[latchkey]

I thought this was a pretty good analysis:

https://www.academia.edu/7666373/An_Order-of-Magnitude_Estim...

tl;dr:

"This means that we can expect our current industry best efficiency of 0.733 W/GH to reach 0.0000000873804 W/GH – so even the most ignorant, arrogant, narrow-minded and pseudo-intellectual critics and arm-chair academics should note that in the event that Bitcoin scales to a million times its current size and market cap over the next 30 years, it’s environmental impact will still be insignificant compared to existing systems."

[wmf]

Regardless of efficiency, there is an equilibrium where energy cost = mining reward.

[tveita]

Moore's law is in effect right now, and yet mining keeps getting more expensive.

The exact number of hashes per joule is irrelevant since the work being done isn't useful. This is purely a competition about spending the most. If hashes get 1000 times cheaper, they'll do 1000 times as many.

[pests]

So maintaining the integrity of the blockchain isn't useful ?

[tveita]

Mining 1000 times more hashes doesn't maintain the integrity any better than the current situation.

[bunderbunder]

According to the third table, corruption, money laundering, and the black market are negligible in the Bitcoin economy.

Fascinating.