After several years of interactions with extremist centrists I wrote this piece:
On the general futility of political discussions with people
https://t.co/k44uHZTUVP
and this is confirmed every time you interact, even indirectly, with them:
They read nothing, impute everything, and hate everebody.
Thanks, Joshua -- quite interesting. And the result completely confirms my experience so far.
1. Overall, it's just a long-winded methodological summary telling us what to do, but not doing it itself.
2. There are no concrete results, which would include posting the dataset in free access, showing how it was analyzed, and summarizing results with p-values or equivalent. Compare to what a proper cliodynamic article does, e.g. DOI: 10.1126/sciadv.abn3517
3. In fact, some results from our article are reflected, e.g. identification of military technology and agricultural productivity as most important technological factors.
4. Generally, it's pretty accurate, if bland: "semantic ablation" describes it quite well.
5. I stand by my assessment in the post that started it all: 'So far, their answers are pathetically unhelpful -- essentially summarizing "common wisdom" from the internet or simply hallucinating.'
6. As an example where AI got it wrong: it claimed that wage compression raises instability risks. In fact, the opposite is true.
7. Conclusion: "AI is not capable of generating truly new science. At least, so far."
After a long break, I'm back to devposting! I argue that the development econ focus on "institutions" was a mistake – not (just) because the evidence was bad, but because the concept of "good institutions" is tautological and uninformative for policy.
Hoy en mi clase (licenciatura) les recomende a los alumnos que trabajen alguno de estos tres libros:
- Into to Stats Learning (Garrett et al): cuatro pajaros de un tiro: 1) repasar econometria, 2) aprender ML, 3) aprender R & Python, 4) ver como andan con el ingles 1/n
[RECURSOS 💿] - Los mejores datasets del mundo están a 3 líneas de código de distancia.
Tu guía completa de paquetes R para DATOS REALES
Kit de paquetes que te conectan a bases de datos del mundo real:
🌍 ECONOMÍA GLOBAL
📦 {wbwdi}
→ Banco Mundial | +24,000 indicadores
→ Ejemplo: PIB per cápita de todos los países desde 1960
📦 {OECD}
→ Países desarrollados | Empleo, inflación, comercio
→ Ejemplo: Tasas de desempleo comparadas
📍 GEOGRAFÍA & POBLACIÓN
📦 {tidycensus}
→ Censo USA | Hasta nivel de manzana
→ Ejemplo: Mapa de ingresos por condado
📦 {geobr}
→ Brasil oficial | Límites, municipios, regiones
→ Ejemplo: Población por estado
📦 {rnaturalearth}
→ Mapas del mundo | Vectoriales, alta calidad
→ Ejemplo: Base para cualquier visualización global
💰 FINANZAS
📦 {tidyquant}
→ Yahoo Finance | Precios en tiempo real
→ Ejemplo: Serie histórica de S&P 500
📦 {Quandl}
→ Millones de series temporales financieras
→ Ejemplo: Precio del petróleo vs Bitcoin
🏥 SALUD
📦 {WHO}
→ OMS | Estadísticas de salud global
→ Ejemplo: Evolución de mortalidad infantil
📦 {nhanesA}
→ Encuesta de salud USA | Nutrición, enfermedades
→ Ejemplo: Relación entre dieta y salud
📊 BONUS
📦 {gapminder}
→ El dataset de Hans Rosling
→ Ejemplo: Riqueza vs esperanza de vida
Instalás el paquete, escribís 3 líneas de código, tenés los datos.
Sin buscar CSVs rotos en internet.
Sin lidiar con APIs complicadas.
Sin problemas de encoding.
¿Conoces otro paquete que te permita acceder a datos?
#RStats #DataScience #OpenData #Analytics
This book coming out in early 2026 looks interesting.
"The authors show how easily observational data analyses, experimental designs, and causal models are tweaked in ways that are extremely difficult, often impossible, to detect. As a consequence, the credibility crisis in science is even more severe than both scientists and the public believe."
“It is clear there is an intent to destroy the Palestinians in Gaza through acts that meet the criteria set forth in the Genocide Convention.”
UN Independent International Commission of Inquiry's report shows #Israel has committed #genocide in #Gaza.
https://t.co/UGjepPwmwo