(it seems reasonable to ask prospective users to bear much of the risk of these new projects, so if they do prove stranded as overbuilt, the cost doesn’t just get passed onto traditional rate payers.)

I think it’s probably worth decoupling the question of “do we want this new, more electricity intensive economy?” (maybe, maybe not) from “do we want it to be a durable source of rents for Amazon/Google/Microsoft etc?” (absolutely not, maybe some first mover thick margins, but not durable.)

Sure. It’s reasonable (always) to discuss externalities. Plus, even before externalities, this is the kind of casino capitalism boom that tends to overbuild things. (One way to understand AI as it was the tech industry doing what was needed to extend the “cloud” boom.)

dark matter was big when i was young and hale, back in my day.

will be meaty i think.

it makes me sad when i make a new friend then my friend gets deleted.

energy consumption for bitcoin is orders of magnitude more than AI stuff, as best i’ve been able to tell. bitcoin (not all crypto, hashtag) really is a crazy consumer of energy. unless and until we address that, i find the catastrophizing over AI energy use disingenuous.

we have an easier time extending solidarity to dangerous, powerful adversaries than to those who share the same adversaries but who disgust us with views they would be powerless to impose upon us.

all of my acts are official. i should have absolute immunity, citing the US Supreme Court.

it’s working really well. and we feel so morally righteous playing into it.

a hard thing is that, in order to overcome divide-and-conquer by plutocrats, we will have to find solidarity with groups we think are profoundly wrong on some issues, whose positions and advocacy we think are outright immoral.

what is rule of law among people for whom words have no meaning?

(ideally you’d want recipients to feel some obligation to work the sponsoring job, but not so much obligation that if wages and conditions are below recipients’ market value they stay anyway. probably it’s not too far from this? change on Day 2 maybe it looks bad, but after 6 months life goes on?)

maybe Day 3 of 2025 will be better.

“In a time of acute loneliness, the proliferation of AI-generated content seems not unlike an act of pollution, compromising the integrity of the social ecosystem.” @lmsacasas.bsky.social theconvivialsociety.substack.com/p/the-cat-in...

On the collapse of US NIIP. So much for “dark matter”. Yet more testament to the profound but often overlooked importance of pure revaluation in aggregate accounts. www.rabobank.com/knowledge/d0... cc @steveroth.bsky.social

📌

let's do it! the Nobel beckons!

I agree. It's just nothing absolves us from having to do our best to decide how to nudge, despite knowing we are incapable of predicting beyond a few direct consequences how things will work out. We have to act despite limited visibility into the consequences of our actions.

they will, but policy is about shaping how people cross those gradients. of course nothing is deterministic and social systems are things we live inside of and never truly understand. but that does not mean "random" or (much worse) "natural" is wise policy. we have to think and do as best we can.