Sounds good to me. 
2 Likes
I had a read through the updated whitepaper on Friday, and have been mulling over some elements.
It was great to read through again; the opening part reminds me of how important this network is and how transformational it could be… it’s exciting to see it finally getting towards launch 
Here are a few things that I’d like to note briefly:
- I like the clear explanation for why pay-once store forever pricing model can be sustainable.
- Great to see the clear intent for the Native Token to be developed outlined explicitly.
- Network Node Reserve is an interesting new concept. The potential for ‘token burns’ through this mechanism is an interesting development.
- Nodes referencing market price of token is a big improvement. I hadn’t previously realised how important this could be to market dynamics, particularly in the event of sharp token price rises. This seems to avoid some serious risks.
- Further reduced supply & reduced emissions was a nice surprise
I have a few suggestions / questions:
Native Token:
- Would it be helpful to clearly outline the benefits of the Native Token vs ERC20, and what impact it could have on the network? (E.g. instant confirmation, scalability, low/tiny fees, reduced TX cost from uploads making small uploads significantly cheaper etc)
Network Node Reserve: (I still haven’t got my head around this one yet, so apologies if any of this doesn’t make sense!)
- What determines the level of payout from the pool when uploads are calculated to be ‘below average’?
- What time-frame is used to calculate the average upload cost?
- With it being stated that it’s important in the early years to support incentives for nodes and help provide competitive pricing to uploaders, is an effective ‘tax’ of 18% on payments to store data while the Network Node Reserve is filling up not counter to the incentives that are desired? (though, on the other hand, the tokens being taken off the market will benefit all holders, including node operators)
- It seems there could be significant unpredictability with the reserve switching from filling, to full, and back again. E.g. when full, node payout goes up 18%. If reserve is later called upon, payout will begin, but so will the 18% ‘tax’ on storing data that won’t make it to nodes. Have these dynamics been well considered?
In a follow-up post, I will outline why I think the reasoning for the emissions is still flawed, but I am pleased that the risk of emissions doing much damage has been greatly reduced by the reduction in their scale, limiting to 12 years, and also the new distribution curve (ramping up then down over the first 12 years).
So, a lot of good news with these latest changes, and I’m very excited to see this project getting into the wild. Amazing work Autonomi team!
19 Likes
This is (as usual) great, deserves a longer answer ofc but wanted to say thank you (thank you!) and let you know I’ve seen and that it’s been snapped to respond to 
9 Likes
Before I start this follow on post, I’ll re-iterate that due to the significant reduction of the scale of emissions, I am not overly concerned about their impact on the network, but still want to outline some of my thoughts as to why I don’t think they’re a great idea… also my brain is a bit frazzled, so if something here makes no sense I’ll come back to it another day 
Is the reasoning for emissions sound?
The whitepaper says that emissions are needed to:
- Support the incentive to run nodes in early years
- Support uploaders to provide competitive pricing by subsidising the upload cost
- Ensure and encourage network growth
- Counter the initial price rise in ‘fair storage’ price
I will try to argue that the planned subsidies would be low in effectiveness in achieving any of these aims, and also that the reasoning given for them being needed may be flawed.
Probable low effectiveness of any subsidy
This is a fairly simple point that the effectiveness of any subsidy paid out in the currency of the network will be reduced because it’s emission to node operators is likely to lead to an almost immediate increase in sell pressure, and therefore to some degree negate any short-term impact it is intended to have.
E.g. 10% subsidy = up to 10% more selling pressure from node operators who choose to sell. With that group likely being a significant source of market token supply, it’s likely to lead to a corresponding token price drop that reduces the effectiveness of the incentivising power of the subsidy.
It’s also worth noting that any effect of subsidising upload cost will be eroded by this same effect, as downward token price pressure means people will need more tokens to buy a given quantity of network resources than they otherwise would.
Notes on the Network Node Reserve:
This is also a potential problem for the Network Node Reserve, and in this case it could be more damaging. Because the Node Reserve is intended to be used to boost incentives specifically during times of reduced upload activity & therefore lower token demand, it’s likely to come at a time when the market for the token is already weak (unlike emissions, which are on a pre-programmed schedule). Increasing sell pressure specifically when the market is already weak seems risky and potentially counterproductive.
If the Network Node Reserve kicks in at a time when the reserve is full, not only would there be the negative price impact of increasing market supply, but the reserve would soon cease being full, leading to the 18% ‘tax’ on transactions being reinstated. This would put further pressure on node operators, negating the benefit of the pay-out to some, and perhaps a very large degree.
The reasoning for the reserve’s purpose seems backward to me; in times of low demand (lower than average uploads), surely boosting demand should be the aim, rather than taking action to try to boost supply beyond what the market requires? Could the reserve be targeted at ‘ecosystem development’ actions that boost demand instead of paying out to nodes?
I certainly don’t have a good grasp of the Network Node Reserve, but in its current form it seems to add significant risks, and be an ineffective attempt to solve a problem that I don’t think will be an issue; I don’t think node operators will be phased by short-term luls in demand, and the reserve couldn’t do anything about long-term demand reduction.
Possible flawed reasoning for emissions
If I understand correctly, one of the main threads of reasoning for the emissions is based on the ‘bump’ in the green line on the following chart:
The system must overcome this initial price resistance to benefit from the subsequent decline in price, driven by increasing data upload rates and decreasing raw resource costs. To address this initial friction, The Network will provide additional incentives to node operators during The Network’s early stages - derived from an emission allocation.
The biggest reason given for emissions is to help the network to overcome the upward sloping part of this green line.
Unfortunately, it doesn’t matter what you do with subsidies; that line is going to rise significantly because the ratio of cumulative uploaded data to newly uploaded data / ‘network data overhead’ will be rising for the first few years of the network’s growth until it stabilises.
Any successful subsidy could only shift the curve rightward a tiny amount, but the trajectory would remain the same. A subsidy can’t do anything to significantly reduce the magnitude of the impact this changing ratio will have on the cost of uploading data to the network.
I can see why it’s a bit concerning to envisage how the market will respond to this cost increase over the first few years, but I think the solution offered by the subsidies seems like a futile attempt to reduce the impact of this dynamic.
One way to look at it is the problem is the network adapting to it’s growing ‘overhead’ as it matures to the point of a stable overhead to new-data ratio. The network will find a balance between supply and demand along the way, but the ‘problem’ is that the network is too cheap in the early days to be a good indicator of what use cases will be in the long term.
Framing the problem as the network being ‘too cheap’ initially exposes a flaw in thinking that a subsidy is a sensible solution to the problem at hand.
Summary
In summary, emissions very likely can’t do anything significant to counter the initial rise in ‘fair storage’ price, and won’t be effective at incentivising nodes to do anything beyond the incentives provided by those who are paying to upload data.
Given this, while 20% of max supply is a MASSIVE amount better than the initial 70% planned for emissions, I feel it’d be a lot more impactful to direct these valuable funds towards ecosystem development.
This would likely provide a far bigger tangible benefit to the network through financing high-quality app development, core-network improvements, beneficial infrastructure etc that will spur demand via real usage, rather than the probably ineffective, and potentially market-distorting subsidies to node operators.
4 Likes
How so, the emissions is not like dropping a large %age at any one time. It will only be a low pressure if any on sell price. There will always be those who hold before selling too so they can get a better price in a years or more time.
False equivalence. Does 50 BTC from mining equate in immediate sell pressure forcing the price down.
If anything it will create greater activity around the token and encourage trade, which is an upward pressure.
So far in my reading network reserve has more of a smoothing effect on the market. When its being released people will equally wait till the market is higher and equally not wait more than about 1 month. Remember its not one or 10 people getting a huge windfall, but spread across many thousands who all have different ideas on when they have enough token to sell.
For instance if one average person is running 20 nodes and is getting one token a week, the price is 1$, then its highly unlikely that average person getting 1.1 token for 4 weeks will be running out to sell those tokens tomorrow.
Your analysis is based around the emissions being given out to a small number of people, as small as the large node operators we saw in beta. But in fact if Autonomi is any sort of becoming successful then there will be thousands of node runners, with only a few large players and you cannot predict the actions of the large players since they will 99% be seasoned in the market and NOT sell into a low priced downturn.
As I understand it it is for market conditions and user uploading behaviour. Not so much the price of storage which is what the curve is about. Yes there is interrelationship and I am not sure the emissions is specifically about the bump although it’d help as a consequence.
tl;dr
One reason for the emissions is that early on (18-48 months) we are trying to attract people to be storing their data on an unproven network. The emissions are more about ensuring the network survives during this period by helping the uploader and node operator alike. Without consideration of this then your analysis might make more sense, but no emissions and a network that tanks as soon as the price drops is not a good place to be in. While the network might not die, it might set back the progress by months/years, at least the emissions will aid in not seeing this happen.
Concerning the size of the emissions pool then I’d leave that to more seasoned people and I noticed Dimitar had something to say on this. As to directing a portion to the ecosystem development, then I agree, and I thought there was supposed to be some of that. The partners will be benefiting the ecosystem, but I don’t think you were thinking along that sort of help.
2 Likes
There is no market. There is literally 66k pooled USDC, which is practically 0 liquidity:
We have been here for 10 years, we understand that there is no market and we are locked in until one appears, which may take years.
At this stage, people running nodes do it because they believe in the network. When there is a market, faith will not be necessary. But until then, the believers are mainly in these 2 camps - people protecting their investment by providing resources and people wanting to get more tokens at a lower price than the market price (“buying” them directly from the network, by providing it with resources also known as subsidized mining, which is predictable unlike uploading to the network by people and partners).
For the first type of people, subsidies are not important, they have a certain amount of resources that they will use to support the network and more subsidies will not make them increase the number of nodes.
The second type of people are ready to put more resources into speculation for more tokens. I do not believe there are stupid people here who do not understand that there is no market and cannot sell, i.e. these people are ok to provide resources and wait…
Check out the Dev Forum
2 Likes
Sorry if it wasn’t clear; my posts in this thread are discussing the plans of the Whitepaper for the network economy post-launch when there’s an active & growing market for data uploads that require ANT, many users and node operators etc.
I’m not discussing anything about beta rewards or the current illiquid market for eMAID, so the post doesn’t apply to this stage.
Post launch, the network economy will become a very different beast to what we’ve seen this far.
2 Likes
How do you know? What normal person would pour money into something unproven? What I’m speculating is more likely to happen - a multi-year near-zero market supported by the core of people here.
Whatever our hope, it is wise to prepare for this less favorable option.
Check out the Dev Forum
4 Likes
If you really believe that the offering Autonomi is working on is so low in value nobody will want to use it, why are you here?
Assuming a functioning network that is easy to use so that it’s at least simple to upload & download files, publish websites, and manage token transfers easily, I am confident there will be a growing interest in the network, a thriving ecosystem of apps and developers, and demand for the token due to the valuable resources it offers.
- What normal person would want to use a fully decentralised internet?
- What normal person would want to backup files to a decentralised cloud service with no ongoing costs?
- What normal person would want to share videos / other files in an uncensorable way with no infrastructure?
- What normal person would want to use computing resources to earn tokens that let them do all of the above and more?
There are many normal people that I could imagine making very good use of Autonomi once it’s up and running.
So, I’m working on the assumption of a successful launch… if there’s no successful launch or demand for Autonomi’s services, who cares if it lives / dies anyway, and any ‘subsidy’ in a non-traded token is going to be useless.
If your scenario turns out to be right and there’s close to zero early interest in Autonomi, then it’s kind of a moot point; a small subsidy won’t save the day, or make it much worse.
I’m discussing the economics / tokenomics of a well functioning and growing Autonomi network over a long time frame, which I’m sure are the assumptions the team is working on as well.
3 Likes
There are no normal people in crypto, meme tokens with billion capitalizations are proof of that. The main group of people are speculators and miners, who are also speculators of a lower order. We have countless examples of successful projects supported by these two groups of people.
My personal speculation is that after enough time passes and the network is proven to store data without losing it, things will start to be built and used on it. Let’s say 12 years subsidized is enough to give it a good chance.
Check out the Dev Forum
2 Likes
I expect emissions will cause an almost immediate increase in sell pressure because I expect commercially minded node operators to make up a significant part of the regular flow of tokens onto the market. As they’ll be an important part of the ‘supply’ part of the market, any increase / decrease in the flow of tokens to them is likely to be felt in the market.
The token’s short-term market price will set by those who are actively buying and those who are actively selling, not by those holding & not participating in the market.
Assuming an active Autonomi network, I expect the most active buyers / sellers who will be key in setting the spot price of the token on exchanges will be;
- Users buying tokens to upload data
- Commercially focused node operators selling tokens they earn to fund their operations
- Speculators buying / selling tokens (trading / investments)
- In the early years; sell pressure from eMAID/MAID/Shareholders exiting their positions
Nobody can accurately model what the speculators and early holders will do, but it seems likely that the flow of tokens to the market will to a large degree come from the node operators who choose to regularly sell some of / all of their earnings.
Due to the nature of the Autonomi network primarily being a market for resources, I expect the flow of the tokens to node operators will be very big as a portion of market cap vs the scale of flow to miners / minters etc in other projects, where it’s only block rewards to play for. I expect that this means the sell pressure from the node operators will have a bigger impact on the market than miners/minters have on a token market in other projects.
You can see in my quote - it says ‘up to’.
And, yes, of course; if the Bitcoin block reward suddenly jumped 10% through a subsidy I would expect an immediate increase in sell pressure from miners who are selling a portion of their block rewards.
If increasing token supply would add upward price pressure to price, why not add a 5000% subsidy? At what point do you think increasing token supply would start to lead to downward price pressure?
Interest in buying the token won’t come from releasing more tokens to the market, but demand for buying network resources and speculation.
What makes you think this? I’m assuming a vibrant and growing market for resources, with many thousands of node operators, and many thousands of users.
My analysis does focus on the outsided impact a sub-group of node-operators (those who regularly sell a portion of earnings) will have on the short-term token price, but even they are likely not a small number of people if the network is growing nicely.
Seasoned or not, many large operators will need to pay bills, and that may mean selling a portion of earnings whatever the market is doing. Hopefully most will have a buffer and avoid selling into downturns, but still, increasing supply to a market at a time when demand is already lower is likely to put downward pressure on price.
Can you explain how the emissions to node operators might make people more likely to store data on Autonomi who wouldn’t if there were no subsidy?
Can you envisage a situation where the network wouldn’t survive purely from data payments from paying customers, but a subsidy would make the difference?
What scenario do you envisage that would lead to this, and how might emissions change that scenario? What kind of problem could the network be in where a bit of a dilution of the supply is the solution?
In all of these scenarios, I am assuming there is a functioning market for network resouces; I understand why the incentives work in the beta program, and why nodes drop off when the incentives stop, but in the live network it will be a very different situation.
Yeah, I was thinking more of directing it to developers in some way that reflects the desires of node operators, and probably with a role for the foundation in that too (but ideally not holding all of the tokens, or having full control over where they go). Partners are fine, but I expect their role will be more influential at the very start of the network, reducing in influence as the network and ecosystem mature.
1 Like
I expect that a live Autonomi network will be very useful to many people once it’s live with good UX and apps will be significantly more interesting to people than a beta network that you can’t do anything with.
It’s also a hugely different dynamic when you can pay for network resources, and earn rewards for supplying network resources, with a token that can be traded for anything you want (FIAT/BTC/ETH).
How do you not expect that to change things a bit from the experience we’ve had until now?
Dabbling with something isn’t the same as pouring money into it. There are many token holders who would probably like to use some of their tokens to upload data. Once a few cool use cases come up (e.g. video / music / other file sharing, backups, websites etc), more and more people will dabble & explore.
While the longevity of the network won’t be easy to have much faith in during the early months and years, the network can still be hugely useful as a decentralised data network even if the data didn’t last forever.
Some Defi projects get massive volume despite being a risky proposition until proven for multiple years, and many had a very different experience to a multi-year, near-zero market for their first years in the wild.
I think you’re way too pessimistic.
I expect that even if the Autonomi network starts small, with a live token and market for resources, the supply & demand could find a balance that works for the market participants at that time.
So, even if the network stayed small for a fair time and growing organically after that, there’s nothing in that path that would be unsustainable, and nothing that subsidies would ‘fix’.
Why do you think that a small network wouldn’t be sustainable, and how would emissions fix the problems that you envisage?
Running as a side chain on Cardano would solve the unexpected fees, security, scalability and privacy issues associated with erc20.
2 Likes
Subsidies make storage space near-free if they can pay operators’ bills, so storing data for free will attract way more users, than if you have to pay for an unprooven service.
It seems that now you are in perhaps, maybe, best estimate of mine and no solid foundations to your reasoning.
If any of your basis “facts” are wrong then your reasoning fails. I would hate the future economics being based on such a thread of assumptions.
Firstly the network is being targetted at home users and assumptions must be made that they will make up the bulk. And as I said some one getting a token or two a week that even getting double will not be going out to sell quickly. Thus your reasoning is flawed in this area.
Even if we take your assumption that it will be commercial people then they will be seasoned and will NOT sell into a market that is dropping or selling at a fast rate. Also the emission is not that great and simple market activity with this emerging network will outweigh any emission that represents a tiny amount at any point in time. Its not like they are getting 100 or 1000% extra. More like 1%
Like I mentioned we do not see the block mining rewards that BTC generates affecting the trading price, but more what people are using BTC for affecting the market. Also for Autonomi the fact you can store data and people wishing to do that will drive the market 100 times more than any 0.1% emission rate, just like you see with BTC emissions.
Now you are using rose tinted glasses blocking he obvious. Just like supermarkets drop the price on items to get them to buy them, or buy 1st at 10% off and rest at 100%
Only leads me further to think you have immersed yourself in this rabbit hole and cannot see what others see so clearly And using reasoning that sounds like emissions will represent a huge %age (like 50% more) does not help.
2 Likes
There certainly would be more demand for something free than if something had to be paid for.
But if the network hypothetically had such low demand that uploads needed to be close to free to get people to use it, that network’s token price would probably be close to zero, in line with the low demand for network resources, so paying these low/no value tokens out to node operators wouldn’t pay any bills or incentivise much behaviour.
I’m surprised to hear all of these highly pessimistic assumptions of very low demand for Autonomi’s offering from this community, and I fully expect there will be people wanting to upload stull to the Autonomi network as soon as there are some decent apps that make it useful.
One reality that could help offset the issue of early risk due to the unproven nature of the network is the fact that uploading data will be cheaper in the early months and years due to the initially high ratio of new-to-old data.
So, by design there’s a ‘natural subsidy’ caused by the fact that nodes will have a smaller ‘overhead’ in the early stages of the network, steadily increasing until the radio of old to new data stabilises.
This is what’s shown in the upward sloping part of the green line in this chart:
So, even without emissions, uploading data will be cheaper in the young network, and increase gradually as it matures, coinciding with de-risking through the network’s growing track record.
2 Likes
I agree that home operators earning a token or two per week are not going to make up a significant part of token sell pressure on the market.
You have just asserted with no reasoning that “assumptions must be made that they [home node operators] will make up the bulk”. Below I will explain why I expect commercially minded operators will be more significant than you are assuming.
I do expect commercially minded operators will be important for the network. The evidence from the beta testing provides very strong support for this, as a small number of motivated individuals spun up huge numbers of nodes, using remotely hosted servers and beefy hardware in professional level home setups.
There are 2 very important reasons why I expect commercially minded operators will be important for the price of the token and for the network in general:
- Commercially minded operators will be the marginal suppliers of network resources
- Commercially minded operators will provide a regular stream of tokens to the market to finance their operations
Point 2 I’ve already made, and it also follows from point 1.
I believe Point 1 is likely to be the case because in response to any increase in network demand, the price will begin to rise as nodes fill up more and token demand increases. When this happens, the average home user will not be likely to quickly increase supply through buying new hardware etc to match the demand increase. Commercially minded operators will be the ones who spin up nodes at a rate that is required to match demand in line with the market incentives provided.
This shouldn’t be controversial; the operators who are commercially minded, whether small or large, will have the capability and propensity to add supply in response to demand increases more rapidly than home users will, meaning they will play an important role in the ‘supply’ side of the market for resources.
Due to their actions in supplying resources to the network, they will be earning a lot of tokens, and will therefore be selling a lot of tokens as many will have operational costs to cover. They will therefore be very important on the supply side of tokens to the market.
Basically, as soon as demand for network resources outgrows any organically provided home-node capacity, commercial operators will take over as the marginal suppliers of resources to the network… people will respond to incentives, whatever anyone intends.
I guess it’s possible that organic growth in home-nodes keeps the price level so low that commercial operators don’t have sufficient incentive to operate. We won’t know for sure until the network launches, but I’d be surprised if growth in the network wasn’t demand-led as soon as it starts picking up some momentum.
My question was not whether people might upload a bit more data if the price of uploading were subsidised effectively (that’s obvious), it was intended to highlight the unlikeliness that someone would decide to become a user of the Autonomi network or not based on small differences in the cost of uploading. Pricing may affect how much data they upload, but if they have a use case and values that make the network a good match for them, a small price increase/decrease will not determine whether they use the network or not. If this is the case, subsidies would be unable to, as you suggested: “attract people to be storing their data on an unproven network”.
I think you’re great with the techy stuff, but almost every time we have a discussion on economics you seem to quicky move away from actual arguments about the concept and start mud-slinging and making assertions & the discussion gets messy.
Instead of:
… just point out what you don’t think I’ve reasoned well, and ask for clarification. I’ve spent a lot of time providing reasoning. If you don’t think something works, question it and try to provide a challenge that moves the discussion forward.
Instead of:
Ask whether you missed the point I was trying to make.
Instead of:
Explain what you think I’m missing and try to clarify with constructive questions and presenting well reasoned alternatives.
1 Like
You are missing real examples of working projects. We know subsidies work because we see how Filecoin does with 70% subsidies vs. Storj with 0% subsidies:
https://dashboard.starboard.ventures/dashboard
Check out the Dev Forum
5 Likes
There are so many examples of working projects that offer services in exchange for payments that cover their expenses with no requirement for subsidy. For example, OneDrive, Google Drive, any other cloud storage provider, any web hosting company worldwide etc etc.
We know markets for computing resources can work without subsidies because of the many markets for computing resources that work without subsidies.
Even if there were no close real world examples, discussing the economics of a proposed system can still happen using well established economic principles around markets, supply & demand without needing to back it up with examples, though of course relevant examples can help where available.
Do you have any reason to think Filecoin’s subsidy is an important part of it’s success vs Storj?
Is the fact Siacoin has subsidies too, but is even less successful than Storj relevant?
2 Likes
Sorry, in the day, there were “huge” subsidies at the launch of Storj for miners $$$ - the paid rates were high, there were surge payments, and Storj pumped the network with data to generate the traffic to make sure miners had something to store and were paid - there was even free storage for users of 50G a month, needless to say that got abused and has now been decommissioned, but it bought traffic and people to the network. This was all paid for from time-locked ERC-20 token tranches from TGE.
Today you are correct, there is 0% in subs, as the network has paying users who now cover the miners fees.
Subs and emissions are needed in the early days of Autonomi, to grow the network, and stabilize the token from my view, others will disagree, and that’s fine.
5 Likes