4 ms·

With respect, when I hear FANG employees putting forward scale as a reason why "moderation is a super hard problem" I'm bemused. If this was a hardware or soft

With respect, when I hear FANG employees putting forward scale as a reason why "moderation is a super hard problem" I'm bemused. If this was a hardware or software issue it would be obvious that you're trying to scale systems that do not scale. In technical terms, social "algorithms" are O(k^n), exponential. You're trying to do the impossible, eh? (Really, the argument could be made that FANG et. al. are becoming a new de facto world government. From that POV you're attempting to solve the hardest human problem, eh?) - - - - No one is forcing you to be that big. At the end of the day, scraping away all the bright-eyed rhetoric about "connecting all the people, places, and things for great good", you're doing it for money. Building a moderated social network isn't impossible (maybe) but it's woefully expensive. The existing systems are, arguably, harming people and messing up society, tech companies notoriously have a lot of money already, but are unwilling to prioritize and pay for the moderation that could (arguably) improve things. So... what? I don't have any answers, I just want to point out that from my POV the problem is clearly hubris and basic cheapness.

altgoogler 6y ago
Indeed, the question is what should be done, or how should it have been done differently? Should YT have grown more slowly? FWIW, I don't think money is the limiting factor here. If there was a known good solution, but it took $10B to implement, I think it would be done. You could pay X thousands (more) moderators, but there are drawbacks in a human solution as well. First is the fact since there's human judgement, rules will be applied inconsistently, and probably with some bias. Second is the fact that moderation is a very mentally exhausting effort. You can observe some of the cyclical nature of this balance from the news over the past few years. Article about moderators with PTSD? Well, automation is the solution. Automation fails in a lot of edge cases? Well, why not hire armies of people to do it for you? I don't think there is an obvious solution here, no matter how much it cost.
- Majromax 6y ago
  > I don't think there is an obvious solution here, no matter how much it cost. If that's the case, then it may be back to Youtube doesn't scale. > Second is the fact that moderation is a very mentally exhausting effort. You have an impossible trilemma. Youtube: * Allows/must allow anonymous or pseudonymous content creation, without the ability to ban bad actors; * Faces significant social (sometimes legal) liability for content hosting and especially curation/promotion; and * Wants to provide friction-free, speedy, and consistent policies to its content creators. Those can't all exist at the same time. Right now, what gives is a bit of #3 (seemingly arbitrary, severe, and algorithmic actions against channels) and a bit of #2 (occasional stories about how Youtube is evil for promoting flat-earth Nazism vaccine-denial theory to children or whatever). Youtube could go with a heavy-moderation approach, by requiring content creators first register with a government ID (plus business license in each country of intended audience) along with prior human-moderator approval of each uploaded video. But that would fundamentally change the platform into something more like a traditional TV network rather than dynamic social media. Other commenters here go heavier on the "free speech" side of things and think that Youtube ought to face no repercussions (social or legal) for hosted content, freeing it of the need to moderate. I'm not sure how we would get there from here, but nevermind that job for lobbyists. Society as a whole seems to implicitly solve this problem by not being a big giant curator under its own brand. Traditional media still exists as many (notionally-)separate outlets, each with their own branding and editorial views to offer at least limited competition. But if this is the only viable model, it's still another way of saying that Youtube doesn't scale.
- the8472 6y ago
  > I don't think there is an obvious solution here, no matter how much it cost. The issue is centralizing social decisions to a single entity. We already have a model where corporate entities provide services at scale without taking on much of the social aspect: Utilities and infrastructure. So find a way to become a video hosting utility. An uncaring pipe. Create tools for users to build their own fiefdoms and branding and have it moderate themselves. Tying advertising into that might be difficult, but perhaps it would be up to those users, pay for the services or automatically enter an agreement with advertisers. If advertisers pull their content for whatever reason then people could choose to pay themselves or perhaps make a deal with different advertisers. Hosting adult-related content that most brands don't want to touch? Perhaps make a deal with seedier advertisers. Youtube would still take its cut, but only as part of infrastructure fees. Want to promote your content to other circles? Well, you're essentially an advertiser yourself now.
Nasrudith 6y ago
To be frank the only hubris I see is from the critics who complain about YouTube that it "shouldn't be that big" and that "big tech is ruining society and democracy". The critics think that they know better than society and impose demands not codified in any law nor how things remotely work. How the hell should YouTube have decided to keep its size acceptably small then? Youtubes 1 to 99999 and hope people go to the right arbitrary subsegment? The demanda are arrogant and clueless as saying that builders should only use rubber headed hammers and nailguns should be banned because it would reduce construction noise knowing nothing about the field or its practicalities. Now /that/ is hubris.
amanaplanacanal 6y ago
They could certainly be smaller. If they were 50% smaller, which 50% of their current content do you think they should get rid of? Anything smaller or with more moderation will have less content. I suspect what we have now is the best that we can have.
- carapace 6y ago
  Panama! (Nice user name.) > If they were 50% smaller, which 50% of their current content do you think they should get rid of? Couldn't they just split in twain? https://en.wikipedia.org/wiki/Cell_division https://en.wikipedia.org/wiki/Cell_division (I lied when I said, "I don't have any answers". I do have answers.)

altgoogler 6y ago

Indeed, the question is what should be done, or how should it have been done differently? Should YT have grown more slowly? FWIW, I don't think money is the limiting factor here. If there was a known good solution, but it took $10B to implement, I think it would be done. You could pay X thousands (more) moderators, but there are drawbacks in a human solution as well. First is the fact since there's human judgement, rules will be applied inconsistently, and probably with some bias. Second is the fact that moderation is a very mentally exhausting effort. You can observe some of the cyclical nature of this balance from the news over the past few years. Article about moderators with PTSD? Well, automation is the solution. Automation fails in a lot of edge cases? Well, why not hire armies of people to do it for you? I don't think there is an obvious solution here, no matter how much it cost.

Majromax 6y ago
> I don't think there is an obvious solution here, no matter how much it cost. If that's the case, then it may be back to Youtube doesn't scale. > Second is the fact that moderation is a very mentally exhausting effort. You have an impossible trilemma. Youtube: * Allows/must allow anonymous or pseudonymous content creation, without the ability to ban bad actors; * Faces significant social (sometimes legal) liability for content hosting and especially curation/promotion; and * Wants to provide friction-free, speedy, and consistent policies to its content creators. Those can't all exist at the same time. Right now, what gives is a bit of #3 (seemingly arbitrary, severe, and algorithmic actions against channels) and a bit of #2 (occasional stories about how Youtube is evil for promoting flat-earth Nazism vaccine-denial theory to children or whatever). Youtube could go with a heavy-moderation approach, by requiring content creators first register with a government ID (plus business license in each country of intended audience) along with prior human-moderator approval of each uploaded video. But that would fundamentally change the platform into something more like a traditional TV network rather than dynamic social media. Other commenters here go heavier on the "free speech" side of things and think that Youtube ought to face no repercussions (social or legal) for hosted content, freeing it of the need to moderate. I'm not sure how we would get there from here, but nevermind that job for lobbyists. Society as a whole seems to implicitly solve this problem by not being a big giant curator under its own brand. Traditional media still exists as many (notionally-)separate outlets, each with their own branding and editorial views to offer at least limited competition. But if this is the only viable model, it's still another way of saying that Youtube doesn't scale.
the8472 6y ago
> I don't think there is an obvious solution here, no matter how much it cost. The issue is centralizing social decisions to a single entity. We already have a model where corporate entities provide services at scale without taking on much of the social aspect: Utilities and infrastructure. So find a way to become a video hosting utility. An uncaring pipe. Create tools for users to build their own fiefdoms and branding and have it moderate themselves. Tying advertising into that might be difficult, but perhaps it would be up to those users, pay for the services or automatically enter an agreement with advertisers. If advertisers pull their content for whatever reason then people could choose to pay themselves or perhaps make a deal with different advertisers. Hosting adult-related content that most brands don't want to touch? Perhaps make a deal with seedier advertisers. Youtube would still take its cut, but only as part of infrastructure fees. Want to promote your content to other circles? Well, you're essentially an advertiser yourself now.

Nasrudith 6y ago

To be frank the only hubris I see is from the critics who complain about YouTube that it "shouldn't be that big" and that "big tech is ruining society and democracy". The critics think that they know better than society and impose demands not codified in any law nor how things remotely work. How the hell should YouTube have decided to keep its size acceptably small then? Youtubes 1 to 99999 and hope people go to the right arbitrary subsegment? The demanda are arrogant and clueless as saying that builders should only use rubber headed hammers and nailguns should be banned because it would reduce construction noise knowing nothing about the field or its practicalities. Now /that/ is hubris.

amanaplanacanal 6y ago

They could certainly be smaller. If they were 50% smaller, which 50% of their current content do you think they should get rid of? Anything smaller or with more moderation will have less content. I suspect what we have now is the best that we can have.

carapace 6y ago
Panama! (Nice user name.) > If they were 50% smaller, which 50% of their current content do you think they should get rid of? Couldn't they just split in twain? https://en.wikipedia.org/wiki/Cell_division https://en.wikipedia.org/wiki/Cell_division (I lied when I said, "I don't have any answers". I do have answers.)