Safe harbor rules are what makes the internet of today work.

Today, content isn't just generated from large, powerful, corporations... it's generated from us. For example, Hacker News is protected from any liability this comment might have due to safe harbor rules. It's the reason YouTube can exist. It's the reason why blogging services can exist. It's the reason Twitter can exist (you can still argue whether or not it should though... ). The point of safe harbor is to assign the liability for the content to the creator, not the service that publishes it. The company that serves the content is only liable after they've been notified. This is critical for today's internet.

Any company that allows users to post content (messages, comments, videos, anything) must have safe harbor protections to operate. If they didn't, then they would be responsible for any liability for everything, so they'd have to moderate every single comment, every single video, etc... Basically, this would shut down the user-supplied content that rules the modern internet. What we'd be left with is a computerized version of the one-way communications of old-media.

That would be a bad thing.

Also, you're unnecessarily extending safe harbor protections to where they don't exist. For example, Google Maps and privacy. This is Google's content. They created it. They published it. And they are absolutely responsible for making sure that they are compliant with any privacy regulations. This doesn't fall under safe harbor at all.

Sure companies make mistakes with privacy, your example of Google Buzz is a good one. However, don't confuse these issues with safe harbor. In this specific case in Italy, it is safe harbor that is under attack. And that's a bad thing for Italy.

About aggregation sites like Hacker News, you do have a strong point, though.

[1] I mean as easy as Facebook, Blogger, YouTube, Twitter… Any harder and it won't work.

I agree... :) it would be nice for all of us to be able to "publish" ourselves... but for the most part, even if we did have better low-power home servers and enough bandwidth to serve the content, most people won't be able / willing to set it up. So, I think in general, the mass-public will be stuck with the Bloggers and YouTubes of the world.

When you buy hosting online, you buy more than space on a hard disk.

I don't think most people really need high availability, so UPS and redundant connections sound like overkill. For fail-safe hardware, we just need something "unplug proof". A Beagle Board (with a few GB of Flash and a small battery) is more than enough.

I am aware that such a solution will be a little less convenient than storing everything "in the cloud", no matter how easy it can be. But the cloud has its downsides as well: you hardly have any privacy in the cloud, and they can shut everyone down, for any reason.

As for the difficulty of running things without safe harbour provisions, people are quick to assume that there are no or only limited alternatives, and that without the current scheme we would somehow "kill the Internet", but I challenge that assumption, based on the argument that follows.

Firstly, if there's one thing the Internet has managed to do successfully for a long time, it is evolving creative solutions to large scale problems. The Web started as a small network of pages, with only a few carefully chosen links between them. Then we evolved web rings to help people find related content. After that, ever-improving search engines became the dominant way to find material. Today we have link aggregators, status updates/tweets, etc. There is no reason to believe that removing one form of expression would make things worse; on the contrary, if history is anything to go by, it might well catalyse the development of a better replacement.

On which note, I think we are going to move towards "Web 3.0" where user-generated content is balanced with editorial control by real people anyway. Web 2.0 was an interesting experiment, but the signal-to-noise ratio is terrible, the volume of information is overwhelming, and the simple, user-supported voting systems of today are widely abused and not really up to the job. Even the poster child, Wikipedia, now has policies that prevent open contributions to certain articles, internal politics among the volunteer editors about what should be allowed, etc. I think the kind of purely automated content hosting sites with unlimited user generated content are already dead, they just don't know it yet.

Next is a key point that I imagine some here may disagree with: I have never believed in absolute freedom of speech (of the kind where you can say whatever you like with no responsibility for any consequences). Words are very powerful things, and in the Internet era that power is magnified many times over. I believe that people should be held responsible for something they say that is unreasonably damaging to others, whether it be betraying a confidence, starting a malicious rumour, giving misleading "expert" advice in a field such as health or finance... This is just common decency and courtesy, and indeed almost nowhere has really had freedom of speech in law for many years, because of issues like defamation, national security, court-ordered anonymity for victims of certain crimes, etc.

These things have been part of the law in many jurisdictions since long before the Internet. The difference is that on the Internet, there is no such thing as an off-hand comment quickly forgotten. Society needs to adapt to this new reality, appreciating the benefits of giving everyone the power to speak to millions of others, but imposing suitably harsh penalties for those who use that power irresponsibly.

I just don't see how that is compatible with broad measures like absolute freedom of speech or complete safe harbour rules for those republishing material to potentially huge numbers of people. Those are merely convenient blanket laws designed for ease of implementation, not because they reflect the ethical thing to do.

So I'm not convinced that a law that imposed responsibility and as a side-effect made it difficult to run unlimited user-generated content sites would cause the loss of something truly valuable. If you want to write a dozen spam comments on a blog about how hot Hayden looked on Heroes last night, you are welcome to do it on your own personal web site, where your deeply-considered and original thoughts will no doubt receive the attention they deserve. And if you say something unfair in your next blog post that destroys someone's career, you can look forward to a personal letter telling you when to turn up in court and defend that action or be held accountable.

Meanwhile, I don't see the kinds of rules I'm talking about halting the development of the Web, because I think we're going to move toward more editorial control on the big, popular sites for other reasons anyway. And even if you do believe in absolute freedom of speech, you can still publish what you want on your personal web site, you would just have to take personal responsibility for it accordingly instead of hiding semi-anonymously behind some giant organisation who will republish your material but shield you from being identified and thus held responsible for the consequences. I, for one, don't have a problem with that.

This fact is indifferent to whether you use web-rings, NNTP, your own "personal website," or YouTube to post content.

How about your local web hosting company, which " stores persistent data from one party and republishes it to arbitrary others" ?

How about your ISP, running a caching proxy such as Squid which " stores persistent data from one party and republishes it to arbitrary others"?

There are a lot of independent entities out there storing, publishing, and forwarding your content. Most make fractions of a cent on each "piece of data" - photo, video, blog, whatever that flows through their system. They cannot commercially afford to screen it all, or even sample it.

Can you read every single article on the new page of Hacker news? 24/7? How about on Digg?

Why do you expect someone else to?

I'm not saying there should be no provisions in law to support the effective running of the Internet. I've never said that. I'm just saying that companies who want to establish a certain business model on the Internet shouldn't get a free pass just because it is difficult to run a business with that model while still complying with the same laws as everyone else. If that means some companies cannot continue, so be it: as I said before, I don't think anything of significant value will ultimately be lost, and I would rather that than effectively legislate certain businesses above the law just because they can't work out how to do things legally otherwise.

Your final comments are a straw man. Hacker News and Digg aren't republishing those articles, they're just linking to other sites that do.

Also, I am not certain this would be a good thing to always require content to have someone be accountable for it. If I want to express, say, a dissenting political opinion, I may want to be anonymous, and I may not want to pass the burden of accountability to someone else.

This is worth reflecting on it, however.

I am certainly not saying that no protected speech should exist. Indeed, I am all in favour of a law that protects certain classes of speech, with political views probably the most important class.

I just think that such laws should be crafted carefully, striking a balance between freedom of expression and protecting people from the harm when others abuse that freedom, and that once such laws have been made, they should apply on the Internet as much as anywhere else.

There are complications with jurisdiction in the on-line world, but there is no reason that most of the international community can't reach a consensus on these issues, just as they have on many others before.

I have always struggled with the idea of anonymity as a vehicle for free speech, for three reasons. Firstly, most people who think they are anonymous on-line really aren't, if someone tries hard enough to identify them, so it is often an illusion. Secondly, actively protecting anonymity automatically removes any responsibility from the speaker, whether or not what they are saying is within a protected area, creating a huge loophole in the laws. Finally, while anonymity may have a perceived value in protecting those opposing an undemocratic government, we don't have that situation anywhere in the west (to the point where political dissidents are routinely threatened or "disappeared"), and if we ever reach that point again, the correct response will be one of the three boxes after "soap".

Firstly, if you need Wikileaks in the first place, you have bigger problems.

Secondly, Wikileaks actively tries to place itself above the law. No-one should be above the law.

Thirdly, there is little that has been revealed via Wikileaks that could not have been revealed in the traditional way via a free press. Wikileaks may make things marginally easier, but if you're in the business of leaking private stuff only if it's easy, maybe you should reconsider your world view.

Finally, before anyone comes along and tells me how much good Wikileaks does, consider this: they also released the private membership list of an unpopular political party, causing very serious consequences for many members of that party. Whether or not you agree with their politics, that sort of action is way over the line. What about the anonymity of those party members?

If we carry out this reasoning to its ultimate conclusion, platforms like YouTube or Blogger won't last long. It would indeed end the web as we know it. It will not end free speech on the internet, though. Not by a long shot.

Free speech can be achieved through two means: the out of jail free card, or a way for people to publish themselves.

Self publication is possible: we just need personal web servers, symmetric bandwidth, and the promise that we won't sue any ISP for moving bits (we need that safe harbour). Of course, companies like Google or Facebook prefer the out of jail free card (for many reasons, they prefer your data to be on their servers rather than on your own). I think the people will rather take the self publication route, given the choice.

Oops! The search provider indexed it, and is serving it out to the world in the form of results. By duplicating your data without reading it, they're responsible. Jail for them.

Youtube and Blogger are a miniscule step beyond indexing. They're duplicating data you wrote on your computer, and serving it out to the world.

Content hosts cannot logically be responsible for user-submitted content. This leaves two outs: safe harbor, or only commercial content on the internet, so the buck can be passed (note that this is identical to a TOS). Remind me again where free speech fits in there?

edit: as I said in another comment: you don't put the papermakers in jail because someone wrote something mean on it.

A publishing platform can choose to be neutral (no power), or it can choose not to be. Of course safe harbour should apply to neutral media. For those that are not, however, I am not so sure. Almost, but not totally.

About the paper makers, you missed another point: anonymity. It should be very easy to reach the actual author of a paper, so this makes no sense to sue the paper maker. If the paper maker published all this stuff anonymously, and can't (or refuse to) disclose the author's name, however, that's a different story.

More-so than many other content-hosts, Google did act responsibly. If anything, Youtube is a medium just like the paper, but Youtube responds more rapidly and identifies criminals using the system uniquely, which is something paper cannot do.

Shall we eliminate all paper? It's easier to draw kiddie-porn on paper and post it up all over than it is to get away with it on Youtube.

Obviously, harboring criminals is different than helping an investigation to find the criminals. But without a takedown notice being ignored, there's no harboring, because there's no effective warrant. And without a legal takedown notice, there's no responsibility to remove anything, because to do otherwise would encourage people to restrict free speech just by complaining loudly.

Papermakers have the same measure of responsibility in this as Google. They can restrict selling paper to anyone until they see and approve the use of it, or stop selling altogether, or they can have a safe-harbor because they're merely creating a medium. If they sold paper to a book company that went on to write something which later got banned, they'd be in exactly the same position as Google, but nobody is saying we should all make paper ourselves, and affix our name and address to all the paper we make (running a server is easy to trace, telling who made paper isn't, so this is needed to level the playing field). And, as nobody can guarantee that that name and address are genuine on the paper, it cannot prevent anonymity, and loses significantly to a web host like Google who can point you to the source of everything they find.

Please, everyone, stop buying paper because people can abuse it. In fact, stop buying water because someone could use it to drown someone else, and the water company isn't doing anything to prevent this.