The US Supreme Court doesn’t understand the internet
wired.com
wired.com
For example it talks about routing:
> “The Internet uses ‘packet switching’ communication protocols that allow individual messages to be subdivided into smaller ‘packets’ that are then sent independently to the destination, and are then automatically reassembled by the receiving computer. While all packets of a given message often travel along the same path to the destination, if computers along the route become overloaded, the packets can be re-routed to less loaded computers.”
And that was back in the 1990s. What people seem to really mean when they say “the Supreme Court doesn’t understand X” is that the Supreme Court doesn’t share the values they associate with X. That’s almost certainly the case. From the Supreme Court’s view, the internet is just another part of the national economy. They don’t understand or care about the distinct cultural values of people in that sector.
Not sure who they are a mouthpiece for but they certainly don’t mind being one.
[1] https://www.supremecourt.gov/opinions/20pdf/18-956_d18f.pdf
The decision does show that the Supreme Court justices are savvy as to how the tech industry works, at least they know from a policy perspective they can't strike down the whole Android ecosystem without causing mayhem. And they know they can't affirm copyrights of APIs for similar reasons.
But from a legal perspective the way they outright dodged the legal question they were supposed to answer (i.e. are APIs copyrightable?) and the way they just asserted the conclusion about fair use without pretending to apply legal reasoning (together with a disclaimer that they haven't tried to change the existing law even if it looks like they have), I wouldn't say it's one of their better decisions.
As you say, in general people in tech have a tendency to assume judges are idiots. Most often they aren't, because judges sitting in the top courts are a handful of people at the pinnacle of their field, and they don't get there by being idiots. I agree with you that it's a good idea to sample some opinions from Supreme Courts or perhaps even others appeal courts to gain an appreciation of how judges tend to think. I just think Google v. Oracle is one of the worse ones to start with..
Should it? This feels a lot like a "have your cake and eat it too" situation. Either you're a neutral party or you are not. Being a trillion-dollar company doesn't exempt you from that. Letting a computer make the decisions instead of a person doesn't exempt you either.
Youtube would be in no danger if all it was doing was keyword-matching with recommended videos. Instead it created profit by designing its algorithm to maximize engagement by any means necessary, especially via outrage and bandwagoning. What the fuck did they expect to happen when they created a money engine that runs on human emotional intensity?
Certainly. But Section 230, at least from my reading, does not protect them for the promotion of content. I could be wrong about that. The Supreme Court will decide. Personally I'd find it delightful if the rage-engine got smashed with a legal hammer and my Youtube recommendations were as useful as they were fifteen years ago.
> YouTube decided against labeling 22 channels identified by ProPublica, but it's not entirely clear why.
https://www.propublica.org/article/how-china-uses-youtube-an...
> YouTube said the clips did not violate its community guidelines.
> The warehouse accounts on YouTube have attracted more than 480,000 views in total. People on YouTube, TikTok and other platforms have cited the testimonials to argue that all is well in Xinjiang — and received hundreds of thousands of additional views.
https://themarkup.org/google-the-giant/2021/04/08/google-you...
> [YouTube] even suggested videos for campaigns with terms that it clearly finds problematic, such as “great replacement.” YouTube slaps Wikipedia boxes on videos about the “the great replacement,” noting that it’s “a white nationalist far-right conspiracy theory.”
> Some of the hundreds of millions of videos that the company suggested for ad placements related to these hate terms contained overt racism and bigotry, including multiple videos featuring re-posted content from the neo-Nazi podcast The Daily Shoah, whose official channel was suspended by YouTube in 2019 for hate speech. Google’s top video suggestions for these hate terms returned many news videos and some anti-hate content—but also dozens of videos from channels that researchers labeled as espousing hate or White nationalist views.
> Even after [Google spokesperson Christopher Lawton] made that statement, 14 of the hate terms on our list—about one in six of them—remained available to search for videos for ad placements on Google Ads, including the anti-Black meme “we wuz kangz”; the neo-Nazi appropriated symbol “black sun”; “red ice tv,” a White nationalist media outlet that YouTube banned from its platform in 2019; and the White nationalist slogans “you will not replace us” and “diversity is a code word for anti-white.”
Or think of it this way: How is Mastodon supposed to take on larger social networks without recommending people to follow? Should every Mastodon server operator be legally liable for recommending someone harmful?
But I'm just not sure how or why online platforms get to have their cake and eat it too. If the NYTs publishes a story that eating Tidepods is healthy and encourages kids and parents to do so, they get sued. If Facebook creates an algorithm that causes the same or similar to happen they get a free pass. They either have to be a public speech platform where anyone can say anything as long as it isn't literally breaking the law, or they have to follow the same rules as other entities that curate content. If you want to say "why not both?" then that's fine but you have to apply that to all entities, not just online content.
If you said something libelous about me on HN, I can sue HN for publishing and promoting the comment?
A platform that "allows everything that isn't breaking the law" is a platform that is 99% spam.
Why is it different if a "newspaper" does it if HN is curating the content algorithmically? I'm honestly not sure what the difference is.
You could imagine the kind of elements that might matter for a fact pattern that would emerge from a deposition: revenue and size of website, proportion of website revenue directed towards moderation, percentage of requests that identify illegal material that are responded to, manner of response, tools provided to users, the types of content actually hosted on the site, nature of the algorithm itself, discussions that were had internally about access to harmful content. HN is a text-based website (which also mitigates the harm claim), it gets maybe in the orbit of a few hundred submissions a day, the vast majority of possible harm is when a submission is connected to a topic likely to cause legal issues, and in my experience such topics are typically flagged within a few minutes and removed quickly. There's no mechanism to directly communicate with users, there is no mechanism to follow users, there's no mechanism to change what you see on the front page based on what you clicked before. Everyone is exposed to the same content.
By contrast, thinking about the companies that are actually the target of these lawsuits I was at the TASM (Terrorism and Social Media) 2022 conference -- some of my research is adjacent to this but I've never done any work on terrorism and my social media work involves alt-tech stuff, not the big social media platforms -- where the keynotes were harm policy leads for Europe for Twitter, Facebook, and YouTube, all of whom made it clear that their position was that it is incumbent on academics and government agencies to identify harmful content and work with social media, because every region has their own unique content challenges and it's not possible for Tech companies to handle those at a global scale. A question was asked of the panel that went something like "Meta was, as it admits, instrumental in spreading the violence-inciting messages that drove the anti-Rohingya pogroms in Myanmar. The defense is that Meta wasn't prepared for, or staffed for, detecting issues in a single small market, and things snowballed quickly. You could hire 100 full time content people in that country for the salary of a single person sitting on the panel or a single SWE, so how could resource constraints be the issue?" and the answer was "We're already devoting enough resources to this problem." I think that's an incredibly shitty answer to the question, and I think a deposition could surface exactly this kind of logic, and to me that would be a fact pattern that supports liability. I hope they get their pants sued off in every jurisdiction that allows it. It's clear an aversion to staffing is a huge part of the issue.
So from my perspective, I think the typical process for resolving civil liability + reasonable assumptions about how courts should interpret fact patterns is likely to get to outcomes I'm plenty happy with.
(In the two cases in front of SCOTUS right now it seems like the victims, who have basically no evidence connecting the perpetrators of the violence to the social media services used, would win: the argument seems to be of the form "some terrorist killed my family member, and some other terrorists got radicalized online, ipso facto social media is liable". I don't think that'd be a winning case with or without s.230)
Either way I think we're going to see a big swing back to authoritative sources because the very technical problems you mention will be taken advantage of by new tools and so the already meaningless content will not even be generated by humans. The Internet in the sense of "publishing content" will be meaningless and unprofitable [1].
[1] Obviously there will exist use cases where this is not the case
Really everything that you're saying doesn't have shit to do with authoritative sources, but authoritarian sources. If you're a big nice identified company, or you're a member of "the party" you get to post permitted information. If you're not, well, better learn how to post those pics of a your senator pulling some crap on the darknet.
You can just register anonymously like you do on HN. Though for social media sites or similar having "verified human" seems like not just a good idea but ultimately the direction we'll go.
> Really everything that you're saying doesn't have shit to do with authoritative sources, but authoritarian sources.
You are really jumping the gun here so I'm not going to respond to your points here since I wasn't making those.
You're watching Tie Your Mother Down, Queen, Rock in Rio 1985. Would you like to see (select as many as you want):
More videos by or about Queen
More videos from Rock in Rio 1985
More videos from 1985
More video about Mothers
More video tagged Live Concert
More video tagged Progressive Rock
More video tagged Rio de Janeiro
And either way, why bother? Just don't recommend content there's no point except to drive engagement which is fundamentally no different than what Facebook (or whoever) is doing.
Why would it be safe for them to use an older recommendation system? It doesn't solve the problem, if their older system recommends a terrorism video, even if it only did so because that video came up chronologically, they're still liable.
I would think they would need to just stop allowing the general public to upload videos anymore and only permit trusted media companies and influencers (ones known to not create controversial content) to do so. Probably after being approved through a vetting process where their lawyers can look through at least some of the content first.
A system that keyword matches isn't making recommendations, it's just keyword matching based upon the user's request. The law actually cares about intent and how things function, not just hypothetical possibilities that can occur, i.e. the law cares about what does happen and why it happens that way. So it's pointless to characterize a non-recommendation system as a recommendation system as a means of end-running an argument.
Do you not recognize how lousy of a video sharing website this would be? Spammers are going to be constantly uploading marketing and other low-quality content with irrelevant keywords, while users that actually put work into making good quality videos will see their results pushed to the bottom quickly. How will you deal with that without implementing a system that can identify and recommend non-spam videos? Even the oldest versions of Youtube were boosting videos that got lots of likes.
>the closest match
How is deciding the "closest match" not considered a recommendation? They all have the user's keyword, what other criteria will you use?
Not sure why that's my problem, I'm not the one making money by promoting reactionary videos to reactionaries.
>How is deciding the "closest match" not considered a recommendation? They all have the user's keyword, what other criteria will you use?
Because it's not a recommendation, some are better matches than others, thats' all. Some match the entire keyword, some just parts, some in different places... I don't understand what is difficult about this for you.
The reason I think we should see it as our problem is because I think the solution companies arrive at is just to turn the internet into cable TV, where only approved media organizations are able to share content because of liability concerns.
I'll admit, we may even be better off as a society of communication was less "democratized." There certainly would have been a lot less covid and election misinformation out there if every rando wasn't able to have their uninformed ideas broadcasted by giant platforms.
YouTube takes it down. Not as fast as it’s put up. But fast.
I found it irritating because I wanted to know what my daughter had been exposed to, but couldn’t. Her history linked to removed videos.
The titles were nonsensical - mostly Unicode homoglyphs.
It wasn’t child pornography, but was definitely grooming material.
You're acting like any kind of algorithm is automatically a recommendation engine that should terminate Section 230 protections, but I don't think it's that simple.
The later is just sorting by additional attributes (video length, keywords in content, likelihood of clicking->watching, keywords of past content watched, ...). Youtube doesnt care what you watch... as long as they match what you want to watch to a list of videos, you stay on the site. If they dont, then you leave. The actual content of the videos doesnt matter to youtube. In this way, the page that displays the feed is very similar to showing search engine results sorted by best match, where the keywords are pulled from your past videos.
If sorting is now promotion and prohibited by 230, then the internet is f'd. Search engines are going to be completely useless.
You're not wrong, but in addition to the leeway afforded to the rich and powerful by "the law", there is also substantial leeway afforded to every individual under "reality", and one option available is that it is technically possible to behave however one likes, including in a manner that is not compliant with "the law" or "the social contract", neither of which I or most anyone else was consulted on, despite living in a country governed by "democracy".
Interestingly, it seems like it is those who are classically "less intelligent" who are most likely to realize that this powerful exploit exists, buffoonery like January 6, anti-vaxx, and shooting power stations with an off the shelf rifle being prime examples of this.
I sometimes wonder if like corporations or most any other organization on the planet, it might be prudent to review our governmental and legal standard operating procedures from time to time to ensure they are working as intended (underlying, actual intent (as opposed to proclaimed intent) being another matter that more than a few people are starting to become rather dangerously curious about).
Giving the user the ability to search doesn't mean you're curating content with a recommendation engine.
But making editorial decisions downstream—what it chooses to show to users first vs last—doesn't change what people contribute, and this is where Section 230 is trying to provide a balance. People can contribute whatever great or terrible things they contribute, and Section 230s stance is, the provider just has to remove it if it's actually out of bounds. That allows sites to have their own editorial decision-making downstream while allowing user-generated content to flourish upstream, so those decisions matter at all. If the burden of moderation is moved farther upstream, to the point at which people contribute, you've increased the barrier to creating content entirely, you've turned every social network into a newspaper, and you get none of the benefits of the scale, low barrier to entry, and low marginal cost of the Internet to content creation.
In my view, an algorithm is no different than the downstream moderation decisions every website makes with or without an algorithm. Of course an algorithm has values. That's the point! They should have values, and people should choose which sites to use based on how well those design decisions translate into value for the user. Rather than the existence of out-of-bounds content in their systems at all, we should be judging YouTube and others on how well they identify and how quickly they remove that content and how quickly they update their policies (which is just another word for algorithm) to reflect those new values. This is explicitly a tradeoff for making mistakes in the interest of making progress versus safety, because what I'm advocating here does mean you get terrible stuff posted to places where it can get a lot of reach, and it's more likely to get that reach than with a system that moderates farther upstream. Where I think we should be coming down stronger is in regulating reach, not access (at least not entirely). The speed of the stream—the speed of virility—should be slowed to allow for reason to re-enter the conversation. That's where I think regulating content on the Internet needs updating, but not so far upstream that you stop it from getting any kind of reach at all.
The practical framework for regulation that I'd suggest is, if something hits a threshold of reach, it gets moderated more heavily and judged more heavily by authorities.
Related: Deciding on what counts as "out of bounds" content is also extremely hard, but we can put that side for a second, because I think the main issues here are 1) if a moderation decision is required, and 2) where the moderation decision occurs.
The tl;dr is that it may not be possible to split a hair as thin as the difference between an automated recommendation algorithm and automated (or manual) moderation (if the system chooses not to put your tweets in the Trending Topics, are we refraining from up-signalling you or down-signalling you? Is "signal" on a continuous real number line or two separate number lines?).
If the Court rules narrowly against Google, it'll be a major change to the function of much of the Internet (we can expect companies to respond to the new liability by switching off their recommendation systems completely), but the fundamental way-of-life we know today will continue. If the Court rules broadly against Google, it functionally kills S230 and opens the entire Internet up to massive liability lawsuits in a way that may end online fora for all but the "judgment proof" of the world.
Not even HN would likely be able to afford to operate if anyone posting a controversial idea slips through the moderation cracks and gets seen. The site would have to go to moderate-all-by-default, not post-and-then-moderate.
Of course I'm not a lawyer. But section 230 is not the only protection here.
For example, if a newspaper published a reader letter that accused some rich dude of having defrauded them, the newspaper itself would probably be liable just as much as the reader whose letter was sent. Apply this same logic to any HN comment and you'll get a huge issue - and being a non-profit is entirely irrelevant to this.
Basically anything First Amendment-grounded can pass the bar of worth bringing to court because the First Amendment protections are a patchwork of carve-outs, exceptions, and careful interpretations to stretch the ideal of "The government cannot constrain the people here" over the reality of "Some speech is harmful in a way that cannot be made whole."
There's nothing about being a nonprofit that prevents some Internet rando from using your forums to post massive obviously-false defamatory statements. The only thing that keeps the forums themselves from catching a lawsuit when that happens is s230.
Section 230 came about because of suits against for-profit entities. Has there ever been suit against the likes of IRC, USENET, or BBSes for libel from content of their participants?
... But that's a little irrelevant because those services are ghost towns relative to the past now. The more pressing reason nobody would bother to sue those channels is because nobody cares what's said there. If the loss of s230 shuttered the web services, and there were a migration back to those channels instead of just a quieter internet... Eventually a threshold would be reached that would be worth it for somebody to fire off a salvo of lawsuits against those service providers. A lot of IRC and USENET nodes are tied to institutions with enough assets that they aren't judgment-proof.
You could potentially attack a forum by posting verboten content on it then turn around an petition the government/state to sue the forum out of existence. A new form of SLAPP.
What order should those keyword matching videos come back in? By total views, by 30 day views, popularity, by upvotes, by downvotes, by keywords in the title, the description, the comments, the video itself?
Any choice made would be effectively indistinguishable from "designing the [search] to maximize engagement" as it comes to the law, since search itself is a method of surfacing new videos, and the order (even if only the default order) would matter.
I don't think the law has a definition for that term. I think it's -- probably correctly -- up to judges (and/or juries, depending on the type of case) to weigh intent and decide if a search engine to be liable for the results it produces.
This idea that "doing literally anything is an evil engagement maximizer" seems too simplistic for how the world actually works.
All of those approaches can lead to illegal content, and the entire point of Section 230 is that regardless of which they choose, they are immune from any legal repercussions of surfacing that content.
In meatspace we don't generally hold those making recommendations liable for the 3rd party content. We shouldn't online either.
As we should, IMO.
How come you and Jeff Kosseff seem to think opposite ideas about whether 230 was intended to enable content providers to moderate content? I wonder who is right.
Say your friend looks great in a piece of clothing. Based on just this, the next time they ask you what they should buy, you suggest the style and brand they looked good in before.
A day later, you discover that the brand in question clubs baby seals.
Does this suddenly make you liable for all the atrocities the brand commits?
Superficially, a YT recommendation is based on metadata - video length? Did the user watch the video beyond threshold values? Did they comment or react to a video? Where do this user’s metrics like in comparison to metrics of other users who watched the same video?
I concluded that since the algorithms aren’t moderating the content, just your access to it, they are not afoul of S230.
Maybe the answer is to require all social media to provide a strictly timeline based view. Even here, the submission time stamp is purely metadata about the content, as is the watch duration or “comment/react?” flags.
If you ignore the emotional manipulation sophistry and look at the logical content it looks a lot more like a thought disorder than anything else.
A judge's job is to understand the law. It is the job of the lawyers to bring facts and argumentation before the judges to pursuade them that the law should be interpreted one way or the other. The expertise is expected to live with the lawyers, not the judges (and the ability to find expert witnesses and spin up on the details of what they're advocating is a lawyer skill).
Thats probably overanalyzing a narrative piece for their owners but its what this accumulates to at scale.
As it says on the Supreme Court site:
"Issue: Whether Section 230(c)(1) of the Communications Decency Act immunizes interactive computer services when they make targeted recommendations of information provided by another information content provider, or only limits the liability of interactive computer services when they engage in traditional editorial functions (such as deciding whether to display or withdraw) with regard to such information."
https://www.scotusblog.com/case-files/cases/gonzalez-v-googl...
(I'm not trying to imply that YouTube should work how specially I want it to, just using myself as an example of what I think many people use it for)
I think they would be liable if they took emails that weren’t addressed to you and showed them in your inbox and it did some harm.
Or if they showed an ad that did some harm.
Your suggestion is that there would be a carve out for spam filtering? Or that Google deciding what goes in "Promotions" and what goes in "Priority Inbox" isn't a recommendation?
My suggestion is that your hypothetical is useless and inapplicable because it is at best, reflective of your own personal misunderstanding, or at worst, outright disingenuous.
When YouTube recommends content to me, the original author did not target that content specifically to me, and YouTube alone is making the decision to put it in front of my eyes.
And would the developer of a Mastodon client be personally liable if their algorithm recommends someone harmful as a new account to follow?
Instead, HN's view would likely become "everyone sees https://news.ycombinator.com/newest and showdead is set to 'yes'"
I'm not sure how to construct an argument that would allow HN's front page while at the same time curtailing YouTube's not signed in front page - both are recommendation algorithms.
I think most of us would notice the difference if HN was allowed only to rank comments only in chronological order, never mind if for liability reasons the stories permitted to appear in order of upvotes on HN were restricted to the ones Dang was satisfied weren't libellous, or possibly none at all if YC decided it wasn't worth the risk
There is moderation as well with the removal of stories but I’m not sure the responsibility is for removing harmful content. If someone posted a slanderous story or other illegal content and it was allowed to stay for some length of time then I think HN would be responsible. The most egregious would be if a child porn story was on the front page for days because HN staff chose to leave it there.
For YouTube, they are suggesting beheading videos to my child and I think bear some responsibility for doing that, and hopefully to stop doing that. They are making editorial decisions to promote content and so, I think, shouldn’t be protected by 230.
For example if I the moderator check the site once a day, and someone posts 5 minutes after I leave, would the law say it's ok for the content to remain up another 23 hours because no moderative choice occurred? Is there now a legal requirement to ensure you moderate fast enough?
So I think there’s a reasonable argument to be made about the difference here. Agree?
Section 230 is a very short section of law, you can read it yourself: https://www.law.cornell.edu/uscode/text/47/230
In any event, it was just an example. Feel free to substitute "HN decides what to show on the homepage" instead.
Therefore it's hard to say what the gmail user could even base their complaint on. Google can do whatever they want with the incomming email as far as categorization and blocking goes, and nothing in the nonexistent ToS is violated.
Many of the problems from social media are rooted in the idea of having such enormous hoards of content that the only way to trawl through it is with automated algorithms. This was great when it worked, but the content pool seems to grow faster than algorithm design can accomodate and will only get worse as AI content generators mature.
It suggests that there are possibly as few as two futures:
1. The internet is a wasteland of content pollution and the automated tools for sorting and sifting through it are overwhelmed with toxic waste.
2. The internet returns to a network of trust where people are individually, but only marginally, accountable for what they share with others and this accountability engenders thoughtful curation at a manageable scale.
It might be that modern equivalents of web rings, curated directories, group chats, and member forums supplant "Internet Scale" search engines and social media networks. It would be an adjustment, but it wouldn't be the end of the world.
Furthermore, distributing liability among your users is not a great idea. What will happen is that extortion enterprises will be created to sue people on our hypothetical old-web-of-trust. We know this because BitTorrent allowed Prenda Law to make porn, share it themselves to dox users[0], and then sue the people who downloaded or watched it. This scheme worked because copyright lawsuits are expensive to defend against, and defamation is no different. So everyone just quickly settled, which is why it took years for judges to catch onto this particular fraudster.
The only thing that keeps you from being sued for watching an infringing YouTube video is DMCA 512, which works almost[1] identically to CDA 230. Because it's a large centralized service, it's a juicier target, and people with legitimate copyright grievances can get things taken down from them. So nobody bothers to try and sue individual viewers.
Your #2 scenario only works if the Supreme Court is merciful and only kills recommendation systems. If CDA 230 is struck down entirely, you won't get to run a network-of-trust version of the web, because just hosting a public web server will require signing an indemnification agreement and posting a very large bond. In this scenario the Internet becomes more like cable, or perhaps a games console.
[0] Normally an IP address is not dox; but a lawyer and an out-of-order DMCA subpoena can turn it into dox.
[1] 512 adds a notice-and-takedown regime because Hollywood wanted censorship powers over the Internet.
3. Some other country tells the US to fuck off and foots a lot of the lost advertizing bill in order to collect a shit ton of user data like the big US sites do now?
It's a world wide web, if the US screws off too much there is no golden rule that says it has to be the monetary king of the internet.
> "We're a court. We really don't know about these things. You know, these are not like the nine greatest experts on the internet," Kagan said of her colleagues, eliciting a laugh from the courtroom gallery. "There's a lot of uncertainty going the way you would have us go, in part, just because of the difficulty of drawing lines in this area and just because of the fact that, once we go with you, all of a sudden we're finding that Google isn't protected. And maybe Congress should want that system, but isn't that something for Congress to do, not the court?"
https://abcnews.go.com/Politics/supreme-court-wrestles-immun...
> Supreme Court Justice Elena Kagan said one could question why Congress provided such immunity when passing Section 230 of the Communications Decency Act of 1996. But she drew laughter when she wondered how far the Supreme Court should go in cutting back such protection.
> “We’re a court. We really don’t know about these things. You know, these are not like the nine greatest experts on the internet,” Kagan said.
> Kavanaugh said Congress knows that lower courts have interpreted the protections broadly. “Isn’t it better ... to put the burden on Congress to change that, and they can consider the implications and make these predictive judgments?” he asked Stewart.
---
Justice Kagan's exchange for that segment is on page 45 https://www.supremecourt.gov/oral_arguments/argument_transcr...
JUSTICE KAGAN: Yeah, so I don't think that a court did it over there, and I think that that's my concern, is I can imagine a world where you're right that none of this stuff gets protection. And, you know, every other industry has to internalize the costs of its conduct. Why is it that the tech industry gets a pass? A little bit unclear.
On the other hand, I mean, we're a court. We really don't know about these things. You know, these are not like the nine greatest experts on the Internet. (Laughter.)
JUSTICE KAGAN: And I don't have to -- I don't have to accept all Ms. Blatt's "the sky is falling" stuff to accept something about, boy, there is a lot of uncertainty about going the way you would have us go, in part, just because of the difficulty of drawing lines in this area and just because of the fact that, once we go with you, all of a sudden we're finding that Google isn't protected. And maybe Congress should want that system, but isn't that something for Congress to do, not the Court?
See for example https://slate.com/news-and-politics/2023/02/brett-kavanaugh-...
Most technical people don't understand DNS, and it's one of the most basic and core technologies to how the Internet works today. The courts definitely don't understand DNS. The vast majority don't understand BGP, and it's literally the basis for how Internet networking works. Many don't understand the Web as a tech stack, outside possibly the basics of website design. Even people who develop web applications don't understand the Web as a tech stack.
It's actually a major concern for me, because the promise that future generations would be "digital natives" with deep technology understanding didn't come to fruition. The knowledge of how all of this infrastructure actually works is dying off, and very few people are interested in learning it, and the entire global economy is now built on the Internet in a myriad of ways. Reminds me of https://xkcd.com/2347/ except extrapolate from open source software to literally all essential technologies the world runs on.
Could you debunk some of the most common/biggest false beliefs you encounter? I'm curious to see which traps I fall into.
1. Routing paths are primarily decided by path distance as the metric.
2. IP ownership is authenticated and therefore the global route table is deterministic and stable.
3. DNS traffic is always over UDP/53
4. DNS is controlled by the network. Related, split horizon DNS is reliable.
5. Domain registrars and the root nameservers are required to be contacted in order to get a DNS response for a given public domain. As a corollary, seizing domains or blocking DNS at the root is a sufficient method to block access to a site. Secondary corollary, a correctly configured and valid public domain is reachable from anywhere on the Internet that can make a DNS query.
A lot of these are true in /some cases/, or in the general case, but not in the corner cases. A lot of policy and court decisions about the Internet assume the happy path or the common case when dealing with situations where motivated actors may take actions outside the happy path to work around that court order. It is entirely possible to bypass public DNS to resolve a public IP on the Internet as long as it routable, regardless of what registrars or public (and root) nameservers say. (see: Pirate Bay and their shenanigans, also see Tor resolvers, see non-authoritative DNS resolvers, see host file sharing, see alternate roots, etc.)
A lot of policies within organizations, including those set by network admins, are based on a belief that systems in their network obey the things provided by their network, when in fact the client controls nearly everything about how it decides to interact with a network, and the network can only provide configuration as a suggestion or recommendation. There are very limited options for enforcing policy at the network level. There is no guarantee that any given client on your network is actually utilizing your local DNS resolvers or successfully will resolve split horizon DNS.
A lot of configurations in the world are based on trust in systems which are inherently designed in an untrustworthy manner, or which require additional configuration and technology to provide an external basis for trust. This has had consequences with far-reaching impact and will continue to do (see BGP hijacking).
DNS, except when using newer protocols and taking extra measures to authenticate responses, is easily intercepted and hijacked. It is entirely possible to capture queries and inject non-authoritative responses, including responses which claim to be authoritative at any point in the path the query traverses. This is actually commonplace in some regions of the world (see: China). DNS that is authenticated and encrypted doesn't traverse UDP/53.
Nothing gives me Gell-Mann amnesia quite like the comments that get upvoted on HN posts relevant to the particular area in which I work...
I'm currently a PhD student doing research related to Tor, and there was a thread about Tor last month. I didn't even know where to start with the comments, so I didn't participate at all.
Even then, without an understanding of how it works technically, it's not possible to responsibly and accurately adjudicate the law as it relates to these technologies. How BGP and DNS work is actually front and center in many court cases, where the decision rested on a basic misunderstanding, which rendered the outcome either ineffective or unjust. Think of all the cases where DNS blocking is utilized on a court order, not realizing that this is effectively meaningless and is not actually effective.
People don't often understand how deep this rabbit hole goes, even technical people who understand the basics of routing. They don't actually understand what a "peering agreement" means, especially what it means when it crosses an international border. The way that the law interacts with the Internet is an integration point that has an especially high level of complexity, and nearly no-one involved in that complexity in our current system day to day is actually qualified to discuss it.
When we have failures in this integration point is has deep reaching impacts that have broad implications for how technology continues to develop into the future, and there are both positive and negative consequences for this. As a simple example, the stupidity with courts and DNS blocking and the way in which some governments around the world behave in relation to DNS is actually a core reason why DoH and DoT exist, because misapplication of law based on misunderstanding of technology risked fundamentally breaking the basic building blocks of the Internet required for it to function and for networks to successfully interoperate, so the people who actually understand how it works had to come up with new technology to "route around" stupidity in legislatures and courts around the world. These new technologies /themselves/ have geopolitical consequences (see UK and EU stupidity regarding crypto and its implications for DoH and DoT).
It's not JUST understanding the technology, there's the 0th layer and the 8th layer in the reference model.
Both the legislature and the judiciary in the US, and in other governments around the world, are technically and scientifically illiterate.
I'm pretty sure that's not what was meant by "understand the Internet", in this context.
Didn't Isaac Asimov write a documentary on this call 'Foundation'?
That was one of the early ideas for Twitter when Musk took over. Could even have external, independent “filter providers”.
Then the user a chose what bubble they want to live in. Or none at all, if they are brave enough to see opposing opinions.
But if the web site decides the algorithm for the user, well, then the web site is the same as a newspaper or TV channel.
Which is probably the reason that particular promise from Musk never actually went anywhere.
There needs to be some accountability. The "slippery slope" or "chilling effect" arguments are valid, but the scale has tipped too far so that individuals being harmed by these companies have little to no recourse.
These companies make billions of dollars in profit and much of that is because they underfund customer support, or simply ignore it all together. It's like any consumer product safety issue. It's expensive, but should be a required cost of doing business, like product recalls or EPA pollution regulations.
How is YouTube different than this? Or Facebook, Twitter, or any other social media that is free to see once you have an internet connection?
TV network made a deal with the government (FCC): they got a monopoly access to a limited resource (spectrum).
In exchange they agreed to limitations set by FCC, like decency standards.
Limitations that go beyond what the law, as created by congress, requires.
YouTube or Facebook didn't make a similar deal with FCC because they use unconstrained resource (internet bandwidth).
Furthermore congress did the right thing and created a law explicitly calling out this scenario and made internet companies not liable for user generated content.
Why did they do it? Because without that there would be madness.
This is also how is has stood throughout American history. The foundation of the US was built on anonymous/pseudonymous pamphlets and secret communications between the "Founding Fathers."
Sites could still allow anonymous communication, but it would need to vet it first and assume liability.
You can't go out in public in a mask and say "John Smith eats worms" (when he doesn't) and not have any repercussions.
Edited to add: It seems like this is what distinguishes a telephone company from a television broadcaster too.
Hacker News does not directly promote content (aside from prioritizing content based on non-content factors [age, voting, etc.]). It does moderate, but that is not the same as promotion.
The crux of the argument in this case is that the plaintiffs want Google held liable for promoting content, not hosting it. The difficulty the Supreme Court seems to have is understanding what the boundaries are between promoting content and simply delivering a usable view of content choices to consumers, and whether Congress intended for that difference to matter with regards to Section 230.
Is Hacker News default feed that promotes highly-upvoted, newer stories to the front of the list and example of HN promoting those stories? If so, should HN be held liable if that algorithm pushes content that harms others?
Of course HN promotes content. There's much more content submitted than can fit on a screen. Automated filtering, human moderation, and signals from other users decide what to show you -- on both sites.
In my language, I say HN does not "promote" content because the view for me is substantially the same as the view for everyone else. Contrast that with YouTube, where Google's algorithms elevate content specific to individual users, thus leading me to conclude that Google thinks I "should" watch the content. I would argue that far more than automated filtering, signals from other users, and human moderation goes into YouTube's recommendations -- particularly a calculation of the revneue to Google of you watching the video.
But again, this concept that Google "promoted" the content that harmed the plaintiffs and should thus be liable for that action is the heart of the case, not that Google should be liable for the content itself.
Fundamentally any decisions a site makes to filter and sort content to show, including HN ordering by vote count and mixing in new content to allow it to make the top page, is an explicit choice they are making that cannot be differentiated from "promotion"
I don't want to respond to part of your comment and not the other, so I'll just say: I don't know, because exactly defining the specifics is not my goal.
> Fundamentally any decisions a site makes to filter and sort content to show, including HN ordering by vote count and mixing in new content to allow it to make the top page, is an explicit choice they are making that cannot be differentiated from "promotion"
This is the part of your comment that actually matters to me. This statement is so definitive, and yet there are people arguing just as definitively that some ways of prioritizing content for users create liability while others do not. I think there is a difference between "elevating when it otherwise wouldn't be elevated" and "providing a moderated list", but those two states are probably separated by a grey area, not a bright line. I think how YouTube identifies content for users is distinguishable from what HN does, but I also don't know if that difference matters with regard to liability, especially with regards to Section 230, which makes no attempt to legislate HOW content is made discoverable.
There's no natural state of how content would be displayed, any choice of how to do it would result in a moderated list that elevates something that wouldn't have otherwise been elevated with a different approach.
There is no distinction to me between what YouTube and HN do, and I certainly don't think the law should treat them any differently. Both should be legally protected regardless of which specific approach it takes.
Simple orderings like chronological are not considered moderated because the content doesn’t define the order.
Exactly defining specifics is kind of important in having workable laws.
The law can and does differentiate across lines like these even if they're both technically "promotion algorithms".
This, by many of the same arguments people say about Youtube, would absolutely constitute a recommendation. Even if it's primarily user driven, HN is responsible for the synthesis of all these values which results in content appearing on the front page and tuning which content appears or is downweighted. You can try to play word games but the reality is that should Section 230 go, no one's going to risk the cost of lawsuits to discover what the limits of promoted content / algorithms is.
But you're living in a fantasy world if you think that I, personally, could ever recover meaningful damages from somebody with enough reach to meaningfully defame me. Rather, defamation laws are being successfully (ab)used to silence politically inconvenient people like Alex Jones.
>Rather, defamation laws are being successfully (ab)used to silence politically inconvenient people like Alex Jones.
LOL. Strong disagree there. He's not being silenced by any means. He's free to engage in whatever speech he wants, be it defamatory. He only need pay for the damages he causes in doing so. I cannot imagine what's more libertarian than that.
I'd just defame him right back.
All section 230 does is make clear that if you don't have people signing off on each and every bit of content on your platform, but respond with a good effort to reports of illegal content, then you will not be held responsible for someone using your platform as a way to do illegal things.
If all these weird people that misunderstand this get what they want, it doesn't mean youtube recommends their conservative creators more, what it means is that youtube stops doing any personalized recommendations, and all you are going to see on youtube is Logan Paul, Mr Beast, and the other mega creators that little children love.
> All section 230 does is make clear that if you don't have people signing off on each and every bit of content on your platform, but respond with a good effort to reports of illegal content, then you will not be held responsible for someone using your platform as a way to do illegal things.
Web hosts that just serve content when a client requests it should definitely have this protection. Great idea. YouTube promoting outrage bait because it makes them more money? God no, no protection beyond what any other business doing that would have, not on the Internet. The line's somewhere between those.
None that have been made with a straight face before a real court, that I'm aware of. That's the spin that the activists put on this, not something that anyone thinks SCOTUS is going to rule on.
You're absolutely right that 230 as written doesn't really speak well to the modern semi-automated echo chamber. But improperly written laws are congress's job to fix, not the courts. Courts step in when laws conflict, they aren't there to figure out how to solve problems with new laws.
The simplest way to look at the spirit of this law is: Congress said that internet companies shouldn't be punished just for hosting other people's opinions. And at the end of the day, TikTok and YouTube and Facebook are still just hosting this data. They didn't write it. They don't curate it. Anyone can post. Anyone can read.
Arguments about "recommendation algorithms" are legitimate, but not really in scope of the first amendment and liability issues envisioned by the original law. They're just not. It's not something congress thought about. And if congress didn't have a clue, why should the courts?
Section 230 does not speak about this because section 230 wasn't supposed to care about this. Section 230 was entirely about protecting large companies from legal harm when, say, the christchurch shooter posts their spree on your platform, or someone uploads literal child porn to your platform, as long as you attempt to remove such content when it appears. Section 230 is simply a legal admission that moderation is hard but necessary, and a single failure should not doom your platform.
If we think the way large platforms recommend content is harmful, we should write legislation for that, instead of yet again fucking with a working law to hammer it into something it wasn't meant to do. The USA loves to do that and all it does is give us really shitty legislation.
Okay - so in your universe, it's just as easy to punish them, not for the hosting, but the promotion of the content which is explicitly not covered in 230...
Care to share any of that sense?
This mirrors pretty well how the rest of the world operates. If I walk into a Macy's and start screaming libel or inciting a riot, the Macy's Corporation is not liable for that speech. Why should a 'digital property' like a social media site work differently?
Of course, it doesn't totally erase the server owner's liability. If someone posts child pornography that my server displays, it is my responsibility to remove it as soon as I become aware of it, just as it is Macy's responsibility to escort the libel-screamer off the premises. Failure to do that can confer liability onto me.
To break this consensus is to break the ability to foster human interactions on the internet that mirror how humans interact in the real world.