ECWMF's el nino forecast now has all of its ensemble members above 2 degrees C anomalies by end of year, which is pretty yikes

Wicked air quality in Bend, OR today (due to prescribed burns I believe)

5 handles all over the place in Michigan FWIW

I know only enough about atmospheric science/weather/climate to be dangerous, but: it does seem notable that the current ECWMF forecast is predicting at least as strong an El Nino as the April 2023 forecast was predicting (if anything with a higher upper bound).

An AI industry that actually cared about avoiding backlash would have license agreements that set prices based on whether the user intends on automating a role, and charging a per token price such that automation costs about the same as average total compensation for the licensee

If you've ever wondered what its like to hear me ramble about using linked administrative records to study the economic impacts of natural disasters, here's your chance: NBER livestreamed today's PE Meeting, which is on YouTube here (my presentation starts at 2:53): www.youtube.com/live/Aad3Y4A...

Got blocked by Michael Mann, which is a fun fact about today

Apparently doomed to repeat this meme every day on a loop

The lead takes have escaped containment and I don't like it

I would even propose we do a survey of American communities with a big enough sample to represent what's going on in individual neighborhoods

Is it good when the other team has double your score at the end of the third quarter?

Guess I'm gonna start using fuckin' in official contexts now. These fuckin' pretends aren't parallel This fuckin' IV doesn't satisfy the exclusion restriction Etc

Global warming is a slowly meandering ghost

Big disconnect between the "everyone admits current AI models are very capable now and will start replacing human labor" discourse and the fact that basically all AI-related subreddits are entirely composed of complaints about how they can't get any work done with Claude because of rate limits

A large part of the "protein maxing by eating a lot of meat" crowd is just picky eaters who figured out a way to be picky eaters without social sanction; the reaction to anyone else not eating meat is a toxic masculinity friendly way to pitch a temper tantrum about said picky eater tendencies

Inequality is worse than it was 40 years ago

Good story about Claudia Goldin and Economics: www.wsj.com/economy/wnba...

US liberalism and neoliberalism only gave us “on paper” legal equality and rights, not material equality? Well now we don’t even have the on paper legal equality and rights

There is a critical difference between holistic, authentic judgment enabled by embodied cognition, and AI's reductive simulation of the language patterns associated with such thought. papers.ssrn.com/sol3/papers....

We need to bring back the ideal of masculinity focused on quiet self-sacrifice for the sake of others. Alpha shit is for losers. Be a man and humbly forgo the shit you want for years so you can provide for your family, you indulgent little wuss

Please Share! The Environmental Inequality Lab is hosting it's second undergraduate research training program in environmental and energy economics this summer & it's open to all (with current U.S. work authorization)! Full-time, funded, in-person, 10-weeks. More details here tinyurl.com/EEE-URTP

Corollary: 1) Ban corporate owned single family homes 2) poor people now can't rent in high opportunity neighborhoods and also can't qualify for a conventional mortgage 3) you don't have to deal with poor people around the house you inherit

One interpretation: LLMs have learned that IVs are mostly garbage Another interpretation: a good IV requires out-of-distribution human insight that LLMs are structurally incapable of replicating

The fact that vibecoding the same set of tasks in e.g. R vs. python is substantially harder I think is a useful piece of evidence on likely future paths

The administrative costs of doing research with confidential government data (working in secure enclaves, disclosure avoidance, etc) seem like they're going to be the costly signal to maintain a separating equilibrium (in some fields)

About the only thing that dirtbag 19 year old me and current me agree on is that you shouldn't do wars

A lot of the horseshoe discourse around private equity buying single family homes is just a straightforward distaste on the part of homeowners and their children at the thought of sharing a neighborhood with (poorer, less white on average) renters

As we get closer, the population forecasted to be exposed to a foot or more of snow has increased a bit: Around 48 million people are forecasted to get at least a foot with slight probability Almost 26 million with high probability

Rough calcs based on the Gridded EIF and the latest NWS forecasts, over the next 3 days: -At least 46 million people are predicted to experience 12" or more of snow with at least some probability -At least 21 million people are predicted to experience 12" or more of snow with *high* probability

NWS showing 30-40% chances of "Extreme" winter storm impacts (the highest on the WSSI-P scale) for Philadelphia-Boston on sun-tues

I think it's pretty clear that research using any data an LLM can access on the open internet is going to be a gnarly pooling equilibrium that actively decreases our ability to know things about the world; this likely changes the privacy/utility tradeoff for producing public use microdata

I think this is melting snow (which has trapped particulates) + stagnant winds + maybe thermal inversion, similar to what was happening in the Midwest this weekend.

Forever remains baffling that people doing AI related software engineering -- an occupation which did not exist a decade ago that involves any tasks that did not exist five or even one years ago -- cannot possibly imagine new jobs coming into existence, only existing jobs going away.

NOAA historical snow storms are public (www.ncei.noaa.gov/data/snowsto...); we know when the CPS reference week was (the week contain the 12th of the month; we know aggregate CPS response rates. Was January's drop in response rates unexpectedly low compared to other storm-affected months?

My local barbershop has acquired a shop dog, which I have to say I'm a fan of tbh

Severely not liking this shitty air quality today

1) A job is a combination of tasks that are visible to people outside the role and activities/relationships/tacit knowledge visible only to people doing the job 2) If you automate the visible tasks you haven't replaced the job

At least some part of this is the fact that large parts of American life -- pretty much any dealing with the financial sector, every commercial during sports events about gambling, the modal interaction with blue collar professionals -- is either an actual scam or feels like one.

One probably likely future: AI will replace jobs in exactly one profession (software engineering), which is the only job the people working with AI can accurately describe in any detail.

Sharply progressive income and wealth taxation should best be understood as a public health intervention targeted at a vulnerable subpopulation

I'm leaning into being a casual sports guy in my middle age, and the vibes around these parts this winter are extremely atypical -- all the relevant Michigan teams (Pistons, Red Wings, both M+W MSU and UM basketball) are all like, good?

Legalizing gambling and gambling adjacent things (crypto, options trading) was not just a mistake, but potentially will be the cause of the downfall of our current civilization.

Power outage map tells quite a story

The fact of the matter is: a severe #IceStorm is developing across much of the country at this very moment, with potentially devastating and far-reaching consequences extending far beyond the region of maximum #ice concentration. Watch out for the #ice, and help your neighbors weather the storm.

Don't kill people Don't lie Don't lie about killing people

Raw tabulations, intersecting the Gridded EIF with NWS forecast data and electric utility service boundaries -- the ten most exposed utilities. 73% of people in Duke Energy's service area have a high probability of ice or at least 4 inches of snow, 61% of PJM's service area. Seems bad!

10%+ chance of "Extreme" winter weather impacts under the NWS WSSI index for DC, parts of NYC, Charlotte, Memphis

we're getting all the downsides of cyberpunk (social alienation, ruthless hyper capitalism, digital mass surveillance state) but none of the promised upsides (cheap street food, cool jackets, super drugs that make you really fast and strong before they kill you)

Well this looks like fun