PtP, PtD, and/or Pt* Megathread - Pros and Cons

Features
41

The more I think about it, the more convinced I become that this is the best path to ensure a network that is a lasting and useful infrastructure to society at large.

Attracting people to use the network and upload data/content:
You do this by building a great product so people get better service than they get with other products. Getting that better service at an even lower cost is icing on the cake. But there’s no need to pay people on top of providing them with a great product/service at a competitive price. There are many people who would use the network if it works well (e.g., recently, someone posted here a twitter thread from Mathew Green exploring Filecoin where he stated clearly that he’s in need of and would pay for a secure data storage that’s easy to use). AWS doesn’t pay people to use AWS. People and companies use AWS because it’s a great product, and AWS charges them for that use.

The many rewards for uploading popular content:
Popular content provides all kinds of rewards to the owner/uploader of that content that go far beyond what the network can see. For instance, popular content might provide fame to the uploader, and in many ways and depending on what they’re after, fame could be far more valuable to the uploader than any additional token incentive from the network. The network paying them on top of the fame (and attendant benefits) they might achieve from their content would be like society mandating a universal income for good looking people, even though they already draw various direct and indirect societal benefits for their good looks.

All data is the same, but all data doesn’t have the same value (maybe to entropy but not to a well functioning society):
At the byte level, yes, all data is the same. But not so in the reality we live in. Even the safe network differentiates between different types of data as it prioritizes its infrastructure messages (different priorities within those too) over the more numerous client messages. Imagine if the network didn’t do that and on top of it, favored the popular client messages over the boring infrastructures messages. The network would grind to a halt. Such would happen to society too if the unglamorous but society-preserving work wasn’t rewarded, encouraged, and even prioritized in some way.

Human beings appreciate and rewards a useful thing at the end, but they don’t tend to appreciate and reward the long and painful process it takes to get to that useful thing. Most people would agree that bridges are useful. But before they are erected, it takes a long period of conceptualization and boring surveying/planning/calculations work that the average person would not reward with their attention. The safe network’s construction is case in point. The safe network would be useful to humanity and yet the community most likely to appreciate that fact is busy speculating on rug pulls and ugly images of monkeys. How much worse would it be if the internet, in its code, actively made it even more expensive to build something like the safe network by taxing it and sending that tax to the masses who already have plenty in cryptopunk riches?

The risks of pushing people away in an overzealous attempt to attract them, therefore achieving the opposite of the desired aim and killing the network:

  • There are some predictions that rewarding all content the same will lead to the network being dominated by spammy, non-critical, antagonistic, etc. content that provides a quick dopamine hit. Assuming that outcome comes to pass, is the network going to survive? Eventually, people get tired of that sort of content. And once they do, they’ll move on from the network (as people moved on from Facebook) and it’d be hard to reverse once that happens because the network would have already cemented its reputation and alternatives would have sprung in the gap it left.

  • Baking such a strong incentive in the network introduces an uncontrollable, unpredictable variable that might get the network’s usage tied to a specific generation or type of people (a la Facebook being for old people, IG for millennials, ticktock for…, etc.) On the other hand, the internet has been used across generations because it’s agnostic and apolitical. It just sits in the background and works.

  • There would be strong opposition to paying for views or content one does not support. If people know that their money goes to support/subsidize content they find objectionable, they might refrain from using the network. It might be a lot easier for them to co-exist with content they don’t like if they know that the network in its essence is impartial and agnostic (like the internet is), and it’s human beings who are responsible.

  • Political fights over what the network should favor could damage the network. What facts should the network reward most? Popular opinion/belief may be attractive (e.g., crypto at large) but expertise/ability matter a great deal (e.g., it’s what’s building the safe network). In that ebb and flow, what if some demanded that a certain type of data (e.g., scientific) must be rewarded over all other type of data? If such an incentive were baked in the internet, would crypto at large agree to being taxed for building a competing network that they are convinced would never work and is possibly a scam?

There are so many other considerations (a critical one being network security which depends on token economics), but I’ll add just one more. The safe network’s current design is to outsource as much as possible to the client. It would therefore be a continuation of that design spirit to also leave it to the client to determine the value of content on the network.

8 Likes
42

Lets take an example of where uploaders of content are being paid based on views. Youtube pays them according to the number of views on their videos. This currently is calculated by the number of advertisements presented to the viewers.

For all the pros and cons it seems to be successful and a load of great scientific and educational content as well as the fun/crazy content some hate and some love. [TikTok is an example where silly content is favoured, rather like people who like comedy entertainment. Doesn’t mean they don’t like other good content]

For me the problem with you tube is that the uploaders are paid from a portion of the money received from the advertisers. Another problem is the human/AI interference into who/what/how people’s content is judged and censored by youtube/AIs/Political/advertisers.

If we translated this to SAFE with some sort of PtProvider then we remove the negative aspect of a big corporation interfering and controlling the content.

We will see then the uploaders/providers being rewarded with a tiny portion compared to what the farmers get on each GET (like you tube gives tiny portion of advertiser money). 1/10 was suggested early on, maybe lower now since no reserve pool. This means it comes from payments by people uploading and distributed to farmers and PtD, PtP, core devs.

In essence instead of allowing advertisers to dictate the terms, it is a secure system for everyone and these uploaders/providers are given a reward (payment) for content that people wish to view.

You Tube has all sorts of content from very high level educational material to the comedy.

When you have a network that is for everyone then I, nor you, can dictate what is good/bad for everyone and denying people payments for their contributions because some will misuse it is short sighted in my opinion.

Why discriminate against other resource providers to the network while rewarding the farmers for their contributions. No one is suggesting that farmers should not being rewarded the majority since they have the higher costs associated with their contributions. Remember that even if you remove PtD and PtP there are other rewards being given out.

What do you remove since all are providing resources (data resources, storage, programming, etc)

  • Core Developers rewards - the network stagnates in a ground hog day for features provided&bugs
  • Application developers - removing means you wish them to develop for nothing. Works in some areas but creativity for variety drops away very quickly when they can get paid working on old internet applications/games/etc
  • content uploaders/providers. Plenty of places to get paid on the old internet so why would they move to SAFE. Even if they earn a lot less with PtP they may slowly migrate with adding content.
  • Farmers being one of the many resource providers is the only one people think is worthy of being paid for their efforts. But if no one uses SAFE for much more than backing up data and trying it out, then why would farmers remain. End up with a niche network that only has enough farmers to support the occasional data upload.

BTW talk of other data on the network is creating strawmen. Everyone knows the context of data and bytes is uploaded files, no one has ever suggested Pt[P,D] or core devs, even farmers are paid according to number of messages sent, decision packets sent, and other data overheads. That is overheads and why farmers are paid the majority by far compared to the rest.

This of course is my biased opinion, but thought examining the example of you tube could give real life experience to the discussion.

6 Likes
43

Data is data. It’s all the same value to the network. Rewards and payments made by the network to humans must be based on the network’s value system to have any technical merit. Some here are conflating their hopes and dreams for a narcissistic utopia with what an autonomous data network needs to survive in the wild. This is pure folly.

8 Likes
44

The point isn’t to deny someone for their contributions. It’s merely pointing out that some contributions will be rewarded better over others, which some might see as fair but others might see as unfair. For instance, something like the safe network’s design write up will not be rewarded well compared to a colorful video with shocked faces, yet I would personally value the former higher than the latter.

Ultimately, another confounder here is what should safe network be? An infrastructure or a social network? Content rewards veers into social network territory. I personally believe that to minimize risks, the safe network should stay firmly in the infrastructure realm and let others build social networks on top with different reward experiments (e.g., decorum).

4 Likes
45

We already know. An agnostic to content network. No rulers to decide what is allowed/disallowed, what is good/bad, and so on.

To program in society’s biases is practically impossible to do even marginally successfully, but also silly for a S.A.F.E. network since different countries have dramatically different views at any one particular time and any society has different views over time. If you want human controlled network then SAFE is not it.

But if people prefer a colourful video over documentation then so be it. Society of the people decided.

Documentation is not as useful to the majority and in any societal system to date this is a sad fact of life. Back in the 60’s when you had to purchase any/all documentation for a computer system, there was few copies purchased by individuals and the company with the computer may buy 2 or 3 copies depending on the size of the Main Frame system.

Compared to any other journal purchased by people in the company it was tiny. Highly valued to the few who needed it or wanted to learn, but not used by the hundreds/thousands of other people in the company/university. This is one aspect of society that has yet to change.

1 Like
46

Exactly.

But putting in content rewards could jeopardize this. It would be the network as the ruler that further advantages certain types of data that we all know most people naturally tend to prefer anyway and would reward in other ways anyway.

And might never. So the argument isn’t to try to fix it (trying to fix it would be to insist on a reward system that incentivizes the opposite). Rather, the point is to sidestep it and stay agnostic to it.

1 Like
47

How so, it is the people that decide.

The network does no choosing, the people decide by favouring content.

But I feel you are being too focused and missing that the people of the world are very diverse and I doubt any particular type of content will rule. In any case it is up to the people what they decide to watch.

Also how is the network going to promote any content? It is the people and search engines and reviews and word of mouth that will drive people to certain content. Just because people putting up certain content get paid does not cause the network to promote their content.

It is normal for society to do things that make them more income. The network will not direct these people.

Also if PtP is non-existent or too small then there will be other forms of systems that pay people.

I’d fear much more another You Tube springing up to pay content uploaders with advertising money and you get exactly content driven by what pays the most with the you tube system directing people to that content. This is in contrast to the SAFE network which has no mechanism to drive people to content that makes some company (or app writer) more income.

This is the beauty of an agnostic network paying for content that people want to watch.

The biggest reason we see certain forms of media unnaturally being big is the advertising dollar driving companies to direct traffic to that content through suggested media and search engine results. Its not individuals making more but the active driving of traffic to these media by companies

Paying content based purely on views does not have the problem caused by companies and advertising dollar skewing it all.

Which is exactly what having PtP or not having it would be. The Network remains agnostic since it has no clue what the content is.

Having PtP does not cause traffic to be directed to high viewed content even if people want to make that type because it pays better. It pays better because people, not the network seek it out. Unlike advertiser driven systems.

2 Likes
48

It’s interesting that we’re both advocating for the network to be agnostic.

  • But your definition of the network being agnostic is for it to reward people for uploading content (and you agree that people prefer certain types of content, so ultimately, the network will be incentivizing certain types of content and behavior).
  • And my definition of the network being agnostic is for the network to just charge for uploads and not get into the business of rewarding uploads. Being a great product is all the network should need to attract uploads.

Just an interesting note.

Ultimately, it’ll be alright. If safe network makes the wrong call, there’ll be alternatives. If it makes the right call, it’ll remain the leader. Same outcome for the end user. So it doesn’t really matter in the end what choice is made here.

1 Like
49

All i ask is that we try it. IMO there are a lot of worries which are highly speculative and when i look at the arguments seem not to justify the opinions.

For example, comparing tiny indiscriminate rewards to the hell hole of centralised attention and surveillance driven algorithms is too big a leap for me.

To not attempt this, with the enormous potential benefits of no brainer monetisation of apps and content seems a much bigger risk to me.

5 Likes
50

But there’s more than those concerns. Allow me to put it this way. Would bitcoin’s security still have been as strong as it has proven so far if:

  1. a portion of the “allowed” hash power is granted to some of the most popular people on earth, and
  2. that assigned hash power (outside of mining equipment) can be stored overtime and could potentially be used at a later date to attack the network?

It might appear at first glance that that comparison has nothing to do with the topic at hand, but it actually does.

1 Like
51

Again, i don’t see this as a significant argument against trying to do something with enormous benefits. I’m sure we can think of many ways to increase or decrease security by unspecified amounts, but that’s not necessarily important. IMO not important enough to rule out a trial.

We haven’t got a network yet, we don’t yet know what the theoretical or experimental security performance is, so these argumentss are sterile. Let’s get the network built, run lots of tests, code reviews, design reviews and run trials such as PtD and PtP and do some proper modeling of them, and not get into speculative discussions over things we really can’t predict without that data.

4 Likes
52

Fully agree.

I disagree. There’s a lot you can model on this topic without having a network up already. The direction the design work takes will depend on what was envisioned, otherwise months could be wasted. So it’s actually an important topic, and I’d suggest those who truly believe in these sorts of rewards to prove that there’d be no security risk for the network. It’s just math, no network needed.

Anyway, just to wrap up the convo. As I said above, whatever path is taken doesn’t really matter too much because there’ll be options. I’m basically agreeing with you that many things will be experimented with, and there’ll be winners and losers.

1 Like
53

Without knowing network characteristics you can’t model how anything affects those characteristics. Also, proving no risk is not a useful or even possible measure. We’re not aiming to maximise security at all costs, but to provide a balance across a range of criteria, including security, utility and appeal.

3 Likes
54

I’ve always liked the idea of PtP and PtD etc, but can’t see how it could work in an automated way that couldn’t be easily gamed.

If rewards are based on how often something is accessed, gaming is a doddle, and it incentivises spam gets.

If rewards are based on quantity of data used, it’ll incentivise over-consumption of bandwidth, which is the opposite of what the network needs to be healthy.

With the Safe Network, even without Pt*, micro payments to content creators / app developers / protocol developers etc will be so easy that I expect it’ll be much more common for people who provide value to be rewarded by users / fans without anything specific at the protocol level. This alone could be revolutionary and allow for avoiding the corruption of advertising & centralised platform dependency.

But, if we assume there is going to be a fundamental under-rewarding of important creators by users despite built in, easy to use micropayments (which surely is a required assumption for pt* to be seen as desirable?), then could a system be made that directs a portion of farming rewards to go towards Pt*, but with the target of the rewards being decided by the farmers / users rather than the network?

Here is a suggestion of how this kind of thing might work:

  • Farmers / users receive ‘Pt* tokens’ as a portion of farming reward
  • Farmers / users receive ‘Pt* voting tokens’ as a very tiny portion of rewards
  • Farmers / users vote for content providers / developers who they think should be eligible for rewards
  • Producers / developers become eligible to receive Pt* rewards by receiving a threshold of ‘Pt* voting tokens’, possibly from a threshold of unique farmers / users,
  • Farmers / users reward eligible producers with Pt* tokens at their discretion.
  • Received Pt* tokens can then be exchanged by eligible producers / developers for Safecoin with the network.

This way, a set portion of network rewards can be reserved for developers / content producers / network developers etc, but farmers / users get to apportion these resources according to their perceived value.

Something like this could be made far harder to game than programming bots to geep getting certain huge files, and I think it’s inevitable that this would lead to a better distribution of the rewards than any automated mechanism / algorithm ever could, if the goal is incentivising things people value without needing corrupting advertising etc.

Sorry that’s a lot of waffle of a raw idea, but hopefully it’ll contribute to the discussion!

4 Likes
55

This has also been my basic attitude to it -

great idea, how to make it work in manner that will cause the least moaning - cos someone will always feel hard done by…

was a figure of 15% to Pt* not part off the plan for quite a long time? @neo would know.

Seems a reasonable scheme but it also looks like awful lot of extra code when we will be able to do micropayments on top anyway.

1 Like
56

That is not part of my definition. Agnostic network is an attribute of the network towards the data. It has to be according to the core principles of the network.

That is actually putting it into what you claimed I was saying. Charging or paying is just + or -, both are done according to the quantity of data and you said that was part of my definition.

But I take agnostic as an attribute unrelated to charging/paying. It means that the network has no say over the data it receives or gives to people. It has no knowledge as to any aspect of the data’s quality. As long as it is packaged correctly and follows the protocols for uploading then its accepted and will be available to be retrieved.

The charging for uploads or paying for resources/services provided is another attribute unrelated to agnostic, but affected because there is no way to identify quality to humans of the data. Agnostic applies a constraint on the paying to/from the network, but they are still not a part of each other.

Every attribute of the network has constraints on them from unrelated attributes. I used unrelated as to what defines the attribute.

Paying the network or being paid by the network is simply a function of the network to control flow of data and prevent spamming.

  • Network is filling up
    • cost to upload is raised to slow down uploads till more farmers come online to increase free space
    • payments to farmers increases inline with upload costs to encourage more farmers in order to increase free space
  • Network is not being used, initial state and need to grow from there if adoption is to happen and maintain a healthy inflow of data.
    • Adoption means healthy inflow of data since data in the world increases daily
    • Only a niche sector will use network if little mainstream APPs/Content is on the network.
    • Internet is the success it is because it was commercialised, but also is its failing due to control/censorship by these commercial entities. And loss of privacy.
    • Before commercialisation we would be back to the early 1990s where it was specialised tech and university related stored content. And email and news groups. Very niche market. And that is what SAFE will be if we do not get adoption
  • Question is how to encourage adoption.
    • David wanted to turn the commercialisation of the internet on its head and proposed some simple ways to do that.
    • First was to remove the companies from the commercialisation because they impose control, loss of privacy, censorship, unfair profit taking at the users expense (money and privacy)
    • at the same time as 1st Prevent spamming filling up the network.
    • paying for uploads reduces spamming to more acceptable levels, never able to stop it completely
    • Second was to provide some basic payments for people who provide resources and services to the network which increases adoption.
    • Use the payments into the network to pay out to people providing the resources & services.
    • Privacy is another attribute of the network and requires the removal of these entities controlling the internet, but the removal is not a part of the privacy attribute.
  • Thus commercialism is removed and everyone is now on an even playing field.

Now what makes up the paying into the network

  • uploads, thats it

Now what makes up providing resources/services.

  • Farming, we all know this involves
    • services - node operations, including but not limited to consensus, message passing, caching, handling clients
    • resources - storing data and retrieving data. Storage resource
  • Investors
    • resources - some of the capital for writing the initial system 5% of total SNT supply reserved
  • Maidsafe
    • services - writing of the initial system
  • Core Developers
    • services - maintaining the core systems
  • Application Developers (Writers/providers).
    • services - providing the Applications that will encourage adoptions. Not cheap to do this.
    • Pay to upload their Application since its uploading of Data
  • Content Providers
    • Services - creating or curating content to upload - again not cheap
    • Resources - Content on the network that people want to watch causes adoption.
    • They pay to upload the data
  • All the above was proposed by David to be included in the payment +/- ecosystem of the network.

For an agnostic to data network there is no distinction as to the quality of any data nor to any value system used by the multitude of societies in the world. Why should the network favour one type of resource/content provider over another. Although the network can adjust payment levels according to the function of the provider

  • Uploader
    • requesting network to store data
    • is sole source of income for network
    • no knowledge of content usage other than private/pubic for purposes of generating address for chunks. Private data has ID included in the address.
    • Is charged the minimum amount to ensure network meets outflows.
  • Farmer
    • multiple services and resource provider
    • Provides more services/resources than others
    • paid accordingly (Initial suggestion was at least 10 times any other provider)
  • Investor - Not an ongoing operational provider
  • Maidsafe - Not an ongoing operational provide (as a company generating initial system, but as an App/Content provider it will)
  • Core Developer
    • providing updates and bug fixes.
    • Initially proposed collectively at 5% of what farmers received during the period the work was done.
  • Application Developers
    • Initially proposed to be paid according to application code usage at 10% of farmers for those GETs on App program
  • Content providers
    • Initially proposed to be paid according to content usage at 10% of farmers for those GETs on content used/viewed

The network needs all the above operational providers in order to be adopted. If anyone can name one of those providers that can be removed and have success then let us know how.

The operational side of the network is provided by the farmers + core devs.
The usage side of the network is provided by the Uploaders, App and Content providers.
Each of these will not work without the other, they are in a co-dependant relationship for a new internet.

The operational side may work for a company since they have set requirements for the company. Few Apps developed and their own content stored.

But for a world wide internet replacement we need both sides to be healthy and vibrant. The current internet showed that once a content provider was paid for content the usage skyrocketed. Yes Porn showed this truth, then some companies went online providing their limited content (business transactions, tech data). But it was not till the content providers like You Tube, Twitter, Facebook and the multitude of others (netflix, etc) that the internet took off and was being adopted by grandparent to young children.

We cannot underestimate the power of the cat video to create adoption, and while people can get/upload their cat videos free on the old internet they will see no reason to pay to upload onto a new system that isn’t yet adopted. It could be the greatest system for privacy, but the ordinary cat video uploader couldn’t give two hoots about it.

3 Likes
57

This is incorrect comparisons in my opinion.

Those popular people in your comparison do nothing to help the hashing power or the bitcoin network. That invalidates the comparison.

For the network to be paying for services/resources the provider is doing valuable work for the network in causing and maintaining its adoption. So your comparison is really a strawman, changing the argument into something else and arguing that other thing.

The 2nd point can be more applied to farming since farming is rewarded at a much higher rate and has no reliance on uploading something that will become popular. It is more sure to simply run a large farming “farm” and collect payments for every GET from every farming device. And the payments are at least 10 times the “popular” content that is not guaranteed to become popular long enough.

TL;DR

Point 1 creates a strawman argument since SAFE providers are active helping the network adoption whereas your strawman is just giving a limited set of people large amounts without them doing a single thing for the bitcoin network.

Point 2 applies more to farmers who get way more in comparison. Popular person uploads 100s of videos compared to farmers running 100’s of farms - the farmers gets way more and thus has more ability to attack at a later date AND is running the nodes to help attack.

3 Likes
58

Farming is easier to game. It provides higher rate of return.

Bloating is one of the easiest ways App devs and content providers will game it and the solution will be greed since someone will upload a non-bloat version and get the usage over the ones gaming.

All payments to providers will be gamed to some extent, that is the nature of humanity as a whole, we game the ground by our farming of the land.

This is more profitable by the person setting up their farming devices (maybe 100’s or more) and finding the chunks stored on their devices then accessing those chunks repetitively. Something similar to rainbow tables will allow this to be found in reasonable time by having known hashes for known chunks spread across the address range, then knowing each node’s address one can determine what chunks to try.

Caching prevents massive gaming on any of the providers payments since trying to do massive GETs on those chunks (Farming, App Dev, Content) will just be hitting caching in the end with a significant cost to do those GETs. Dimishing returns.

Cheaper in the long run to just farm with those bots and let the world help earn by accessing data on your farm(s)

I fear that like the current internet we will see commercial enterprises providing this functionality and be back to something akin to the current internet in that area.

The way to help prevent this is to get adoption faster and commercial enterprises do not have time to get a foot hold. One way to help that is to allow the network to pay all those proving resources/services to the network.

The google attack becomes possible here in that Google runs a lot of farms (millions or lot more) and thus control the content through selecting content to be watched. Paying producers by the network prevents this simple attack on control of the network.

Leave the decision in the hands of an agnostic to data entity, The Autonomous Network.

4 Likes
59

It was 10% for APP and 10% for Content

The farming rate (FR an amount per GET) was calculated and farmers got the FR then PtD got 10% on GETs to the code and PtP got 10% on GETs to the content. And Core Devs collectively got 5% on every GET.

So between 105% of FR to 115% of FR would be given on a GET. (115% since only one of PtD or PtP would apply)

But this is to be revised when the papers are rewritten soon as mentioned in an update

3 Likes
60

Sorry to all about the long posts. Thought it’d be good to revise some of the history, facts, etc. I’d like to think people will read them for the information and my opinions if they are interested in the topic. But Hey I know “walls of text” are not always enjoyable to read, but maybe you’ll find my errors in logic :joy:

The most important thing from the posts is that

  • the network is agnostic to data.
  • the network is not imposing anyone’s or any society’s values onto the data or what is to be popular.
  • that commercialisation creates an illusion of what should be popular by manipulating people. (suggesting whats popular, censoring, manipulating search results and so on)
  • Farming, Core Dev, App Dev, Content creation/curating are all essential services/resources that together are needed to make a desirable network to use thus being adoptable
  • All providers are doing valuable work for the network and not free loaders.
8 Likes