From what I understand the data which a user wants to upload, there will be a bidding at which price farmers are willing to accept to store the data. Free float supply/demand price.
Let’s be realistic about users. Today a content producer can put his content on YouTube for free and be paid. This remuneration can be prorated with Youtube (if it is Google ads) or it can be self-profit (sale of products or bilateral advertising). Why would he prefer to put a video on Safe Network? If multiple videos are being viewed on the SAFE Network, that new user will have to pay the average cost for those content, even if his video does not get any results.
A Farmer is not able to evaluate this cost, since the cost is generated by the number of requests that a content receives, not just the size of the file.
Getting strong vibes of this:
4 Likes
It has nothing to do with relative popularity. It has to due with data perpetuity. Although the algorithms aren’t finished yet, the philosophy of a perpetual data model is pretty inuitive. When a client requests to upload a chunk, the network determines a price that reflects the cost to maintain the present state and growth of the network due to the extra burden . A portion of this payment contributes to maintaining all chunks that came before and a portion contributes to its own maintenance going forward. Potentially, some chunks could be valued higher than others when rewarding farmers. For example, it might be beneficial if old chunks receive a higher relative reward than new chunks to incentivise long term stewardship. In 200 years, serving a rare chunk from 2020 could be like hitting the lottery. However, when it comes to upload costs, those are simply based on supply and demand for resources at the moment of upload.
3 Likes
It would be great if we could have a SAFE version of this old clip A Packet’s Tale. How Does the Internet Work?
I get the OP was asking for a wider spectrum of information but a nice simple ELI5 like this re the basics would be fantastic.
2 Likes
I sympathize with @gan_rl point of view. But these are medium to long term concerns. The network needs to start somewhere. Keeping it super simple and as currently envisioned is the fastest path to launch while respecting the pay once store forever fundamental.
That said, there are serious potential side effects that could have the opposite effect of that originally intended if not adjusted for depending on how things progress. For instance, in the long run, uploading could become so expensive (to account for the high bandwidth usage of popular data) that only those with means can upload and the less fortunate might end up excluded. That would actually become the capitalist point of view @gan_rl, not socialism as you put it. The less fortunate/popular pays even more than the already fortunate (as in traditional banking and other societal setups today). Trickle up if you will.
It’s great to think that educational sites will benefit from the payment system set at the outset. But for every Wikipedia, there’ll be several and far more bandwidth intensive TMZ, Worldstar, Dailymail, etc.
For instance:
A 2kb SAFE improvement proposal document would be downloaded maybe a couple of hundred times. A QANON conspiracy document of the same size would be downloaded millions of times. The SAFE improvement proposer would subsidize the conspiracy theorist and manipulator.
Or:
A 10mb video file of a townhall meeting discussing the town’s budget allocation decisions versus a porn video of the same size (again, the former is far more likely to subsidize the latter)
In the end, “boring” content might find themselves increasingly unable to afford upload onto the network having to pay with the assumption that their content is “valuable” enough and will have high bandwidth usage. That’s pure market.
To recap, we should focus on the simplest system to start, which is the current one. However, it rests on the assumption that upload costs will be inexpensive in perpetuity. That may be the case, but if it were to change, what then? If farmers are rewarded precisely based on the amount of work, can uploaders be charged precisely based on the amount of expected work on the data? After some time of network operation, maybe the network can learn so much that it can predict an estimate of the amount of work it needs to perform on a piece of data and charge the uploader accordingly?
2 Likes
But they are disconnected concepts. The cost of GET is different from the cost of PUT. The current model is creating a connection that does not exist. If the number of current GETs increases, this model will reflect that cost for the new PUTs. If the number of new PUTs does not balance the current GETs, the system collapses.
1 Like
If the cost of GETs is too high, this should be reflected in such a way as to discourage new GETs, not new PUTs!
Agreed. I think the system may be such that it isn’t so much (or only) that unpopular data subsidizes popular data, but more (also) that GETs subsidize PUTs. In this world the cost to upload data to the Network may become extraordinarily prohibitive in order to keep farmers who are servicing both PUTs and GETs solvent/profitable.
One solution to this is to charge a fee once GETs exceed a certain level. The data would still remain on the Network perpetually no matter what, and if it’s public data it should be such that anyone can pay the fee to reinstate access. The assumption here is that say the fee kicks in after a million GETs, by that point the “owner” of that data is likely harvesting some kind of profit from all the views anyway. Thus, it only makes sense that the owner of the popular data should give back to the Network (I.e. global SAFE community) in the form of a fee. More so than capitalist, I think this is more socialist (if we see the Network as the “state” who controls the means of production) or communist (if we see the SAFE community as controlling the means of production). Perhaps even, this would suggest a new governance paradigm that breaks the bounds of Capitalist/Socialist/Communist thinking.
Perhaps the Network launches as pay per PUT only, and eventually evolves to pay per PUT and pay per x number of GETs as the workload changes. To that end, by what mechanism would such a change be ratified. Rigid systems break, so some governance framework needs to be in place to provide the Network the flexibility to evolve as the environment evolves around it.
To that end, what precisely the governance framework for SAFE will be is a question that remains unanswered as well.
3 Likes
And this is just one asset unique to SAFE being overlooked here.
Depends on who it is and the app they would publish to on the network.
Maybe they are aligned with the ideals of SAFE.
Maybe they don’t want to be demonetized by a monolithic company’s algorithms and not receive answers to why.
Maybe they don’t want the video to be deleted or censored.
Maybe the platform takes the zero ad model because it already gets PtD rewards and that will draw in a large audience that are sick of ads stealing their attention.
I can’t pretend to know all the reasons a person would or wouldn’t do something but those are some that I would consider.
Why would or wouldn’t you?
3 Likes
I do agree SAFE won’t be the optimal solution for EVERY type of data, but at the same time there still are significant reasons to pay more then your share of GET costs to do a PUT. For example if you have some sensitive data that maybe doesn’t need to be accessed alot but you need it to be secure and always online. At the same time the current model does kinda bias against smaller operations since the cost of entry is the same no matter how many “customers” you get.
In a way I would contend SAFE is not “for everyone” as the name claims. It will be for some people but others will do better using old technology. That doesn’t mean the new tech is useless… just some people will already have good enough solutions and SAFE will not be enticing to them. Other’s will get great value out of it they could have never achieved with legacy internet.
2 Likes
In the worst case, the economic model of the network will not work long term. This means that there will not be one huge network.
But we, the humans are genetically predisposed to help our loved ones and our immediate community for free. In the worst case scenario, we will have thousands of local SAFE networks in the cities, which will be maintained free of charge by volunteers.
2 Likes
I think we will have to see how SAFENetwork sites and apps evolve. They will be quite different from what we have on the clear net today, due to the users retaining ownership of their data. This reduces the opportunity to create a walled garden to monopolize the data to transfer its value.
In this model, YouTube are not providing any data storage services. They would simply be a portal to watch videos users had submitted or perhaps found through a search bot.
If they want to make money, they could embed ads in their videos. Maybe this would become more popular if Google ads become less popular.
We also don’t have a complete picture of network cost. The more popular data is, the less effort is required of the network for that data (per user request). The caching mechanism will likely do well with viral and ever popular data.
I’d also suggest that videos are probably the extreme case too. Uploading a text and images will almost certainly remain very cheap to upload.
There was also discussion about site owners being able to pay for the upload fee on behalf of the user (not sure the implementation status on this - @dirvine, any feedback?). The user would (potentially) relinquish ownership of the uploaded data, but the economic model would then become more similar to the clear net. That is, the site would own the data and could monopolize it, advertise on it, etc. So, the user may have a choice of whether to pay or not.
Imo, the upload costs will not be an issue, as long as they remain tiny. In either clear or safe net, the upload costs have to be born somewhere. It just depends on whether users will be prepared to pay for it (for ownership, anti-censorship, control, etc) or whether they will want another user/site to subsidise it.
3 Likes
Btw, thanks for @gan_rl for some great debate. This place has been a bit of an echo chamber recently, with more talk of marketing and market cap recently. It is good to actually talk about the technology and its advantages, disadvantages and initial limitations.
9 Likes
People who value their data and want to own and control it will pay for the service. Sorry but the safe model is not suitable for “troll” data.
He has valid concerns. YouTube allows people without money to try their luck in “show business”. The SAFE network will not be competitive in this niche.
1 Like
I dont agree with this. YT is “free” but takes large pert of the profit and can end your career any time because some algorithm decided you did something against the rules. On Safenetwork, you have to invest little bit in the begining, but what you create is your and that is a big thing. Actually the tought you are paying fort the data can make the content better, because you will thing better about what you want tell. I see it as a bit different market, but competetive in many ways.
Of course some things will stay on YT, and it may be better 
https://www.youtube.com/watch?v=PLOPygVcaVE
3 Likes
Content creators use YouTube because everyone is there, not because there are no free alternatives that allow you to own your products.
See Dtube, this is what the SAFE network will be for YouTube + it is free and still not competitive to YT.
People generally do not understand that size matters. Things have different characteristics when they are big or small. YouTube is like the sun, which attracts everything because it is to big. One day SAFE may be a competitive sun, but it will take a very long time (competitive to YT)
2 Likes
There are no free markets, no systems that attribute costs to consumers based on the principles of free markets etc. All markets are flawed, unfair, inefficient etc, just in different ways, and we all live inside different mental constructs that determine how we regard this.
As I said, SAFE is what it is, and there are other models out there, and if you believe those are “better” that’s completely fine, you’re welcome to them. And maybe you’ve learned what you needed to learn about SAFE and that it isn’t for you.
4 Likes