Emily Grubert
banner
gruberte.bsky.social
Emily Grubert
@gruberte.bsky.social

civil engineer / environmental sociologist. energy, water, climate, buildings, justice. fossil phaseout / universal programs. she / her. bunnies.

Environmental science 38%
Engineering 24%

Oh yeah. Agree and it’s a good distinction!

Yeah and to be clear I’m talking about data as a decision input around infrastructure management vs data for the sake of understanding, etc. I think your analog here is right.

Yeah, though I’ll shout out that we do need a lot of *social* science. Physical data about emissions and climate response are not super decision relevant at this point IMO. Methane measurements are what really pushed me over the edge on this maybe 10 years ago

I am honestly shocked that I have not heard an actual use case that makes any meaningful sense re: climate. like it’s been years and the best I can usually get is “well the models are getting better”

🫡

Oh man I have this same theory about apps but also about things like recycling and the IPCC after about TAR

there are real opportunity costs to how humans spend our wild and precious lives!!!

I just went to a talk where someone described an AI model that can test trillions of degrees of freedom as good. My bro: spend another second thinking about it and I assure you that the decision space is not that wide. I am not impressed that you’re brute forcing this without care.
I genuinely do not care if your model is faster at telling me how high the waves will be under 3*C. The value of that information in actually supporting a normatively defensible decision about climate change action is 0.
I think I’ve underestimated how radical it is to be an “actually, we don’t need more data” guy in the age of AI

Reposted by Emily Grubert

Just FYI - there is a cost curve for RE projects ( a fairly basic / uncontroversial point I think!)

New loads (...looking at you data centers) necessarily push you along the cost curve.

All else being equal, RE used for data centers means more expensive RE for everyone else. Not rocket science.

is ND the windiest coal-heavy state?

exactly!! and then it comes out: the separations step is actually quite different!

As another analogy, it’s like someone telling you you’re overreacting to “mosquitoes, everyone has mosquitoes” without differentiating between 2 and 200 or 2000 of them.

A common turn of rhetoric is to make everything about the number of things vs the size — it’s just one data center (and let’s not comment on the point that it’s unrecognizably large)

my first intense encounter with corporate PR was during the frac boom, when companies made it sound like you were a naive dunce bc we’ve done this exact thing for decades — while failing to comment on scale and key differences. Same thing w AI now, w data centers getting orders of magnitude bigger