Hegseth: "He [Trump] called the last 20 years of nation-building wars dumb and he's right. This is the opposite." So opposite of nation-building here we come...?
www.cbsnews.com/news/iran-op...
Hegseth: "He [Trump] called the last 20 years of nation-building wars dumb and he's right. This is the opposite." So opposite of nation-building here we come...?
www.cbsnews.com/news/iran-op...
from the NYT: "A journalist asked Mr. Trump last week who the arch is for. He responded: “Me — it’s going to be beautiful.” " www.nytimes.com/2025/10/21/u...
Trump's proposed Memorial arch
Interesting "birds" on the Trump's proposed Memorial arch en.wikipedia.org/wiki/Memoria...
Gambling everything, everywhere… while chatting with friends - this post sees such a future where we can also gamble in group chats as a good thing. sounds healthy blog.shanemac.com/the-future-o...
You can read the full working paper and methodology here - comments welcome! arxiv.org/abs/2511.11532
I find a weaker, or non-existant relationship to coverage by MSNBC and CNN (not statistically significant). As I was submitting this, a wave of new attention broke, so I will need to re-run the analysis when I have data on the new window 3/
So I thought I would test it. I put together data on Trump's Truth Social posts and TV transcripts from Fox News, CNN, and MSNBC. I find there is clear movement towards novel content in Trump's posts in the days following a spike in attention on Epstein in Fox News TV coverage. 2/
Timeline of key events and density of appearances of 'Epstein' in Fox News TV transcripts. Source: Author (CC BY 4.0): Peterson (2025), "Distracting from the Epstein files? Media attention and short-run shifts in Trump's Truth Social posts"
𝐖𝐡𝐞𝐧 '𝐄𝐩𝐬𝐭𝐞𝐢𝐧' 𝐢𝐬 𝐨𝐧 𝐅𝐨𝐱, Trump's Truth Social posts change
(new working paper)
Earlier this year, I woke up thinking about the commonly-cited claim that Trump tries to distract from the Epstein files by talking about the new ballroom, Federal troop deployments, Venezuela, etc. 1/
I hope this work sparks a conversation about shifting our focus from what's easiest for AI to teach, to what's most valuable for humans to learn.
Full paper here: arxiv.org/abs/2508.19625
The problem is worsened when considering unpriced, non-cognitive skills like persistence, which could be crowded out by an over-reliance on AI tools that are intended to make learning "easier."
The central proposition is that this mismatch grows as AI becomes more prevalent, leading schools to over-invest in skills that are becoming obsolete.
The risk is that AI encourages educational systems to specialize in precisely the wrong areas. My paper formalizes this as a coordination failure, where planners who adopt AI for its immediate teaching benefits, systematically generate a skill mismatch.
Pleased to share my new working paper, "Training for Obsolescence? The AI-Driven Education Trap." I explore a fundamental tension, motivated by a small pre-registered survey, suggesting a positive correlation between the skills AI excels at teaching and those it is most likely to automate.
The attempts at illegality are so brazen that they confuse and distort - it’s so openly unlawful that it confuses people into thinking he has the authority to do these things
It’s like the corruption - if it’s so out in the open, it must be the case that it’s lawful
Same thing here
I can only hope that there will criminal trials for Musk, Rubio, and Lewin for this crime against humanity
OpenAI rolled back their model because it was praising all kinds of crazy stuff, but the problem goes much deeper than this, as my recent article in AI and Ethics discusses. Discussion here open.substack.com/pub/elenchos...
Exactly! I have a research article about this, and a more recent write up here - open.substack.com/pub/elenchos...