> Despite that, Pai did not criticize Facebook and Twitter for restricting Trump this week. When asked if he "agree[s] with Facebook and Twitter's decision to pull the president off of social media,"
> Pai responded, "Given the circumstances we saw yesterday, I'm not going to second-guess those decisions."
Some people believe that having a choice between 128Mbit/sec DSL and a cable company that does everything it can to bundle services and make you pay for stuff you don't want is enough competition to ensure companies don't misbehave.
I’m paying €32 / mnth for 1000/1000 fibre in an urban, but not mega city urban, area of the Netherlands. Internet access should be a commodity like electricity and gas by now. It’s become that important.
So you had a right taken away and replaced with the “benevolence” of corporations and you’re happy with this? You gained nothing, you only lost things.
Speaking as someone who use to work for R&D within the industry (in the resource allocation space), the OIO (the name for the FCC's implementation of net neutrality) injected a significant amount of regulatory uncertainty into our work resulting in dried up funding from certain sources. The rules themselves were expansive and went well beyond the existing status quo in a number of different ways.
As an example, the OIO had lots of carve-outs for what they termed "specialized services" in the text; however, what would be considered a specialized service in the future was ambiguous as the definition wasn't governed by any principled approach. Some of my colleagues more familiar with regulatory policy said it would be a minimum of 10 years before we even began to get a clearer picture of how the rules would work in practice. To give a concrete example, we're essentially in year 6 of Net Neutrality in the EU and we still have no clear guidance on how certain important aspects of 5G (such as network slicing) will interact with it. (source: https://ieeexplore.ieee.org/document/9017846)
This was never going to go anywhere. The same companies who supported and appreciated Pai’s net neutrality ruling (big ISPs) are all content companies now too, and they benefit as much as Facebook and Twitter from Section 230.
Unlike net neutrality, which split the business community between network owners and application companies, I can’t think of a category of business that benefits from repeal or weakening of Section 230. IMO it’s a purely performative policy discussion.
I don't disagree with you but it makes me sad that we accept it as axiomatic that any law change is completely untenable unless it financially benefits some business sector.
I don’t think that is true in general; labor and environmental law stand as two huge counterexamples. What I’m saying is that Pai specifically is one of the regulators motivated this way.
I have many disagreements with Ajit Pai, but I think this is a smooth move. I think there needs to be some regulation on content moderation policies for the platforms which are becoming increasingly entrenched as the new public square. But even just the last few days have proven it's a difficult subject on which there is little public consensus. I haven't read up on Section 230, but I'd rather them not rush a solution out the door in the final hours of the current administration.
I want to point out there really are two sides to Section 230. On one hand, it's absolutely crucial to have it if we want most of the online communications tools we've become accustomed to to exist. On the other hand I think some companies, like Facebook specifically, have massively abused the protections of Section 230 to benefit themselves at the expense of the general public. When you're to the point of hiring psychologists to manipulate your users into "engagement" regardless of the impact to mental health and public discourse, I think that's the point where everyone needs to recognize that _something_ needs to be done in terms of regulation.
When regulation _does_ happen, I hope lawmakers around the world give no consideration to the impact on Facebook just as Facebook's given no consideration to the negatives they're creating for society. Let them fail if necessary IMO.
And finally, the best middle-ground I've seen for Section 230 was here on Hacker News. I don't remember which user said it, so I'm sorry I can't attribute it, but the idea was solid. Things like reverse chronological data feeds that a user is specifically asking for should be protected under Section 230. Things like recommendation and engagement algorithms should be considered publishing and platforms should be liable for that content since they're surfacing it and putting it in peoples' faces.
I've paraphrased that, but I really hope the general idea gains traction. I've been very anti-censorship my whole life, but there's no denying that social media can have a severe, negative impact if the platforms are susceptible to propaganda and misinformation, so I think it's time to talk about moderation.
I would love a "Choose your algorithm" law for content recommendations. The default would be chronological order, and you can opt in to whatever other algorithms these platforms offer.
It would be hard to limit algorithmic internal promotion without limiting algorithmic external promotion, including the entire ad bidding ecosystem. And perhaps that’s a good thing - in addition to Facebook taking responsibility for Facebook links its non-opt-in algorithms promote, it should take responsibility for ad content. There are arguments that this could disadvantage smaller businesses in advertising. That’s a secondary concern to the deprogramming of a radicalized populace.
> Things like recommendation and engagement algorithms
So, how does this work for something like a search engine? Does Google/Bing/DDG need to get an audit from the government that their recommendation algorithm is sufficiently non-creative so as to be not publishing?
I hadn’t thought of that before, but you’re right.
I’m not sure how to properly distinguish between the case of “I’m a veterinarian so when I search for ‘mouse’ I mean the animal not the peripheral” versus the case of “I click on all the conspiracy theories so when I search for ‘president’ I mean the Satanic space-lizard who kidnapped Elvis and eats babies”.
This is, I'll note without any judgement, a surprisingly anticapitalist opinion. You're essentially saying that you don't want to allow companies to compete on how good the product is from a consumer perspective (as a consumer, I value not having to precisely describe my search queries, I don't want to have to write SQL in my search tool!).
But also, when I search Google, I want to know if what category the results list is coming from. The recommendation engines have gotten so bad that I quite frequently will use private browsing on a lot of my searches that are intended for a single-use purpose (like when I read about something and want more info on it), without having to have future searches (or advertisements!!!) coming back in my face.
A simple example is when I bring up a playlist on Youtube for when one of the grandkids is over. Well I really don't want my default recommendations to be mostly 70's metal but contaminated with The Wheels On The Bus peppered through the playlist.
I agree there needs to be some reform of Section 230 in that direction. The law simply can't handle what the internet and social media have become.
When a platform is curating and recommending content to users, it should be treated more like a publisher and be held responsible for the content it promotes. It doesn't matter if the curation is done by hand or with algorithms.
Providers that simply allow the hosting and basic browsing of content should be protected as long as they are willing to remove illegal content when it is found.
Also, I wouldn't call it censorship at all. If a newspaper publishes an article they claim some responsibility for its content. The same should apply if Facebook highlights an article or YouTube recommends a video.
> think some companies, like Facebook specifically
Meta comment, Facebook seems to be the main go to comic villain. Things that were described in the comment like driving engagement is also done by platforms like Twitter. But twitter seems to be in peoples good graces for some reason.
I wouldn't say that Twitter is in my good graces, just that I think Facebook is worse. I don't have any empirical data to back that up, so it's just a general assessment from watching the industry. In my mind, Facebook pioneered a lot of the bad practices and has been the leader in unethical practices.
For me, that makes Facebook worse than the others because increasing profits by being unethical is easy and as soon as one company in an industry starts doing it, others are left in an impossible situation. Either they become unethical to compete or they get outcompeted and fail.
Facebook has also demonstrated they're not going to change until someone forces them. Even with all the pushback in the last couple years, they're still pulling anti-consumer stunts like the forced account tying / data collection with Oculus and Instagram. I tend to judge companies by their actions rather than their words and Facebook hasn't ever done anything to make me think they care even the tiniest little bit about the public good.
IANAL, but if you remove Section 230 protection from algorithmically-curated content, Facebook would have to switch entirely to manually curating everything to avoid accidentally including libelous user-submitted content they could be sued over.
Even basic features like ordering new posts by people you follow to show the most popular ones first are algorithmically-driven.
There's no way Facebook could manage to do anything individualized, so it would turn into a chronological feed + manually-curated, Taboola-style content (which I doubt anyone wants).
I agree with your perspective on the negative impacts of social media, and I'm open to the idea of some kind of regulation to improve things. But I think effectively banning algorithms from internet services is going to hurt more than help.
My general reaction to that is "tough luck!" If a site algorithmically generated recipes, but accidentally created recipes that poison people once in a while we'd ban those algorithms without thought or, at the very least, make the algorithm owners liable.
If social media is destabilizing the mental health and critical thinking skills of the general public to the point that a subset of the population is becoming dangerous (and violent), isn't that basically the same thing?
I don't care if Mark Zuckerberg needs to buy a smaller mansion if that means my friends / neighbors / family aren't going to get brainwashed by propaganda.
Yeah, based on the really awful cases of my wife's family, the problem exists between the keyboard and the chair. When faced with fact-checking or truth that refutes their insane conspiracies, they double down and call the fact-checkers or truth-tellers liars and invent even more insane conspiracies about how the "real truth" (read: heinous virulent BS) is being suppressed because the clickbait article about how "liberals are planning to euthanize all whites!" now has a disclaimer under it that says "this claim is disputed" (a caveat so weak that it'd be almost funny if it weren't depressing how many psychos eat up the totally-untrue headline).
It's the only thing that is starting to turn them on Fox News for example; though they still watch it religiously, the fact that the most feeble, CYA-grade questions were asked during their brief "news" segment that slightly challenged Trump's baseless "stolen election" narrative now means that they've moved from "only Fox is telling the truth" to "well, I don't like everything they say, bit they do a good job, like Tucker Carlson" (who is probably the most prominent disinformation propagandist on the whole festering network). In response, they've turned to just any old 2-hour-old "freedom eagle dot biz" site as their "real true sources", because Fox had a brief lapse of allegiance to the Trump cult.
You can't fix that with an RFC, a law, or a moderation policy because they've rejected any possibility of accepting the truth. They're thoroughly deceived, most especially by themselves, and any attempt to un-deceive them is seen as a sinister plot. They refuse the truth outright because it's not what they want to believe.
You can't fix it by shutting down or "breaking up" the site where they go to get their much-desired disinformation, because there's always another con artist with a Squarespace or WordPress license who'll gladly capitalize on having less competition in the "cult propaganda dissemination warehouse" market (like OANN capitalizing on Trump's brief lashing out at Fox, or Parler springing up because Republicans started getting "this might not be true" disclaimers on their lies).
I've always seen "you can't fix stupid" as a harsh redneck insult, but I'm starting to believe a slight variant on it: "you can't fix willfully-ignorant." You'd have to change their thought patterns and their habits and often their motivations, and that kind of change doesn't come from outside but from the person themselves.
I think having massive, mainstream social media sites acting as onboarding for lunacy is a major contributing factor. Competitors trying to capitalize on the fringes wouldn't have a viable market to capture if Facebook et al wouldn't have created it in the first place.
The antivaxxers of the world might be a counterexample here; for example, here's [0] a newspaper cartoon from the 30s showing "misinformation" as a cliff which "antivax" and "anti-fad" and others willingly walk right off. The antivax movement, fueled by misinformation at every step, long pre-dates Facebook or the internet, instead living on backwater gossip and sketchy snake-oil salesmen, who have since moved on to running their own websites and TV shows (see also: "Dr. Mercola" and "Dr. Oz").
The internet didn't create the market for crazy fringe beliefs; like most markets, the internet just accelerated and connected it.
Banning algorithms that push users into extremist positions far beyond what is generally acceptable to the public would be a net improvement over the situation that exists today. Conspiracy theories used to be fringe content that one had to make an effort to seek out.
The biggest problem facing society today is the complete lack of discourse across segments of the population. The filter bubble exacerbates this. It the tech platforms were to sprinkle in even just a little bit of content from outside of one's filter bubble, there would be a benefit to society. Just look at the platforms of old (broadcasting, newspapers): you could only target general populations. Today's platforms can target across so many variables there is no need to publish content that's generally acceptable when you can limit its viewing to those that your algorithm has determined can be easily manipulated. Maximum effect, minimum cost.
Removing Section 230 would be horrific because it would prevent anyone else from being able to make competing platforms to those that have the massive ability to moderate everything.
However, the platforms are too large to be allowed to moderate whatever the hell they want as they are now, while slurping up all our personal information and data. I think we need real reform, and people need to leave the big platforms and find better alternatives. I wrote about this last month:
45 comments
[ 2.9 ms ] story [ 108 ms ] threadPlease don't fulminate
> Pai responded, "Given the circumstances we saw yesterday, I'm not going to second-guess those decisions."
As an example, the OIO had lots of carve-outs for what they termed "specialized services" in the text; however, what would be considered a specialized service in the future was ambiguous as the definition wasn't governed by any principled approach. Some of my colleagues more familiar with regulatory policy said it would be a minimum of 10 years before we even began to get a clearer picture of how the rules would work in practice. To give a concrete example, we're essentially in year 6 of Net Neutrality in the EU and we still have no clear guidance on how certain important aspects of 5G (such as network slicing) will interact with it. (source: https://ieeexplore.ieee.org/document/9017846)
https://arstechnica.com/tech-policy/2020/04/fcc-blasted-for-...
https://arstechnica.com/tech-policy/2020/04/ajit-pai-uses-ba...
https://arstechnica.com/tech-policy/2020/05/fcc-fines-sincla...
https://arstechnica.com/information-technology/2020/01/ajit-...
Unlike net neutrality, which split the business community between network owners and application companies, I can’t think of a category of business that benefits from repeal or weakening of Section 230. IMO it’s a purely performative policy discussion.
When regulation _does_ happen, I hope lawmakers around the world give no consideration to the impact on Facebook just as Facebook's given no consideration to the negatives they're creating for society. Let them fail if necessary IMO.
And finally, the best middle-ground I've seen for Section 230 was here on Hacker News. I don't remember which user said it, so I'm sorry I can't attribute it, but the idea was solid. Things like reverse chronological data feeds that a user is specifically asking for should be protected under Section 230. Things like recommendation and engagement algorithms should be considered publishing and platforms should be liable for that content since they're surfacing it and putting it in peoples' faces.
I've paraphrased that, but I really hope the general idea gains traction. I've been very anti-censorship my whole life, but there's no denying that social media can have a severe, negative impact if the platforms are susceptible to propaganda and misinformation, so I think it's time to talk about moderation.
So, how does this work for something like a search engine? Does Google/Bing/DDG need to get an audit from the government that their recommendation algorithm is sufficiently non-creative so as to be not publishing?
I’m not sure how to properly distinguish between the case of “I’m a veterinarian so when I search for ‘mouse’ I mean the animal not the peripheral” versus the case of “I click on all the conspiracy theories so when I search for ‘president’ I mean the Satanic space-lizard who kidnapped Elvis and eats babies”.
mouse => computer mouse | house mouse | feeder mice | common mouse diseases
president => PTA president | US president | lizard president
A simple example is when I bring up a playlist on Youtube for when one of the grandkids is over. Well I really don't want my default recommendations to be mostly 70's metal but contaminated with The Wheels On The Bus peppered through the playlist.
When a platform is curating and recommending content to users, it should be treated more like a publisher and be held responsible for the content it promotes. It doesn't matter if the curation is done by hand or with algorithms.
Providers that simply allow the hosting and basic browsing of content should be protected as long as they are willing to remove illegal content when it is found.
Also, I wouldn't call it censorship at all. If a newspaper publishes an article they claim some responsibility for its content. The same should apply if Facebook highlights an article or YouTube recommends a video.
do we?
Meta comment, Facebook seems to be the main go to comic villain. Things that were described in the comment like driving engagement is also done by platforms like Twitter. But twitter seems to be in peoples good graces for some reason.
For me, that makes Facebook worse than the others because increasing profits by being unethical is easy and as soon as one company in an industry starts doing it, others are left in an impossible situation. Either they become unethical to compete or they get outcompeted and fail.
Facebook has also demonstrated they're not going to change until someone forces them. Even with all the pushback in the last couple years, they're still pulling anti-consumer stunts like the forced account tying / data collection with Oculus and Instagram. I tend to judge companies by their actions rather than their words and Facebook hasn't ever done anything to make me think they care even the tiniest little bit about the public good.
Even basic features like ordering new posts by people you follow to show the most popular ones first are algorithmically-driven.
There's no way Facebook could manage to do anything individualized, so it would turn into a chronological feed + manually-curated, Taboola-style content (which I doubt anyone wants).
I agree with your perspective on the negative impacts of social media, and I'm open to the idea of some kind of regulation to improve things. But I think effectively banning algorithms from internet services is going to hurt more than help.
If social media is destabilizing the mental health and critical thinking skills of the general public to the point that a subset of the population is becoming dangerous (and violent), isn't that basically the same thing?
I don't care if Mark Zuckerberg needs to buy a smaller mansion if that means my friends / neighbors / family aren't going to get brainwashed by propaganda.
It's the only thing that is starting to turn them on Fox News for example; though they still watch it religiously, the fact that the most feeble, CYA-grade questions were asked during their brief "news" segment that slightly challenged Trump's baseless "stolen election" narrative now means that they've moved from "only Fox is telling the truth" to "well, I don't like everything they say, bit they do a good job, like Tucker Carlson" (who is probably the most prominent disinformation propagandist on the whole festering network). In response, they've turned to just any old 2-hour-old "freedom eagle dot biz" site as their "real true sources", because Fox had a brief lapse of allegiance to the Trump cult.
You can't fix that with an RFC, a law, or a moderation policy because they've rejected any possibility of accepting the truth. They're thoroughly deceived, most especially by themselves, and any attempt to un-deceive them is seen as a sinister plot. They refuse the truth outright because it's not what they want to believe.
You can't fix it by shutting down or "breaking up" the site where they go to get their much-desired disinformation, because there's always another con artist with a Squarespace or WordPress license who'll gladly capitalize on having less competition in the "cult propaganda dissemination warehouse" market (like OANN capitalizing on Trump's brief lashing out at Fox, or Parler springing up because Republicans started getting "this might not be true" disclaimers on their lies).
I've always seen "you can't fix stupid" as a harsh redneck insult, but I'm starting to believe a slight variant on it: "you can't fix willfully-ignorant." You'd have to change their thought patterns and their habits and often their motivations, and that kind of change doesn't come from outside but from the person themselves.
The internet didn't create the market for crazy fringe beliefs; like most markets, the internet just accelerated and connected it.
[0] "Here's What A Depression-Era Cartoonist Had To Say About The Anti-Vaccination Movement" https://m.huffpost.com/us/entry/6608366/amp
The biggest problem facing society today is the complete lack of discourse across segments of the population. The filter bubble exacerbates this. It the tech platforms were to sprinkle in even just a little bit of content from outside of one's filter bubble, there would be a benefit to society. Just look at the platforms of old (broadcasting, newspapers): you could only target general populations. Today's platforms can target across so many variables there is no need to publish content that's generally acceptable when you can limit its viewing to those that your algorithm has determined can be easily manipulated. Maximum effect, minimum cost.
Like I said, I'm open to the idea that something should be done, just not sure exactly what.
However, the platforms are too large to be allowed to moderate whatever the hell they want as they are now, while slurping up all our personal information and data. I think we need real reform, and people need to leave the big platforms and find better alternatives. I wrote about this last month:
https://battlepenguin.com/politics/is-meaningful-section-230...