabbit holes for some reason. For instance my friend sent me a video by some cable news neo-philosopher (a Jordan Peterson type) and since then I can't get rid of low-quality vids in my recommendations that are trying to make me angry. "{Person} totally EVISCERATES {other person}" being the calmest of them.
The recommendation algorithm is also just not even good at its job. I am Jewish and the number of videos I have seen recommended to me that are thinly-veiled anti-semitism is pathetic. Do they really think I am going to watch those?
The one thing from the article I do believe is that these extremist videos (mostly) don't monetize. So why do they dominate? Honestly I think we just suck at building recommendation systems.
I am not a big fan of Facebook, but I applaud their recent ban of white supremacist content. Freedom of speech is one thing, freedom to use someone else's site to speak to millions is another thing. YouTube should just take a harder stance on all of this dangerous crap. The Sandy Hook deniers and the Antivaxxers will have a hard time finding a new home.
</rant>
I pay for YouTube to avoid ads and wish I didn't have to use their home view which shows me a lot of content I don't care about. I want my homescreen to be my subscriptions (instead I have to manually switch to this every time).
I would entirely turn off recommended content if I could, in favor of channels suggesting other channels or something more human curated (like Spotify discovery which in part uses other user's playlists).
People are vulnerable to bad information delivered in a manipulative way - especially when this is reinforced by a bunch of recommended videos.
You wouldn't say that a drunk who walked into a nice restaurant yelling hateful things was being censored when they are asked to leave. Neither would you say that a man trying to convince kids to get into his van outside of McDonalds was being censored for being kicked out of the McDonalds. Both of those are private businesses regulating the behavior of their customers, and not in any way censorship.
> Randomly clicking on videos
The entire point is that people don't click on random videos, they click on videos they think are interesting. Call it "rage clicks", curiosity, or some people are just bad people and want to see bad things, just don't call it random, because thats obviously wrong.
Nowhere. It always puzzles me why this response always crops up, even when I never even remotely try to claim that platforms have any legal obligation to host content.
I did claim that making their site have a partisan recommendation algorithm is not a good idea and may be bad for society. I have zero doubt that this is their right, but people and companies have the right to do plenty of things that aren't good for society. Saying that it's bad for society to implement a partisan recommendation algorithm is not at all the same thing as saying it is or should be illegal.
My theory of HN comments: People prefer to respond to familar arguments, rather than the arguments you make in your comments. People prefer to discuss topics that they are familiar or passionate about, rather than the specific topics addressed in the articles or parent comments.
Posting Fast and Slow
So who gets to decide what is good and bad for society?
It's entirely valid to say that tech companies' censorship or forms of manipulating views is bad while acknowledging that this is their legal right. This comment applies just as much to you: https://news.ycombinator.com/item?id=19525998
Everyone does. Everyone makes their own decisions about what is right or wrong.
This is setting aside issues with bots, trolls, and brigades. In real life with people you know and can talk to there can be value in these discussions, but never anonymously and never online. It’s pseudointellectual wank dressed up as reasonable talk. There is nothing wrong with simply saying, “I’m not interested in having this conversation with you, sorry.” and moving on.
That was a pretty normal position to hold on the US until about 5 years ago and I’d say it still is.
Basically you have people that demand others be censored. That they don't deserve to speak freely. And they'll come up with all these mental gymnastics as to why other people's speech isn't acceptable. And for most of us we understand that the first amendment is the most important and that's not acceptable.
Just look thru the last 2 years of left leaning websites. Anyone who claimed trump was innocent (clearly...) was a "bot, troll or brigader" and ought to be censored because it was "hate speech".
A world where we censor is a dark world. There's a reason free speech was right number one in this country.
This gets tricky though when you're talking about things like being blocked by a domain name registrar or something.
Even with speech you publish yourself there are some existing restrictions (libel laws, hate speech targeting individuals, etc.)
Moderation is also what allows communities like HN to exist and remain interesting places - I'd definitely prefer it to every website turning into 4chan.
Furthermore, deplatforming has always been a thing. It basically happened to the dixie chicks after they spoke out against the Iraq war and GWB in England way back in the early 2000's. I don't think they have ever been on fox or a sinclair broadcasting station ever again.
No one is saying proclaiming trumps innocence is hate speech, that is an absurd claim, where are you getting that nonsense?
The concern is that people are able to click through to the extremes of what _is_ allowed on YouTube. My point is that this is not in and of itself a valuable thing to be concerned about. Even if YouTube actively tweaked their algorithms to push people towards the center, it'd still be possible to click through to the extremes - unless YouTube totally stops all recommendations that don't lead closer to the center.
Its not like YouTube is bound by some cosmic constant that makes it hard to police the content on their platform. They are Alphabet for god sakes, they have so much money they don't know what to do with it. They can, and should, solve these problems, and we can, and should, hold them to account until they do.
Copyright filtering isn't a good example IMO. Lots of people consistently claim that the content ID algorithms are crappy and have high rates of false positives.
It's easy to do when it's the same files or data trying to be uploaded, or when there's a way to consistently identify it as prohibited. It's hard when people are actively trying to evade the filtering algorithms. Here is one easy example: a lot of copyrighted work like to shows can be found on YouTube where the copyrighted video is displayed on a smaller section of the screen (e.g. a rectangle in a corner with some animated stuff around the rest of the screen). This is because such editing makes it harder to detect. Maybe YouTube has beaten this countermeasure by now - it's a cat and mouse dynamic by nature.
To be fair, this is just my perspective. If someone with greater authority on this subject can give me technical explanation as to why Facebook's 80% preemptively blocked rate is not good I'm happy to hear it. But from what I know, an 80% preemptive block rate on an hours-old video with a significant number of people actively trying to avoid the filtering logic is an impressive feat.
This is an incredibly hard problem space. Machine learning is already hard before you start introducing adversaries...
I read an article the other day about the Australian response to the New Zealand shooting, which had this choice gem:
'He said it remained online for 69 minutes. "That is a totally unreasonable period of time and represents a complete failure of Facebook's own systems," he said.' [0]
That was a very senior government minister and they are about to start legislating on the subject. That has to be one of the more breathtaking expectations out there - this is asking a media company to judge what the community standard is and deploy widespread containment measures in under 60 minutes. That is comparable to the response time of our emergency services (~10 minutes) in a life-or-death situation.
I can't really grasp what it is people think that Facebook is doing wrong here that requires a response of that precision; accidentally watching a video simply isn't comparable to having a heart attack or being on fire. I was hopeful that the response time would be something out of the ordinary and measured in hours or days, I'm doubly impressed that Facebook managed to respond in 70 minutes and surprised at how well developed filtering mechanism to control communication on their platform.
[0] https://www.abc.net.au/news/2019-03-29/social-media-executiv...
You can simply reduce you complaint from ‘hate speech’ and ‘violent speech’ to ‘speech’ to reveal the nuanced perspective of grey thinkers.
Dismissing the societal effect of censorship by dominant public mediums because their profit structure is private is - regardless of your position on said censorship - a very incurious and unserious to adopt.
The contrived examples to support your position are unlikely to resonate with anyone who spends the briefest moment in consideration of all the game theoretic outcomes that are possible with your attitude.
I presume in your world it is okay for telecommunication companies to deny cell phone access to individuals based on arbitrary, internal and opaque reasons because they’re privately held.
Attempting to deny the reality of societal responsibilities when a service hits societal scale is just silly.
Indeed, especially when the primary purpose of the service is to be a platform for speech.
Hotels don't get to violate peoples rights just because they are private businesses.
Hm. I see some obvious holes in that line of thought.
I feel that the answer to “why is this advertising company so much more interested in selling ads than promoting free speech?” is more obvious than some give credit for, as is the answer to “why isn’t this massive for profit advertiser not on board with my narrow absolutist view on speech?!”
We believe that everyone deserves to have a voice, and that the world is a better place when we listen, share and build community through our stories.
Our values are based on four essential freedoms that define who we are.
Freedom of Expression
We believe people should be able to speak freely, share opinions, foster open dialogue, and that creative freedom leads to new voices, formats and possibilities.
Freedom of Information
We believe everyone should have easy, open access to information and that video is a powerful force for education, building understanding, and documenting world events, big and small.
Freedom of Opportunity
We believe everyone should have a chance to be discovered, build a business and succeed on their own terms, and that people—not gatekeepers—decide what’s popular.
Freedom to Belong
We believe everyone should be able to find communities of support, break down barriers, transcend borders and come together around shared interests and passions.
I don't get out of my obligations as a public accommodation just because my hotel makes money by also having billboards on the property.
I don't take this position seriously anymore, because I've seen time and time again that people who espouse it immediately flip 180 degrees when some large platform starts censoring content they want to post or see.
The opposite is also true. Some people who speak against censorship flip 180 degrees when they don't like some content. (Though a bit less often, it seems.)
My conclusion: before saying anything of this sort, you have to be able to present a clear general rule (that other people can apply without asking you) that distinguishes between cases of censorship you're okay with from other cases. That's the only way such conversation can start in a rational manner.
Realistically, very few people truly believe that big companies should be home free to delete anything they want, solely their own discretion (including content from human right activists, opposition parties, corporate whistle-blowers, scientists and critics of the platform in question).
Of course they are. If they start showing stuff that scares off advertisers, they will be out of business. Don't see a lot of porn on youtube, do you?
If you think that platforms like youtube, twitter, and facebook should be forced to show all legal content, I think you will find yourself to be in a very small minority position.
And it's a position there is no way the US Supreme Court will back you up on.
Censorship does not require it being done by the government or be morally or legally wrong.
Let me know if this is free speech:
- I have an extreme view I want to share
- I tried sharing on facebook but it got banned
- So I hosted it on my own website but then Cloudflare and AWS dropped me
- So then I hosted it on my own physical box in my house but then Comcast disconnected my service
Do I still have free speech?
> I tried sharing on facebook but it got banned
You still have free speech.
> So I hosted it on my own website but then Cloudflare and AWS dropped me
You still have free speech
> So then I hosted it on my own physical box in my house but then Comcast disconnected my service
Now you don't. Is that legal? If it is, we should consider regulating ISPs in the manner we do phone companies. Someone call Ajit pai.
Which was subsidized, made scarce, and then sold to Comcast by the government. There's a reason people are calling for broadband providers to be considered common carriers, because the market isn't "free" (as in competitive).
Money quote:
> The State attempted to analogize the town's rights to the rights of homeowners to regulate the conduct of guests in their home. The Court rejected that contention, noting that ownership "does not always mean absolute dominion." The court pointed out that the more an owner opens his property up to the public in general, the more his rights are circumscribed by the statutory and constitutional rights of those who are invited in.
Another phrase I've heard recently is along the lines of "a company's ability to refuse to do business with you should be proportional to your ability to refuse to do business with them".
It's not too much of a stretch to liken some of the big platforms and service providers we have today to company towns. While, strictly speaking, one doesn't have to (live in a company town)|(subscribe to Comcast), for some large segments of the population who (live in economically disadvantaged regions where the only jobs are available via the Company)|(are forced to choose between dealing with Comcast and cutting themselves off from social and economic activity), that's not much of a choice in practice.
For the moment, I think that line of "inescapable enough to warrant restriction of property rights" falls to include infra providers (think cartel ISPs, some PAAS, DNS, maybe CloudFlare if you squint) but not platforms like YT/FB. That's probably down to my bias towards technical solutions over regulatory intervention - there's at least a reasonable hope of displacing YT/FB with decentralized platforms, but short of mesh radio you're SOL if your only ISP that offers >56k doesn't want you.
What exactly is "their own platform"?
If I post to a blog site, that's not my platform and can be censored.
If I host my own blogging software on a kubernetes service that's not my platform and I can be censured (AFAIK all providers have TOS on the type of content allowed and largely reserve the right to remove content for any reason).
If I host my own OS with blogging software on a IaaS provider again I may be shutdown/refused to do business with.
If I host my own hardware, with own OS with blogging software in a colocation datacenter it may also be removed for TOS violation.
Same for simply hosting it on my own computer connected to an ISP, the ISP has TOS for the types of content allowed through their networks.
So how exactly do I setup this (mythical?) "own platform" where I can post anything that I am constitutionally/legally allowed to and how do I make it available over the Internet?
So, in theory (at least in my perfect world), you could run a server in your basement and host your controversial views there.
This seems like a bad analogy to me. No loud drunks are forcing anyone to listen to them on YouTube. Yet some people are angry that YouTube lets loud drunks post videos of themselves yelling hateful things, and angry that YouTube lets some people who want to listen that do so?
I think we're at a point (or fast approaching it) where these huge platforms are so integrated into our society that perhaps we _should_ consider them public platforms.
>You wouldn't say that a drunk who walked into a nice restaurant yelling hateful things was being censored when they are asked to leave.
I don't think this analogy holds up because of the sheer magnitude of difference in size. These platforms are bigger than most countries. Countries have governments and laws that serve the citizens and keep society running. Imo, the users of these platforms should be treated like citizens of that platform.
I don't want to be a citizen of a dictatorship.
People are ultimately choosing what to watch.
It's a clear conflict of interest for traditional media to blame their younger competitors for all the world's problems.