3 ms·
That's not the same at all. You left out an important step in the first model: Energy -> Factory + Labor -> Goods and Services -> Revenue Goods and services i
by ezrast 7y ago
That's not the same at all. You left out an important step in the first model:
Energy -> Factory + Labor -> Goods and Services -> Revenue
Goods and services improve the life of people, and we have deliberately organized our society in such a way that providing goods and services is incentivized with revenue.
That you can also leech revenue for yourself by shuffling liquid assets around while providing no net benefit to the world is a side effect nobody intended. If you happen to reinvest that revenue into something more positive than what the other suckers would have, then no harm no foul, I suppose, but then "Revenue" shouldn't be the endpoint of your model. Something like "Well-being" should.
- fermienrico 7y agoThis is a really good counterpoint, thanks. What, if any, government interventions could be put in place to somehow add value to BTC mining? What are your thoughts of governments using a resource such as hydroelectric dam or wind power to mine BTC and using it to build infrastructure (Goods), healthcare (Service), etc.? It kind of sounds like trickle down economics though, I am afraid.
- ezrast 7y ago> What, if any, government interventions could be put in place to somehow add value to BTC mining? Well, the typical answer to "what can government do about this thing that is bad for society, but not bad enough to ban outright" is "tax it", though as another poster pointed out offshoring limits the efficacy of this strategy. I'm going to interpret your second question as "What are your thoughts of governments using a resource such as hydroelectric dam or wind power to mine BTC and using it to increase their budget?" since in general, money spent on Expense X doesn't remember whether it was generated from Assets Y or Z; money in the budget is money in the budget. That said, I suppose I'm not strictly opposed to the idea, though I'd be interested in an analysis of the returns you can get on those CPUs when you're optimizing for a community rather than an individual - for example, how much benefit does society get from putting all that CPU to work for something like folding@home?