Thanks, that's a little boring to watch since I already know what to response (I watched the prior chat about the conclusions and a little of the talk itself)
He's making one fatal error in his assessment: Bitcoin mining scales linearly with market cap.
The amount of value representing Gold & Fiat Money aren't considered in his comparisons, but they should.
The ecology should be represented in value per energy not absolute energy. The amount of value both in market cap and transaction value is so much higher with Gold and Fiat money that it distorts the conclusions he makes. I also don't think that energy efficacy will improve to that extent (it won't tend to zero if Bitcoin remains important). So far the power consumption has been increasing linearly with market price and more efficient machines simply did only offset it during the transition with even higher power consumption overall in the long run.
I'd like to see a trend of power efficiency in [Transactions/Joule] and see where that is headed. I doubt you can extrapolate an asymptotic trend. I think it's linear perhaps slightly logarithmic.
The last point is you wouldn't want to exclude the manufacturing costs of mining equipment nor the cost of exchange infrastructure if you compare it to the operational costs of banks.
Making sure that I'm following you: Are you saying that if the Bitcoin market cap increases to the level that fiat money or gold currently maintain, that you think the total amount of energy required to secure the system could never be reduced by way of efficiency gains to have it be relatively similar to the total energy that is used to secure those other stores of value? Just putting this out there, but this phenomenon of increased energy efficiency developments, whenever involving something that humans value, has historically caused the total amount used to increase with most, if not all, such resources.
