Alexander Newton

this corresponds to the expectation of the elasticities. PPML in this case does not go to the arithmetic mean (as the arithmetic mean is a function of x).

@guoxu_econ great stuff from Eric :)

This seems like a really exciting JMP - looking forward to reading it in more detail! ericnrobertson.github.io

@BBCNews needs some actual statisticians on its team maybe? Mental health problems can be both over-diagnosed and under-diagnosed so where is the option for this?

@NathanpmYoung Maybe Obsidian + AI assistant plugin?

I want cursor but for writing. Give me a VScode instance with a nice white instead of black, LLMs tuned for writing text, not code, good speech to text and an LLM that sits over the top. "Reformat the essay so that it has five distinct sections" Anyone doing that?

@NathanpmYoung The stakes could be very high indeed

okay hear me out: e/acc vs EA high stakes poker game. Televised.

@ShakeelHashim I also have quite a big bet riding (hah) on them being in London by 2026 so would be great if Sadiq could listen maybe a little

When I was in San Francisco earlier this month, I took a Waymo for the first time — and got very, very cross that we don’t have them in London yet. I wrote about my gripes (and how to fix this) for the Greater London Project:

@lastpositivist @lastpositivist do you know the answer to this?

@lastpositivist Separate question: what's the position called where probability statements represent how often the stater expects their universe of probability statements at the same probability to come true?

Does anyone know of any psychological work on how people interpret probabilistic claims? like do people generally hear "There is 50% chance of XYZ" as making a claim about uncertainty, frequency, propensity, or something else?

@jessRmorley Did you see Acemoglu's paper on the macroeconomics of AI? It's a conservative estimate but gives quite low macroeconomic boosts to GDP from generative AI

@benhayesnyc @Noahpinion Is it because there are so few nuclear power plants - only 430 worldwide - while there are huge numbers of solar and wind installations? It makes the nuclear line look much lumpier. The largest nuclear reactors in the world each produce like 1% of all nuclear power generation.

@Noahpinion Why does the nuclear line have so much more resolution? The solar + wind chart looks like it is just a continuous curve not real data.

Dear Lord, now that is a chart.

@jessRmorley Lol I remember thinking the exact same when I left. 15 additional exams later with 2 to go... Smh

@lastpositivist I like the masque scene. Earth's increase. Foison plenty, Barns and garners never empty

Even out of context I quite like the lines: "Be not afeard; the isle is full of noises, Sounds and sweet airs, that give delight, and hurt not." Very simple yet evocative of such loveliness. The Tempest actually has many good lines. Much to consider.

@krishnanrohit @economeager Condition was not to tweet about it

@economeager What was the condition?

not even 24 hours later we just got a rejection lmao

@lastpositivist @d08890 Holy shit

@d08890 I mean if there's one field the Irish have excelled at in the 20th century it's applied chemistry tbf

In the name of equity henceforth we'll take all the neurotic white women + all the white guys seething with barely contained resentment and we'll divide academic jobs ~evenly between them and... huh? that's exactly our current distribution? Run the numbers again this can't be ri

@MeiselZachary @alanklement @hubermanlab I think we aren't talking about the same thing. Have a good day!

I strive to get things right on the podcast, and if I don’t, I do hope people will point it out. I re-record audio to correct it and post corrections in the show note captions on all platforms and pinned comments on YouTube. Below is the correction now posted for the fertility episode. At the timestamp related to cumulative probability of pregnancy (2:01:30), I made a math error. The appropriate equation is 1-(1-p)^n where p is the probability of getting pregnant in a given cycle, and n is the cycle number. Therefore the equation for this scenario is 1-(0.8^6). Using the equation, cumulative probability over 6 cycles is 73.8% (rounded to 3 significant figures).