[lmm]

Fiat is just an IOU, it's worth exactly what it's worth. There's a small admin cost to e.g. printing notes, but it's tiny in proportion to the overall currency value. Bitcoin is unique because it has a big negative externality - a huge amount of energy is simply thrown away, powering computers to do something non-worthwhile. A bitcoin is an IOU for something that's already been spent.

[joe_coin]

I would point out that you're not making a fair comparison to the cost of administering a fiat system. In the US, it includes the cost of the entire Federal Reserve apparatus dedicated to managing the supply of FRNs. Treasury's printing of them is only one small part. The Fed has buildings to maintain, support employees to pay, behind-the-scenes political processes that play out and who knows what changes hands there in exchange for appointments and other policy stances.

You're right that Bitcoin by nature includes a huge energy cost, but when you take the narrow view that it's simply being thrown away, it's both short-sighted that it ignores the theoretical possibility that its heat "waste" could be channeled into something economically desirable (e.g. steam to drive turbines), as well as not an even-handed to comparison to the true costs of its alternatives.

[lmm]

> In the US, it includes the cost of the entire Federal Reserve apparatus dedicated to managing the supply of FRNs. Treasury's printing of them is only one small part. The Fed has buildings to maintain, support employees to pay, behind-the-scenes political processes that play out and who knows what changes hands there in exchange for appointments and other policy stances.

Sure - the physical printing is just one of the many overheads of the monetary system. But the point is that these costs are small and, importantly, more-or-less constant. Whereas with Bitcoin the costs scale up (and inherently have to scale up) with the Bitcoin economy. Indeed, if we assume rational miners with accurate price predictions, then the amount wasted on mining Bitcoin always exactly equals the total value of bitcoins in circulation.

(So far we don't see this because the price is much higher now than it was when most of the coins in circulation were mined. On the other hand, if the Bitcoin price drops to e.g. $10, then the current expenditure on ASICs is going to look even more extravagantly wasteful).

> You're right that Bitcoin by nature includes a huge energy cost, but when you take the narrow view that it's simply being thrown away, it's both short-sighted that it ignores the theoretical possibility that its heat "waste" could be channeled into something economically desirable (e.g. steam to drive turbines)

Even if you find a way to use the heat, you're wasting the computational power that could be allocated to something more productive e.g. protein folding.

But even more importantly, and what I expected to be half the point of the article, the work put into validating the blockchain inherently has to be wasteful. If it isn't, the economic incentives against cheating no longer apply - you can just do a leveraged buyout of a huge turbine company, put their heat generators to work doing a 51% attack, and double-spend all your bitcoins.

Whenever someone comes up with a way to do hashing for half the energy, the Bitcoin network doesn't start using half as much energy - it starts doing twice as much hashing. The wastefulness is built-in and unavoidable.

[Terr_]

To play with that idea and rephrase it: Fiat-currency is backed by a potentially-repayable debt held by a government, the original credit having been spent on infrastructure, social services, weapons, etc.

Bitcoin is backed by an irrecoverable debt held by... an abstract math problem, because the original credit spent on computing answers (of no external use) to that arbitrary problem.