This is pretty interesting from a economic standpoint. The USD price for file space would likely hover around the same price of something like S3. If the price paid by the user is too high, they are better of using S3 for storage. If the price paid to the user is too high, it creates an arbitrage situation in which someone can simply use S3 to host Filecoin data. This should theoretically provide more stability to the price of Filecoin in comparison to other virtual currencies which is simultaneously the biggest strength and weakness of most of these things.
Filecoin providers may feasibly beat S3 on price in USD if (i) they're betting on the protocol because up-front losses may be more than compensated for in investment returns on Filecoin or (ii) the alternate for them is making no money on unused disk space.
The problem is that the investment possibilities for Filecoin might be limited since it will be tied to a real world service unlike most other digital currencies. People are free to speculate how much Bitcoins are worth because there is no company out there saying "Bitcoin are worth X". Filecoin will basically be saying "Filecoins are worth X GB per hour". The way to combat this issue is to tie the price to some other currency (e.g. X USD or Bitcoin per GB per hour paid in Filecoin), but it would be tough to convince other people to use your currency if you aren't even confident enough in it to price your own products using it.
For your second point, we need to keep in mind the free space might be empty but it is not free as in beer. Using the disk will cost extra electricity and CPU cycles. Plus we shouldn't forget that disks are still the most common piece of hardware to fail inside a computer. Increasing usage of that disk increases the risk of failure.
>The pricing is left fluid precisely to avoid this. The price can change over time. 1 Filecoin / N time today. 0.1 / N tomorrow.
That makes sense from an economic perspective but how does that work from a customer service perspective? Is the price pegged to a different currency? How would a user be able to predict the cost of the service if the price is not pegged to anything?
How can you predict the value of a dollar? It's not pegged to anything either.
Although in fact the decline of the value of the dollar is relatively predictable because of the Federal Reserve's policy goal of keeping inflation relatively low and the law of large numbers as applied to the generation of the real rate of inflation.
At my local deli a dollar used to buy a bagel with the works; and now buys less than half of a plain bagel. That's over the course of 20 years. Any unit of exchange is valued anew in each transaction; the persistence and trends in value exist mostly due to social pressure and the cost of bargaining. I don't make offers and counter offers on each bagel I buy; I just look at the sign and accept the price or leave without a bagel.
> Although in fact the decline of the value of the dollar is relatively predictable because of the Federal Reserve's policy goal of keeping inflation relatively low and the law of large numbers as applied to the generation of the real rate of inflation.
That's a crucial point. If the Fed didn't pledge to keep inflation under control, the dollar wouldn't be such a great store of value or medium of exchange. It's the stability promised by central banks that allow you to "predict the value of a dollar" in the immediate future.
If Filecoin works, the price to store, say, 4 TB of data will be bounded by the cost of a 4 TB disk and the cost of 4 TB of S3. That's a wide range, but it's not as bad as Bitcoin's 1000% fluctuations.
>The problem is that the investment possibilities for Filecoin might be limited since it will be tied to a real world service unlike most other digital currencies.
That certainly hasn't dampened the investment possibilities in actual, government-issued currencies.
> The way to combat this issue is to tie the price to some other currency (e.g. X USD or Bitcoin per GB per hour paid in Filecoin), but it would be tough to convince other people to use your currency if you aren't even confident enough in it to price your own products using it.
You can't peg a decentralized currency. You need to be able to control supply in order for the peg to work (ie. to sell more into the market to lower the price, if needed), and then it wouldn't be decentralized.
What if both the price paid by the user and thus paid to the miner is lower than S3, how would you arbitrage that? What would drive the price up to s3 prices? Scarcity?
There wouldn't be a direct arbitrage opportunity if the price was lower, but we have to remember there is a reason why S3 has been so successful. HDD space and CPU cycles aren't free. Plus there is the redundancy issue that others have mentioned. Amazon can help address this because they control the servers and can nearly guarantee their uptime. Filecoin doesn't mention any requirement of uptime for its users, so it will likely need to combat the problem by just throwing numbers at it. We then have the user's costs (electricity, wear and tear on their machine, and opportunity costs of doing something else) multiplied by whatever number Filecoin feels is enough to guarantee redundancy.
This is a long way of me saying I am not sure how much lower than S3 pricing they can realistically accomplish.
