The continuum of action discussed yesterday ranged from hands-off to intervening. For many, the hands-off approach is unpalatable when one contemplates the most harmful content being dropped on the IC. Others fear intervention opens the pandoras box to influence by external entities and unsustainable bureaucracy.
We listen to the arguments of others with our own filters. What I thought was an expression of a libertarian perspective may have been an impassioned defence of social scalability. (Unenumerated: Money, blockchains, and social scalability).
Pathfinding how and how far any of such will fit into existing traditional jurisdictions is a non-trivial exercise we all have still to figure out. There is no single juridical system on this planet that a-priori trumps all others.
Yet, there are pluralistic liberal values that many (most?) of us want to see reflected in the way this whole thing is governed. This specific case is but (one of) the first nudges for the involved community and perhaps local countries legal frameworks to put effort into governing such things âright and justâ in ways where there is little to no precedent on doing it globally.
Perhaps down the road there is a case for forming an WHO/WTO like governance organisation that decides and articulates on legal principles for interfacing with on-chain governance rules and rule changes that local legislation commits to reflect, adhere to or respect.
For the moment, the DFINITY Foundation necessarily holds dominant voting power to be in a position to if necessary undemocratically correct unfolding developments when they go wrong. Abstaining from voting in this particular case is an important step to mature the communities perspective on not only the particular the problem at hand, but the class of problems where there are no off-the-shelf answers available for yet in a global, de-centralised system that down the road will have to interface with local jurisdictions around the globe one way or the other. It will happen by design, by accident, depending on the local legal system by decree, precedence or derived from roman law principles.
Letting us all figure this out in this specific (first) case by ourselves like teaching a kid you raise to own the consequence of its own decision is what we as a community are going through with this initial conversation. It is massively relevant to our growing up and mature as a community that over time will entirely have to self-govern itself through the NNS. As as with kids, many of the ideas and thoughts coming up are still understandably immature and rooted in the limited perspective of the world we know until now, not the world that we are growing up towards, yet.
Where is the 21st century Plato writing our next gen âPoliteiaâ when you need her?
I hope I donât come across as too bitter here, but having spent the last 2 years of my life working 12+ hour days on the IC (alongside people who have been doing it for much longer) I find many of the opinions expressed here incredibly entitled.
Yes, the IC should be able to eventually do pretty much all that the above posters expect from it. Eventually. (Whether much of it is a good idea or not, is a completely separate issue.) But right now we donât even have the possibility to move a canister from one subnet to another; or run within encrypted enclaves; or allow node providers any freedom to choose what should and should not run on their nodes. All of the above above and more is still being worked on at a rate of well over 12 hours per person per day. And honest-to-God efforts are being made to allow the community to actively contribute to it all (via proposals, code and whatever else).
But as many have already pointed out, attitudes like âif the IC doesnât do this now, then it is dead to meâ or âthis is no different from AWS, except slowerâ, with an implied expectation that things need to change here and now to allow for unfettered freedom on the IC are, to put it mildly, unhelpful.
There are things that can be done today (e.g. decide how to handle obvious IP infringement, so there isnât a big argument every time it happens). And there are things that can be considered for the future and introduced gradually, as protocol evolution allows. Itâs just reality, not any oligarchic plot to turn the IC into Web 2.0 on steroids. E.g. there is no reasonable way (given the currently implemented capabilities of the IC) for a node operator to protect themselves against a legal DMCA request that is not acted on by the NNS. And the community (via the NNS) making a public statement that IP is old school and the IC is alegal is not going to help their argument.
A few posts later in this thread you indicate that you will exit participation in this debate. I would like to encourage you to remain engaged. You bring a valuable perspective. Iâm not sure I agree with everything you have written yet, but this particular comment was a turning point for me in understanding your perspective. I think you are bringing intellectually honest contributions that need to be part of the conversation.
i think the idea of allowing canisters to move and perhaps establishing a free bid/ask market for canisters and node operators to participate in would be ideal. in that case a node operator in a jurisdiction where they could not be sued, or where the possible liability is very limited could bid low prices on âbadâ canisters and ones which are in other jurisdictions could put in higher bids. canisters could just run on any node initially, and with the ability of the node operator to kick them to the market if and when they receive some legal notice and make a business decision that they donât want to deal with that particular client.
for example, canisters which offer gambling would want to be hosted on nodes in gambling friendly jurisdictions.
free market is the way to go on this jmho. of course if the canister is running something that is an assassination market itâs possible that no node op would place an offer on the market and the canister would be SOL.
@ajismyid , what about Sci-hub ? For me, it is a major example of revolutionary content that The IC should host and stand for. It is about Science made available to everyone, with no cost-barrier. It has been revolutionary in web2.0, and we should let it be revolutionary in web3.0. If we bend to big-corps, Scihub would never exist in The IC. Is that what we want? ( Just for context, scientists are usually supported by public money. Scientific results go published in Refereed Journals. Scientists act as invited referees (and gain no money for that). Scientists have to find ways to get funding and pay for publication fees (that goes from 100-200 $USD per page). After the work is published, the editor charge readers a fee of 30-100 $USD per paper! Can you imagine that? Fee over fee. And if you canât pay, you canât read and access the scientific results (even though it has been financed -mostly- with public money).
Scihub is revolutionary because it goes against that. It takes the scientific papers, hosts them in stealth servers, and anyone can get any scientific work, for free. It has created a huge counter-move from scientific editors worldwide. But still, the -editor industry- want to keep enforcing the old model. Charge for publishing, and charge for reading (while using established scientists as free referee-labor-force).
The IC could host Scihub. And editors would try to put down our Scihub canisters. Are we going to let that happen?
Thanks for your perspective! Itâs way too easy to get hung up on ideological views while forgetting about the reality we are facing right now. Good to have this vision of ultimate freedom but important to keep in mind that it will be a long road and trying to take the path of most resistance is unlikely to get us there.
On a side note: My understanding is that the NNS, being pretty much the shareholders voice, is exactly meant to be able to make decisions like that, where the success of the project will most likely be the common denominator and decisions are made accordingly. Itâs impossible to please everybody, but as long as we can agree, through voting, what are the best decisions for the IC to succeed, I donât really see an issue here.
On the contrary, Dfinity announcing they wonât vote and let other shareholders/participants make up their minds was a huge step in showing everybody that they are ready to listen to everyone regardless of what they think/know to be best for the overall success of the IC. So more than everything Iâm encouraged we are on the right path here. keep up the good work!
Once again, we can satisfy everyone with varying content moderation of subnets. Some subnets could remove IP infringements, some just remove CP, some remove nothing.
We could have âModeration Groupsâ that consist of moderator neurons and subnets they moderate. Each group posts an acceptable use policy (or none).
Then we let the node providers decide what legal risk to take based on their location (and morals). @Ciaran can run his moderation-free nodes from a boat in international waters where he is free from prison risk if someone uploads CP.
@Ciaran said âThatâs exactly what people said about Bitcoin in 2012.â
No, the IC is entirely different because of its capabilities. It is not economically feasible to host content more than a few MB on existing chains. The IC opens up much broader possibilities of copyright infringement and morally controversial content. And because the nodes need data centers, it is much easier for people to point fingers at node providers as âresponsibleâ for the content they serve to clients. At the end of the day it comes down to what a non-technical judge/jury can be convinced of.
However, in theory, Bitcoin could face the same problem. Someone could spend millions uploading CP to the Bitcoin blockchain that would be then hosted on every node for people to download. If this happened, governments would be faced with an outraged populace demanding a solution. It could end up with legal consequences for Bitcoin node operators, effectively banning nodes in many countries, and severely degrading the performance of the network and disenfranchising many users. We would likely see a fork of Bitcoin that removed the offending blocks gain popularity, a form of meta-governance. Amidst all the chaos, networks that have built-in governance to avoid such disruption will thrive.
In my opinion only canisters that host content that can endanger other peopleâs lives such as: hitman services, pedopornographic content, weapon marketplaces, etc⌠should be taken down, not only due to moral reasons but also cause the infraction is evident, a picture either is CP or isnât.
DMCA on the other hand is much more complex and thats why claims are escalated in court if necessary, the NNS doesnt have the manpower to handle the whole internetâs DMCA claims, just a few days ago a YT channel had 120 videos wrongfully taken down due to DMCA, I dont want to experience this on web 3.0 and Iâd rather have thousand of pirated content hosted on the IC than IP owners abusing their rights.
Its clear tho that as of now if things worked like I wish node operators would have legals issues and the IC would suffer as a result, so I think the foundation should prioritize systems to remove legal liability from node operators.
Iâve only walked away as I donât feel like Iâve anything more I can say on the issue really. Iâm just absolutely convinced - from history (liberty reserve, e-gold) and present day actions against governed DeFi apps - that the only way to provide a Web 3.0 platform that can actually function as a Web 3.0 platform is by being autonomous and pushing politics up to the nation state and app layers.
The protocol, like HTTP, needs to remain completely agnostic to who uses it or what itâs used for, or it will otherwise be sucked back into a Web 2.0 structure. If you have the power to remove a copyrighted piece of work, you have the power to alter account balances. If you have the power to alter account balances, even if you donât use that power, youâre a money transmitter in every jurisdiction on the planet. Youâd almost certainly also be a publisher because you can curate content - note thereâs a movement to revoke Section 230 in the US right now. If youâre registered as a money transmitter and also deemed a publisher all the clever cryptography and consenus mechanisms you invented donât matter a jot. Nation state rules rather than economic consensus rules the roost.
This situation simply does not apply to Bitcoin, Ethereum, TOR and BitTorrent and itâs precisely why theyâve survived outside of regulated structures. States treat them like the treat HTTP: Something that just is so they pursue their policy goals in ways that donât involve tampering with how they work. This is the difference between decentralised protocols and those that merely have a veil of decentralisation.
As I said Iâm not sure what else I can say to convince people of this reality but it is the reality.
Actually thereâs one more thing I could say that I havenât. Apart from opening up a pathway for governments to override the consensus mechanism to the point it becomes meaningless, itâs not as if active censorship would materially affect (for example) the amount of child porn on The IC. Itâd be whack-a-mole just like it is now. The powers that be will find it much more fruitful operating like they do now and setting up honeypots (L.E. actually uploads CP to catch criminals remember) to catch these scumbags.
When somebody uploads CP to the internet now, nobody goes âlook what youâve done Tim Berners-Lee! Youâre letting paedos get away with this!â They correctly blame the people that did it in the knowledge TBLâs invention is general purpose. I strongly advocate decentralising The IC to the point itâs completely agnostic and people view it like they view HTTP and itâs a given that criminals are pursued outside the protocol. CP wonât disappear. It never will. But it will be pushed off the mainstream apps on The IC (which I assume will be moderated by DAOs) and into the dark corners where 99.9% of people donât frequent. By pushing the scumbags to those corners, L.E. can isolate them and have success in catching them.
I get the impression many here think Iâm some crazy libertarian. Iâm not at all. Iâm merely a pragmatist that understands thereâs only two ways a network can be administered. 1. Autonomous constraints on power (Web 3.0) or 2. By plugging into nation state rules with its separation of powers and legal systems that also constrain power. I think itâs a fantasy to think thereâs a middle ground where you can choose to follow 1 some of the time and 2 some of the time.
Appreciate all the work you have done and are doing. The IC is undoubtedly a technical marvel, albeit one rapidly developing rather than available in anything like a finished form. I can understand why some of the discussion might have felt unfair to you and your colleagues. However, we on the outside have been exposed less to the innards of Dfinityâs functioning than to the way it has been marketed. A central plank of that marketing has been that, unlike AWS, it is censorship resistant. While the present case might not seem related to censorship, it actually is, because a barrage of DMCA demands are commonly used as a silencing tactic by authoritarian regimes. Dfinityâs response to the notice seemed immature: the idea that we immediately go for a vote without getting details of the complaint. Obviously, we canât have such votes each time some node providers gets a DMCA demand. The knee-jerk response suggested the organisation had given insufficient thought to what is bound to be a common stumbling block for any expanding network. That, in turn, left the entire claim of being censorship-resistant in question. Maybe thatâs a very entitled viewpoint, but I and most others staked a portion of our assets in the belief that Dfinity had thought these things through thoroughly, and I have seen little evidence of that being the case during the ongoing discussion.
I gather from some of the quotes that this is partly aimed at me. Iâve made no demands for technical fixes to be magically conjured up. Iâve said nothing derogatory about the technical work thatâs been done, I have in fact only complimented it.
Iâm taking aim at a culture, a philosophy that thinks itâs a viable strategy to actively moderate The IC ledger. This culture is an existential threat to the technical achievements you guys have made. I think the philosophy is delusional and the protocol will simply be forced to devolve back into a Web 2.0 structure. Thatâs what I mean by âhave fun reinventing Web 2.0â. Itâs a not an attack on developers.
Having read that blogpost that Dominic wrote about The NNS, I do see that that culture hasnât come from nowhere though. So Iâm assuming there is widespread buy in to this culture at the highest levels in Dfinity. Iâm merely here exercising my voice trying to engender a shift in thought. Iâm sorry if you personally feel attacked.
Yes, because on the Web, it is perfectly possible to remove such contents from a web site again, such that it can no longer be accessed. You are arguing for depriving the IC of that possibility entirely, which would create a completely different situation.
This is super insightfulâŚthanks for coming back and contributing. I think you will find consensus that what you are describing is an IC that we all want to existâŚas @free has pointed outâŚwe arenât there yet. Letâs keep moving forward. I want an enclave-base, node-blinded, with node shuffling that âjust worksâ. If we all push forward and build together we will get there.
The caveat here is enterprise and how quickly we want to penetrate that market. They say they wonât deploy to an environment that can host CP, but yet they are all deployed on the Internet. I think this example backs your point. Until we can put the pure P in ICP, if we want to make inroads into the broader market we may have to do some specific governance.
âWhen somebody uploads CP to the internet now, nobody goes âlook what youâve done Tim Berners-Lee! Youâre letting paedos get away with this!â They correctly blame the people that did itâ
OK. I challenge you to stand up a public file sharing server in North America or Europe, advertise it to the world, wait for someone up upload CP, and keep the server running for as long as possible. Are you sure you wonât get any blame by criminal justice systems? If we force node providers into this dilemma, we wont have any node providers.
Just imagine a node provider in court.
Judge: âSo, I can open your server in my web browser, and your server, which hosts CP, will send CP to me?â
NP: âYes, but sir its a decentralized networkâ
Judge: âBut you have the option not to run the node, correct?â
Thanks but my big worry from hearing attitudes on this forum is thatâs not where we all want to get to. I also worry that if we introduce short term moderation while we wait on technical improvements to protect nodes weâll never let go of it. âThereâs nothing more permanent than a a temporary government programâ ,etc, etc.
Sigh, this is why I originally gave up. I keep getting strawmanned.
âŚIâm obviously talking about structuring and decentralising the network to the point node providers arenât put in this position. As a member of the team has chimed in to state, this is going to happen.
Yeah, I wish we could have complete plausible deniability and not worry about node providers legal risk.
But I havent seen a proposal that completely gives plausible deniability. Even with secure enclaves and shuffling, if data is publicly accessible, nodes can find out they are hosting it (network traffic, logs, etc). We dont know whether courts will absolve nodes of responsibility with those hurdles. And it will be a long time before shuffling and secure enclaves are implemented. In the meantime nodes bear the full risk.
I would like to know what you think about having per-subnet content moderation and give node providers the choice of which ones to host, rather than forcing node providers into an all-or-nothing situation. If content moderation devolves into a shitstorm as you predict, the free market will run all the apps in the non-moderated subnets. If the ability to locate nodes in US data centers is crucially important, apps will trade censor risk for latency.
