Poll: Should MaidSafe implement PtP (Pay the Provider)?

Features
688

I’m extremely happy to hear that people prefer quality over quantity or what’s “fresh” (that was a way of saying “hip” in the 90’s) but it is entirely subjective. Could get philosophical on it really. I smell a weird analogy…

Imagine junk content is a high sugar fruit that gives you more of a boost. Cultural works are another fruit that isn’t really sweet to most but far more nutrient rich. The junk content might well win at first or maybe completely (look at bananas), it’s sweet and gives a hit of dopamine, feels good man.
The nutrient rich fruit sustains life and those wise enough to eat it will likely protect its existence.
Though going further in a banana analogy just shows how poorly things could go. Monocultures are bad, usually bad for soil health, risk of that plant going extinct from disease (as opposed to having biodiversity), etc. relate this to echo chambers, extremism, violence from consuming nothing but junk.
In this analogy though PtP is just a pinch of sugar. It adds an equal amount of sweetness proportional to each fruit, so really no perceivable difference between the two, it can’t make people choose one or the other, it can’t ban one fruit to make people stop eating junk, that’s up to the people or how these fruit are sourced or brought to market. It’s up to those whom consume to decide what to consume and since they were equally sweetened it isn’t much different besides maybe speeding up the pace of this experiments conclusion.

If you stuck with me through that, bless your heart, you deserve a pinch of sugar!

I think the only compelling argument I’ve heard so far is being able to load content without ones knowledge of consumption, or if someone is researching or investigating something vile and it is at the same time (granted very marginally) rewarding the vile human, and you can’t retract that. That bit seems unfair to me on many levels but it’s also something that should be considered and then worked out because if people were like “yeah but you can’t build a decentralized internet because blockchain is too slow and there’s no other way”, then we wouldn’t be here…

I know Safe predates blockchain but you get my point.

1 Like
689

Now all I have is a hunch but I don’t think any crypto currency will be allowed save for bitcoin. They will even ban arcade tokens. As the world has gotten smaller the states have wanted to shore up the holes between states. They won’t share their currency space any more than a corporation is apt to share control over its branding or any more than a state would delegate its law making authority (ALEC is small but lame counter example,) they simply won’t do it. They already have the one world currency in the crypto SDRs which state banks use to settle international trade balances. They also have crypto fiats. They will take tech like what SAFE is developing for that purpose and integrate it. And can’t even rely upon them to credit or pay for IP on this issue. Why bitcoin? Because its their fake psychological equivalent of a redlight zone
but has been designed as a state controlled complement. What David said about our intrinsic inability to audit has never been more true (I’d guess.)

690

The conclusion I’ve come to is that rewards should be tied to performance of work required for the network to function. As I see it, that’s farming rewards to vaults retrieving requested data.

Sites/apps that add utility to the data and network as a whole would also qualify assuming the reward mechanism can’t be exploited.

691

If I understand correctly, the goal is to reward creators, since they will always get the first shot at uploading their content to the network. The fact that others will be rewarded for content they didn’t create is incidental and not all that important.

Plenty of people have criticized the currently proposed model(s), so I’ll instead focus on the principle of rewarding creation. I believe no one should be rewarded merely for creating something original; creation is not inherently good or bad, desirable or undesirable. That’s not what’s being proposed here, however, since attention is also a factor. Rewarding content for the attention it receives has the exact same flaw though. Just because something gets attention does not mean it deserves to be rewarded.

I believe the general argument then is that PtP will do more good than harm; it would be a band-aid to help alleviate some of the flaws in society caused by our current political/economic systems. A noble goal perhaps, but I really don’t think rewarding original content that gets attention is the right solution. If those willing to support your creative work can’t afford to pay with anything other than their attention, something is very wrong with the society you live in. Which may very well be the case, but using the Safe Network as a band-aid just adds complexity and obfuscates the real problems.

If someone appreciates what you have created and/or want to support you in creating something new, then let them actively make that choice. That is something the Safe Network can already help with by making it easier (built-in currency) and cheaper (lower transaction costs and fees) compared to, say, Patreon and Kickstarter. I would much rather see recurring payments and escrow implemented, though perhaps not on such a low level.

The only way I see PtP as perhaps a net benefit is if the goal is to promote the network and make it more competitive. Even then, I would consider it a necessary evil at best. Perhaps with proof of personhood (identity corroboration through vouches in a Web of Trust-style system?), people could actively vote for the content they want to see rewarded and the network could use that information to best distribute rewards. Hopefully with zero knowledge proofs or whatever is required to preserve anonymity.

2 Likes
692

Today’s post from Vitalik couldn’t be more à propos. Please read the whole thing if you can spare 5 minutes. Some choice quotes:

the organisms that are the Bitcoin and Ethereum ecosystems are capable of summoning up billions of dollars of capital, but have strange and hard-to-understand restrictions on where that capital can go.

The powerful social force that is creating this effect is worth understanding. As we are going to see, it’s also the same social force behind why the Ethereum ecosystem is capable of summoning up these resources in the first place (and the technologically near-identical Ethereum Classic is not). It’s also a social force that is key to helping a chain recover from a 51% attack. And it’s a social force that underlies all sorts of extremely powerful mechanisms far beyond the blockchain space. For reasons that will be clear in the upcoming sections, I will give this powerful social force a name: legitimacy.

A major power of cryptocurrency (and other digital assets such as domain names, virtual land and NFTs) is that it allows communities to summon up large amounts of capital without any individual person needing to personally donate that capital. However, this capital is constrained by conceptions of legitimacy: you cannot simply allocate it to a centralized team without compromising on what makes it valuable.

It’s too risky to start doing public goods funding by printing tokens at the base layer. Fortunately, however, Ethereum has a very rich application-layer ecosystem, where we have much more flexibility.

From his list of legitimacy types, I believe legitimacy by fairness, legitimacy by performance, and legitimacy by participation present the greatest risks for the Safe network. His other legitimacy types would be strengths for the Safe network.

8 Likes
693

I’ll add an observation that I see missing from the discussion.

The network is an autonomous network. Basically its own entity and built to fulfil some goals as set out by MaidSafe for the network.

While the network is not an AI or similar, it is written with the algorithms to grow and reward for things that are beneficial to that growth.

A very good exercise is to place yourself as the network itself, you deciding things. Now in order to survive you need to provide storage that can be easily retrieved, provide applications that people want to use and provide content people want to access.

  • providing storage is fundamental to providing everything else. (Node providers - farmers)
  • Applications gets people using the network and it is expected that as people use the network they will also decide to store (some of) their data on the network thus proving income to pay node providers.
  • And of course alongside applications is having public content that people want to “consume”

Without those 3 aspects the network cannot grow as fast and may eventually fail from lack of use.

So as you in the shoes of the network consider how you can encourage growth in all three areas?

This is the basis of

  • farmers being rewarded
  • application developers being rewarded (based on popularity of their apps), and
  • content providers (based on the popularity of their content provided)

These were proposed in the beginning as working hand in hand to build the network as universal application and content network. The network is autonomous trying to grow into the new internet and your challenge is to view it from its angle and work out how you are going to do it.

Challenge - put yourselves into the shoes of the network and view these from that angle.

This is not a crypto project and to view its economics that way is to have already failed

9 Likes
695

I think this can be summarised as objective vs subjective value. We can objectively see that farmers retrieving data is objectively delivering on what was requested (data). I think we can all understand that the job being done is clear and the reward is obvious.

When it comes to the content of that data and how much it should be valued, it enters the realm of subjectivity. One person’s art is another person’s scrap. Therefore, the network cannot be the judge of this, individuals must be.

Of course, we can ignore this and reward scrap as equally as art. We may as well say any data request is worth rewarding. Where does that lead us in terms of incentives, when there is a cost to retrieve? Indeed, all other crypto storage projects, they actively charge to retrieve data. Given having free downloads is a point of economic dispute already, moving to a position of paying someone (the provider) for requests would surely tip this balance further.

Notwithstanding the above, my fear is less about the economics and more about it incentivising network abuse, potentially causing DDoS. Factoring this and the above in, is it worth it? IMO, no and I’m not convinced that testing it will change these realities substantially - data has to be retrieved and the network has to pay out, one way or another.

7 Likes
696

You nailed it. The humans here don’t like the fact that the network places no subjective value on the data. A 1MB chunk has the same value as a random set of random bits as any other.

From the network’s perspective it is. See @neo’s post above, he summarizes the main points nicely.

It is not an AI based on the typical artificial neurons for deep learning, but it does encapsulate a form of artificial intelligence based on the way complex behavior emerges from simple rules, similar to the way collective intelligence emerges from the simple behaviors of ants or bees. To be more exact, the design/implementation of Safe leads to a form of “artificial collective intelligence” (IMO).

4 Likes
697

This discussion has gone on in the abstract for years and it is pointless to restart it again. As happened recently in the marketing discussion, there’s a point where talking has to be supplemented with experimentation. “Contact with the environment”.

We have a lot of people here who like to express ideas and argue points, me very much one of them. That’s a good thing, but it isn’t sufficient if you want to be able to make good quality informed choices. For that you also need to be open to other ideas, have curiosity and a willingness to experiment and see what actually happens.

People have strong views, people think they’re right, there’s a clear disagreement that won’t be resolved without moving from abstract discussion to collaborative problem solving and experiment.

2 Likes
698

Yes, we need to understand that incentives to use the network - to encourage storage, ultimately - are desirable/necessary. I think the question is whether the network can provide all these at its core or whether they should be delivered on layers above.

Can we remove barriers to allow easier tipping/sponsoring/rewarding directly between user and provider, for example? Yes, I believe we can. It should be no more complex than hitting ‘like’ with a suitable design. We can also take approaches like Medium does, where subscriptions are pooled and distributed according to reader time, likes, etc.

We can also divert some tokens to an organisation which shares them out. I think this is more controversial, but we can make the argument that a small fee can lead to a better maintained, more stable, more usable network. At least the organisation can make a subjective analysis and decide where to push the funds.

However, these custodians of the network will be sucked into political arguments pretty quickly, the more the scope of their funding increases. The governance of such a fund will therefore become a hot topic and the influence they will have to declare winners will come under scrutiny. Some will reject the network on this basis, which is the reverse of the objective it was supposed to achieve.

All in all though, I feel incentives should come outside of the core network. Initially, Maidsafe and community funds will be involved with seeding this. Beyond that, if Maidsafe want to blend in with the wall paper, staying out of this sort of thing may be desirable. The same for the Maidsafe Foundation too, perhaps.

3 Likes
699

This changes the viewing from the network to human again and relying on humans to tip, gift, buy.

My whole post was to get people to consider it as if they were the network.

I am pretty sure that any PtP will needs be supplemented with tipping or whatever if the person is hoping for ‘normal’ income from content. The idea of PtP is to provide a base return, an incentive, for their content and that reward is based on how popular it is.

I’d say for someone who uploads one document (a few chunks in size) will not be looking at PtP to be significant. But for say youtube people where their videos are large then they will receive income based on views, rather like what they get now on youtube. How much i am not sure, but if a person is after a healthy income then they could do like youtube and have like superchats but for videos as well as live streams. Also Memberships like youtube has. Except there is no middleman taking 30+%

Again though my post was designed to have people spend time and let the concept sink in while placing themselves as the network itself. Also that this is not a crypto project with its own issues and realise this is a content network (a new internet) and crypto economics thinking will always hinder being able to consider the network as its own entity selling resources and rewarding actions that encourage buying resources to store data.

What people will come away with when they consider themselves as the network (for a reasonable long time) and work through the issues I do not know, but until they do I reckon they are working without valuable information/wisdom.

8 Likes
700

I’m the network, and I’m faced with two videos of the same length: (1) 5 minutes video of paint drying; (2) 5 minutes video of a maidsafe dev explaining how to use a certain Safe API.

To pay for storing the video or retrieving it, I find the task straightforward because it’s objective: x safemoney/chunk price for either of the videos.

To pay for the content of the video, I find the task difficult because it’s subjective. From my perspective as the network, both videos are identical, so I could just use the same pricing formula as I used for storing them. But the humans who coded me told me it can’t be the same price. Maybe they told me to use number of views instead (or maybe with a NN I’ve learned to reward the content that brings more attention like controversy, misinformation, and extremism similar to the Facebook algo). So I’ll do that instead.

Turns out the paint drying video has a lot more views than the maidsafe API video, which much fewer people found interesting. So I pay the paint drying video more safemoney. But hold on, there’s more and more paint drying videos being uploaded, and what’s more, misinformation and extremism too. That’s great. Let me reward them even more to encourage more of these types of content. I have no idea what humans are doing or if they like my token issuance to their paint drying and controversial videos, but I’m doing what some humans coded me to do, and I hope I’m meeting all of the other humans’ needs and predilections and that everyone is happy.

Yes, Safe isn’t crypto. But that’s beside the point. The point is that crypto reflects a mirror to a certain aspect of humanity, the force that drives us to exchange valuable resource for something. That same force governs many of our social intercourse. And it is a force that has evolved in us for millions of years. As the Safe network will also be subjected to that same force, it is entirely reasonable to try to learn how that force shaped other networks. If you think Safe will not be subjected to the same force or can rewrite it entirely, then please share how it’ll do so in such a short evolutionary timeframe.

I think it’s a mistake to assume that those cautioning against PtD/P adopted their position reflexively or without serious consideration. In some instances, that opposition came as an evolution from actually being for PtD/P initially.

If Safe is to become a force of nature (“like oceans or mountains” as someone else put it more eloquently), the builders of the network have to try really hard to overcome their own biases or any attempt at control. The network should have no business dealing with the subjective. Let humans decide or make their decisions at the app layer. Only that which can be objectively measured by the network should be its concern.

Nature doesn’t know and can’t control if some group of humans decided that one of its rivers is actually the goddess of fertility and therefore holds great value above all other rivers. The natural ecosystem just carries on all the same without favoring any particular tendency of a particular human group. This is what Safe should aim for too in my estimation.

4 Likes
701

The problem is that you missed the point. If paint drying videos are intensely popular and increase network growth, then it is in the self interest of the network to reward the content so that it can grow. Growth is fundamental to success. All those paint drying videos keep the cure to cancer or breakthrough inventions safe and secure.

The broader question to me is not if it should be implemented, but how can it be implemented in a way that mitigates the gamification issues @Antifragile has pointed out.

4 Likes
702

@Bogard
Obviously the network cannot know the quality of the work as humans see it. So what is the measure it can use. Seems accesses is the way a neutral network without biases can do it simply.

Paint drying, if multiple ones show up then it dilutes the original doesn’t it. Same for quality videos etc. Should the network care? Why should it? If humans want to see paint drying then its content that is getting people to use the network. Goal achieved even if humans are stupid.

And to pay PtP when storing the work is an obvious no-starter since it doesn’t account for usefulness to the network (to get people using it).

And caching will mitigate the gaming. And the question is even if there is gaming that “takes” 10 or 15% of all PtP paid out, does that defeat the purpose? If you think so then what about scammers gaming farming by emailing out malware to install nodes on unsuspecting people and earn the scammers 5% or 10% or 20% of all farming rewards paid. Should we stop paying farming rewards because of the gaming that will occur of farming rewards?

2 Likes
703

Gaming of farming rewards would actually help the network however, whereas gaming of PtP is a burden on the network.

704

Not really since those people with the malware delivered nodes will disappear relatively quickly and cause a lot of relocating of chunks and thus unnecessary work for the network.

705

I thought it was obvious that I’m ultimately concerned about the unintended and unpredictable effects that PtD/P could have on the network’s staying power/adoption/etc. Put differently, I don’t think that every possible way by which PtD/P will be abused can be envisaged a priori and planned against. You like KISS. Don’t you think such a risky concern calls for KISS?

It’s a bit more complicated than that. If paint drying videos is what humans want to see and what the network therefore further incentivizes uploading, then the cost for uploading anything else will go up. Furthermore, if humans prefer paint drying videos then the network doesn’t need to incentivize it because humans will naturally upload that content regardless. Further still, one could argue that popular content should cost more because the network does more work for it (is it fair to further tax those who work with difficult subjects that the average person finds boring?) There are other potential second order effects.

I like the idea of the storecost being a rate limit to ensure that the content being added to the network is valuable either personally to the uploader (kept private) or to the public (in the collective subjective opinion). If the network is issuing safemoney to encourage content that isn’t valuable, the social consensus could conclude that safemoney isn’t valuable either by proxy (and as a result, a large number of potential network participants might instead flock to a different network with a token that they see as valuable). A network that isn’t subjectively deciding winners and losers would better satisfy legitimacy by fairness.

[I get the potential counter argument along the lines of the junk data being used to secure the network, similar to say the Burst network or to wasted hashes on pow blockchains. But is this the direction we want the network to take, with data potentially being prohibitively costly for the average person to store anything unless it’s a specific type of content that the network will reimburse them for?]

1 Like
706

All reasonable considerations to raise. I certainly don’t have the perfect answer to this complex topic. It’s why I favor keeping it very simple and letting humans deal with the rest on a separate layer and over time.

2 Likes
707

Indeed.

Again, the point is being missed, and your conclusion is not predestined. External Fiat vs. Internal SNT markets make this too complex to predict. The PtP reward rate doesn’t need to be fixed and can vary with PUT and GET rates. A larger and more popular network can bring in more resources that leads to lower absolute costs for all users than would have been possible without PtD/PtP.

Who decides what is valuable? Forego the human subjectivity. It’s easy to define what data is valuable when looked at from the network’s perspective. Valuable data is the data that results in more growth.

One way to think about PtP/PtD is as part of a continuum of the tipping model. Humans are free to tip or pay for content based on whatever subjective value they may place on it. This could be managed in the app layer, but it might also be useful to have a hook in the core api to make it a more uniform process. From this viewpoint, the PtP is essentially the network giving a tip to the content provider because the data has caused growth, which is preferable to the network than the alternative.

708

@neo

I wouldn’t consider that gaming it because there’s no fault in the logic. The network should have no exploits built into its logic, however people being shaken down by criminals for network rewards is no fault of the network.

2 Likes