[wc-]

There's a lot of misunderstanding in the comments so far regarding bitcoin mining incentives. I urge you to read this thread describing the split-second load-shedding response time from the POV of one of these miners:

https://twitter.com/ogbtc/status/1699588007664275873

It also goes into the other mechanism by which they make money (being natural sellers of future energy demand contracts during times of high demand). This mechanism is similar to how other commodity markets operate with producers, consumers with steady future demand, and consumers with unpredictable short-term demand.

Our energy grids need to keep an equal demand/production at all times, and on-demand load-shedding is a valuable part of this equation. The bitcoin miners are providing a service to ERCOT and being paid for it. If there were a more "productive" source of on-demand energy usage, then it will replace the bitcoin miners, this is how markets work (of which both energy production/consumption and capitalism in general are).

This is the reality of how the texas energy grid works at present. The bitcoin miners, for lack of a (subjectively) "better" option, are filling the two needs of elastic load-shedding and predictable future demand. The first is very hard to fill, the latter can probably be fulfilled more productively with steady demand from other industries (factories, data centers, other things that run 24h per day).

One more edit: this whole equation changes COMPLETELY if we have the ability to store energy production in times of low demand to be used in future times of high demand (batteries). We don't currently have this at any sort of reasonably useful scale, we need this, and the current "market" everyone is upset about is a bandaid on top of the lack of decent storage options. For the climate folks, the anti-bitcoin folks, whoever disagrees with what I've said here: Fix the storage issue and everything gets magically better. Good luck, it's a very hard problem with very nasty environmental impacts, I'm rooting for you.

[gcr]

Alternatively, a better option: as a human need, maybe energy production shouldn’t _strictly_ follow market incentives and shouldn’t be _solely_ governed by supply/demand dynamics

[wc-]

Totally agree! But this isn't the reality we live in today and you have proposed a potentially better option with no ideas on how to achieve it. What you need to focus on is how we can grow a better grid while achieving prepaid usage levels, guaranteed usage levels, and on-demand response. These factors are what lead to a more efficient, more climate friendly, more better etc etc grid.

[mcdonje]

Saying they're "providing a service" by turning off in peak times is kind of a ridiculous phrasing. Normal consumers in normal markets don't buy at peak because peak pricing is quite expensive. Flipping the equation to paying people to not buy something and calling it a service is outrageous. It exposes the grid as something that literally cannot work on free market principles. People are right to be outraged.

[wc-]

This just isn't how the grid works. Texas has added more than double the amount of renewable energy than any other state grid in the last two years. These investments introduce variable production and require on-demand response to keep the demand/consumption balance steady. Normal / residential consumers do not have a steady demand, don't prepay power usage, and don't guarantee future power usage. All of these factors are what make industrial demand response valuable and necessary for the modernizing, increasingly-renewable based, power grid. ERCOT isn't perfect, I think there are areas for vast improvement, but I think your comment is a little uninformed as to how the grid currently works and will be working in the future as we increase renewable production.

[cozzyd]

Putting the electricity into giant resistors would be a more productive source of on-demand energy usage.

[vel0city]

GPUs are pretty much fancy resistors that also happen to do some math.