The work is essential to securing Bitcoin transactions, and thus Bitcoin itself.
This argument has traction only if racing to solve useless hard problems ("waste") is shown to be essential in securing distributed currencies.
While I won't deny for a second that 'Satoshi' had a deep insight, his method is not by any means the only way to achieve Byzantine fault tolerance for the coin database. The literature contains extensive proofs for algorithms that can withstand up to one third traitor nodes without use of PKI, and unlimited traitors with PKI. These algorithms might not be practical from a communication overhead stand-point, but they don't lead to a wasteful arms race.
The second claim for the need of waste is to provide artificial scarcity. It's easy to envision a system where scarcity is generated by a central bank which digital signs monetary tokens; the challenge would then be to either distribute such a system in the absence of a central bank, again with Byzantine agreement on the number of minted tokens and their initial owners.
As it is, the waste should be accepted as a challenge of the current Bitcoin implementation, not embraced as the only imaginable solution.
3 hundred thousand tons of steel ?! For a toy internet currency that, aside from being a vehicle of speculation/investment, is only accepted by a few hundred tiny sites ?
Maybe some scale is required. A phrase from 'Satoshi' in the initial version of the FAQ:
When Bitcoins start having real exchange value, the competition for coin creation will drive the price of electricity needed for generating a coin close to the value of the coin
Let's assume Bitcoin becomes hugely successful and manages to displace the US dollar tomorrow. The total quantity of dollars that make up the
monetary base (the most liquid money that bitcoin would replace) is on the order of 2 trillion, so if their value is substituted with bitcoins, the 12 million bitcoins generated in the next 10 years (
6 million -> 18 million) will be worth today about 1.3 trillion $. If we equate that value with the price of electricity as per the quote above we get 26 trillion KWh at current wholesale prices. That's more than the
entire energy production of the world in a single year ! During the ten year period, minting bitcoins would require 70% of the electricity production of the United States !
While I agree Bitcoin will not replace the dollar, and it will certainly not do so tomorrow, using the correct scale shows just how bad a design is. A smaller Bitcoin is not less bad, just bad on a smaller scale, a local toxic spill as opposed to a full blown Exxon Valdez.
Mind you, I have ignored the hardware requirements which will likely dominate mining and that have more important impacts on the environment than the electricity consumption. In a real scenario, the electricity consumption will be lower, while the environmental impacts will be higher (manufacturing is more damaging than electricity production for a given revenue level; electricity can come from nuclear, hydro, etc. while copper can't be extracted without carving up some mountain).
I challenge anyone to reflect on how all this compares with the current financial sector, and prove how Bitcoin can be seen as an improvement.