feel free to show that calculation and cost for 5 copies of storage, cpu, bandwidth and electricity over lets say 10 years. Don’t forget to calculate for accumulated old data, already payed for, for every year that will spread across the network.
I would very much like to see numbers.
I’m sure an actuary could calculate a price for you. I suspect in the future folks will provide the calculations you require. I’m not an actuary though, but then I’m happy to use a spare laptop and bandwidth either way (I’m not using either for anything else).
For now, rest assured that folks will guestimate. They will decide what price they will run a node for and the point they will stop. The network will adjust the storage cost to ensure sufficient numbers of people run nodes.
Here’s a strange number - the human DNA is pay once store forever.
Is it really forever against the infinity of the universe? No, but practically for human life span it is practically forever. It will be the same with the network - whether this forever lasts 5-10-100 years before it is replaced by the next eternal network, for your life and mine it will be practically forever.
Stoicism has a very good concept - don’t worry about things out of your control.
Privacy. Security. Freedom
The problem is I have given him the figures in days or weeks past elsewhere and they are handwaved away because of a lack of experience and understanding of pricing in the computer storage industry. Even in reading graphs on storage prices he didn’t understand what he was reading and picked one figure that suited his bias
So as far as I am concerned I will not play the game of repeating myself time and time again for the pleasure of some game being played.
But I will counter him to others if there is an attempt to change the course of the project based on his misunderstanding of basics of storage technology/pricing.
At this stage the onus is on him to show hard figures and projections since he is at odds with the current trends and technology experts out there. But he uses non-spare resources, over priced figures, etc in any of the handwaving, and the poor results from the bumpy start to beta
I’ve mentioned this before, but will repeat … I have reservations about the pay once model. I don’t think there is any calculation for it … I think it relies on many things and some of those things could go wrong.
But what people forget it seems, is that the pay once model is part of the code … and the code is not static. The network is composed of people who will protect the value contained within it.
Therefore, if the model starts to fail the network will change the model by adapting the code.
Hence worrying about this as any kind of show-stopper isn’t worth the trouble. Let’s wait and see how it works and if it turns out that it doesn’t then we will come up with solutions as we need.
The pay once store forever* concept is perhaps one of the easiest parts of Autonomi to get right.
Its the concept of paying upfront for future costs.
The actual cost of storage is a reducing series like the frog thought experiment of a frog of no size that jumps half the distance each time, how far will the frog end up at infinity?
So answer to yourself how far does this no size frog jump? Will it be infinity distance? or is it mathematically proven to be so close to twice the distance of the first hop.
As pointed out @Traktion this financial concept is built into a lot of things.
Now it is proven (by experience & known advancements to come) that for the foreseeable future barring war, global destruction, and such extreme things, the price of storage is dropping each and every year/cycle of replacing storage. And like the frog thought experiment this ends up being a reducing series with a limit it approaches. So if one charges up front more than that limit the cost to store today’s data forever* then that cost is covered. Upto about 3 years ago it was around 10 times every 5 years with ups and downs. SSD in last 3 years has dropped to a 1/3.
In this explanation there is no need to say what the current price is since you only have to pay the amount needed to cover future costs. For the magnetic disk series of 10 time every 5 years the raw cost is 1.1111111 times the current cost. The cost to run the drive is the same if you keep the old data or discard it in 5 years. But its like dollars a year
Now Lets come to Autonomi. The beauty here is the network is purposed to run on spare resources and to consider new devices and data centres is not part of Autonomi’s working SOE, so I will not, any analysis is above.
For Autonomi using spare resources the costs can only be talked of in terms of incremental costs and not raw costs. Anyone changing from spare resource use (same h/w, same hours they normally use) and going above that will have reduced earnings vs costs.
For incremental cost we have less than a watt for most consumer grade PCs, less for laptops and even less for tablets. For big powerful PCs we might have a few watts because they want to push limits of spare resources.
At 5 watts for 24/7 its 0.12 KWHr per day and 43 KWHr per year. For the absolute worse case and consumer grade PC it is 1/5th of that or less if laptop/tablet
There are the rough figures of the costs to run Autonomi the way its meant and designed to run
But as an aside, I have a Odyssey, a slightly more expensive SBC computer but much better than RPi. It uses 5-7 Watts MAX, can run 50 nodes. So electricity costs is around 16-25 AUD Raw Total price per year. So all it needs to earn is 32 to 50 tokens a year at the current price. And I use it for other tasks in my network. You can look at it as me running it for free and eventually even paying for its upfront costs and SSD replacements OR I pay the costs and I earn the tokens for free. Take your pick.
When you do this work for a living for so long and doing quotes blah blah you learn a few things. I just want someone to pay me to put lasers on shark heads. Damn it I want sharks with lasers
And please I don’t want to have to repeat this exercise yet again because someone wants to play a game and say prove it. Not you Taylor, but if someone disagrees with this then make sure you are using the spare resource model Autonomi is designed for and make sure you actually do the research and not take a graph of rotating media thats dying off and claim see it doesn’t work. Yes I am pissed off at ignorance demanding proof over and over again. And be assued Taylor it is not directed at you.
I’m not asking anyone for proof - I don’t require such. But I do disagree that proof can be known, and I also disagree with @Traktion 's examples. But I see no reason to debate it. We will see in time if it works or not and if it doesn’t we will change it. It doesn’t stop the network, it doesn’t alter it’s value proposition for me in the least.
So, I’ve nothing more to say.
Well if you want the proof then its in the analysis I gave.
And it goes without saying that it assumes a couple of things
But its OK to remain questioning, and if you want basic proof then its above.
And No I am not trying to control anything here I actually want people to understand the sheer beauty of forever* data storage for ordinary people and educating others using my decades of experience in this field. I kind of feel real bad that it came across that way
Nice work on the calculations!
I know it is obvious, but unused storage space can also be considered a sunk cost. If you have a 1TB drive and are only using 100GB of it, the other 900GB is essentially wasted resource.
Having Autonomi taking that idle - and already paid for - storage and putting it to good use is a net benefit. The few pennies spent on electricity should easily be covered by the value added by putting the spare storage to work.
It actually reminds me of how AWS was born. All those spare resources, just waiting to handle peak demand for sales/christmas etc. Selling those idle resources has turned Amazon into what it is today. That’s what Autonomi can do for consumer devices too.
Yeah, my joke was kind of bad. It was more of a feeling of how people think they can know enough of the present that they feel they can predict (and so control) the future. Fundamentally, this is not rational, which is why we have markets and not central planners.
I provided annuities as a common example, but really insurance covers all sorts of things from the somewhat predictable, to the highly unlikely. They then turn this into a policy with a price on it.
In many ways Autonomi is much more predictable than, say, the weather and the damage it can cause. Even life insurance or annuities can be impacted by freak deaths or unusually healthy folks living into their 100s.
The point I’m trying to make, is that actuaries put a price on things which are much harder to calculate than the evolution of technology. Indeed, it is probably more likely that a breakthrough will make those predictions conservative.
Ofc, we’re not all actuaries and don’t need to be. Big Autonomi hosts may want to consider their minimum price carefully, but for consumers using spare resources, its likely in the noise anyway.
We definitely need to do estimations in order to consider whether something is possible or not … otherwise writing something in code may end up being futile. I’m simply arguing that there’s no perfect knowledge and we don’t need such either - we can evolve.
There is a saying that is along the lines of he who understands the past can know the future.
If you go back 10 years then the prediction for the future of storage was the same, go back 20, 30 years and its the same, 40 is less sure, but it was the same. SSDs though have accelerated it
The forever* storage is the easy part like I said, the hard part is enabling the spare resources to be utilised without effort (mostly) and also the protocols etc.
I understand that people have instinctive trouble understanding how costs to do something (store old data) can be reducing over time just like the frog example.
So I understand uncertainty with that concept, but once you go through it for yourself its a game changer. Big business are doing this all the time. Cloud storage tell you its more and more expensive but that because they have shareholders to give those savings to. LOL
I hope, for your happiness that you are correct, but if not, we’ll sort it out.
lol - this is why I was reluctant to enter into a debate here - whether it works as is or not just doesn’t matter to me, so the argument isn’t helpful. I sincerely hope your views and numbers are correct. I’ve only been arguing for the crowd in this thread that whether they understand your numbers or agree/disagree with them … it doesn’t matter because a network isn’t a centrally controlled entity that may be doomed to fail - it’s almost a living thing in that it can and will evolve to fit it’s niche.
Very true. The algorithm has to account for much.
Why is this discussion in the price thread?
Because they’re talking about price of storage? 
Jokes aside, I do think it’s better fot this topic to get it’s own thread (if there isn’t one yet)
If it’s such a big deal for you, why not spend an hour with a spreadsheet to have a stab at it rather than rehashing your unsubstantiated doubts.
If you factor in the falling cost of storage, bandwidth, and processing power over time, and the fact that old data will be less active on the network (using lower bandwidth & processing per month than newer data), I think you’ll find it’s not as terrible as you expect.
Also, imagine the impact of archive nodes for old / rarely accessed data, which has been proposed (though I have no idea how that would work).
But, I do think it’s analysis worth doing, which could show up some potential pitfalls to cover for.
I would give it a shot if I had time, but I’m moving home & travelling so can’t do it for a while.
as far as i remember the concept of time does not exist in the network, which makes the recurring costs model kind of impossible.
also having perpetual recurring costs would mean that the network would have to be able to delete data that isnt paid for anymore.
On the cost front for people wishing to run 24/7 my machine costs maximum EUR 10 per month. Comparing that to the price of a DropBox subscription it is a small price to pay to support a network that has permanent and almost unlimited storage. It will be unlimited because we are not all as unique as we think, we have a lot of data duplication between us humans.
The deduplication of data is part of why I think store forever isn’t a difficult concept to accept. When this picks up we will have a variety of people keeping nodes running, the people running it for ideology of freedom, people running it because they want their data secure and safe, and don’t forget the people who are enthusiastic about the financial side of things. With the lighter load of deduplicated data across all of the above I think we’ll be just fine.
I’ve heard one or two ‘experts’ accept that you can run a NAS at home, have one disk fail and rebuild the data from the remaining 4… who then have trouble accepting this project because what if one person turns their machine off… personally I think the only reason they are having trouble with the paradigm shift is the amount of shares of Amazon they have invested in or their other IT intrests - which this time next year will bite them in the ass for not taking the risk free opertunity to take part in the launch.
This project is the first of its kinds so it’s going to be interesting watching this play out.
Can’t wait to drop DropBox, and Google storage. My google storage was meant to be free forever but things changed.