iiinova — where your digital self keeps evolving

Reposted by Max Reith

Patrick Bernau @patrick-bernau.de · 08/02/2026

Deutschland braucht Reformen. Doch die Deutschen haben keine Lust darauf, und die Politik verliert sich in Kulturkämpfen. Wer etwas bewegen will, muss anders rangehen. Mein FAS-Kommentar: www.faz.net/aktuell/wirt...

faz.net

Deutschland und Reformen: Wie die Bundesregierung scheitert

Deutschland braucht Reformen, doch die Politik verliert sich in Kulturkämpfen. Wer etwas bewegen will, muss anders rangehen.

051

Reposted by Max Reith

Gabriel Kreindler @thetahat.bsky.social · 15/01/2026

Nobody: Economists using stated preferences:

A screenshot of an illustrated book saying: "Would you rather... jump into nettles for five dollars, ... swallow a dead frog for twenty dollars, or stay all night in a creepy house for fifty dollars?"

0244

Reposted by Max Reith

max power 🌌 @sneptech.bsky.social · 14/01/2026

in a nutshell,

623729

Reposted by Max Reith

Janis Kluge @jakluge.de · 13/01/2026

Which economy is more productive - Germany or the US? It seems the answer would be clear, but it's not. The mood in Germany is so deeply negative, even aggressively pessimistic, that I thought we had fallen behind by a lot. But here are the facts - they surprised me as well:

13515210

Reposted by Max Reith

Wojtek Kopczuk @wwwojtekk.bsky.social · 13/01/2026

The death toll is now higher than Tiananmen Square or October 7th and Iran is barely news. Not trending here, a bit more but also little on X, scroll down the feed to #9 on NYT

2225

Max Reith @maxreith.bsky.social · 18/11/2025

Gemini-3 got this wrong 5/5 times... (But this might just be reduced reasoning budgets at launch or something)

020

Reposted by Max Reith

Sarah Cohodes @cohodes.bsky.social · 10/11/2025

📣 New NBER Working Paper out today 📣 "The Consequences of Faculty Sexual Misconduct" Sarah Cohodes & Katherine Leu

Screenshot of working paper: The Consequences of Faculty Sexual Misconduct

17560196

Reposted by Max Reith

Scott Lincicome @scottlincicome.bsky.social · 10/11/2025

New @nberpubs: "The Economic Impact of Brexit" www.nber.org/papers/w34459 "by 2025, Brexit had reduced UK GDP by 6% to 8%, with the impact accumulating gradually over time." 😲

12518

Max Reith @maxreith.bsky.social · 08/11/2025

Unlike the other models used, Kimi K2 Thinking is freely available. The other models, Gemini 2.5 Pro and GPT-5 Extended Thinking, are only available through a $20 monthly subscription. So overall, Kimi K2 seems like a pretty big deal to me. (I didn’t test GPT-5 Pro, since it costs $200 per month)

001

Max Reith @maxreith.bsky.social · 08/11/2025

Sometimes the LLMs gave wrong equilibria, sometimes they wrongly claimed that there were no new equilibria at all. The inconsistency across all three models is annoying, but that’s just part of working with LLMs, I suppose 🤷‍♂️

100

Max Reith @maxreith.bsky.social · 08/11/2025

The paper is on algorithmic game theory, where I modify existing games in a specific way and examine whether new equilibrium outcomes emerge under the modified framework. I provided each model with a simple numerical example and asked whether new equilibrium outcomes arise.

110

Max Reith @maxreith.bsky.social · 08/11/2025

Another DeepSeek moment? Moonshot AI, a Chinese lab, released its new (open source!) model K2 Thinking, outperforming OpenAI et al. on several benchmarks. I tested it with a question from an unpublished paper of mine. Out of 5 tries, Kimi, GPT-5 and Gemini 2.5 Pro each replied correctly 3 times!

131

Reposted by Max Reith

dame @dame.is · 24/10/2025

chat, is this good? I scored 67 on the AI purity test. post your scores: aipuritytest.org

AI Purity Test
The AI Purity Test is a voluntary self-assessment developed by Tina Tarighian. It provides participants with a structured opportunity to reflect on the evolution of their interactions with artificial intelligence over time.
Caution: this is not a bucket list. Completion of all items on this test will likely result in death.

Your score:
67

25321

Reposted by Max Reith

Aaron Roth @aaroth.bsky.social · 21/10/2025

An interesting debate between Emily Bender and Sebastien Bubeck: www.youtube.com/watch?v=YtIQ... ---Emily's thesis is roughly summarized as: "LLMs extrude plausible sounding text, and the illusion of understanding comes entirely from how the listener's human mind interprets language. "

youtube.com

CHM Live | The Great Chatbot Debate: Do LLMs Really Understand?

YouTube video by Computer History Museum

281

Reposted by Max Reith

Rudi Bachmann @bachmannrudi.bsky.social · 18/10/2025

Dieses Streitgespräch zwischen @clemensfuest.bsky.social und @suedekum.bsky.social in der @zeit.de sollte man in Vorlesungen und Proseminaren zur Theorie der Wirtschaftspolitik durchnehmen. Sehr gutes Lehrmaterial, for the good and the bad. Ein 🧵:

38716

Reposted by Max Reith

Scott McGrath @smcgrath.phd · 15/10/2025

🧪 A new computer science conference, Agents4Science, will feature papers written and peer-reviewed entirely by AI agents. The event serves as a sandbox to evaluate the quality of machine-generated research and its review process. #MLSky

nature.com

AI bots wrote and reviewed all papers at this conference

Event will assess how reviews by models compare with those written by humans.

042

Reposted by Max Reith

Jennifer Doleac @jenniferdoleac.bsky.social · 14/10/2025

I’ve decided not to post my annual “women on the Econ job market” thread this year. Social media has splintered too much, and now that I’ve left academia I’m focused on other priorities.

15513

Reposted by Max Reith

Mauricio Drelichman @mdrelichman.bsky.social · 13/10/2025

Elated at Joel Mokyr's Nobel Prize! You can find numerous accounts -now multiplying by the minute- of his scholarly contributions. Today I want to celebrate the man and the mentor.

Joel Mokyr at the 2011 conference in his honour at Northwestern.

2418

Reposted by Max Reith

Ethan Mollick @emollick.bsky.social · 12/10/2025

I don't think people have updated enough on the capability gain in LLMs, which (despite being bad at math a year ago) now dominate hard STEM contests: gold medals in the International Math Olympiad, the International Olympiad on Astronomy & Astrophysics, International Informatics Olympiad...

712919

Max Reith @maxreith.bsky.social · 12/10/2025

These results are somewhat at odds with the mistakes Gpt and Gemini keep making when working on my proofs. I have the $20 subscription though, could that be the reason?

280

Reposted by Max Reith

Our World in Data @ourworldindata.org · 09/10/2025

How over- and underrepresented are different causes of death in the media? Another way to visualize this data is to measure how over- or underrepresented each cause is. To do this, we calculate the ratio between a cause’s share of deaths and its share of news articles.

625797

Reposted by Max Reith

Jason Furman @jasonfurman.bsky.social · 10/10/2025

The other day a student asked me about the prevalence of insider trading in prediction markets. I now have an answer.

8627162

Max Reith @maxreith.bsky.social · 03/10/2025

Wohl nicht. Any suggestions?

130

Reposted by Max Reith

Alexandra de Gendre @adegendre.bsky.social · 02/10/2025

The best post I’ve seen on Bluesky in a very long time! Brilliant idea and brilliant accounts out there !

0132

Reposted by Max Reith

Joshua Gans @joshgans.bsky.social · 30/09/2025

Back in graduate school, Paul Milgrom asked me to examine a published paper from 1984 by another person that he suspected had an incorrect proof. I found the error. I decided to see if LLMs could. Only Gemini 2.5 Pro did so. Claude Opus and GPT-5-pro found no significant errors.

1111

Max Reith @maxreith.bsky.social · 21/09/2025

Income Effect: Analyst become more productive -> hire more Substitution Effect: Fewer analysts are needed per project -> hire less. Both effects exist, it’s TBD which dominates. If a job is fully automated (AI can do all tasks), employment should def. fall (think Waymo replacing Uber drivers).

000

Max Reith @maxreith.bsky.social · 21/09/2025

I think it does help! AI today mainly augments labor: AI substitutes some tasks that analysts do, but not all. Analysts are more productive now. Does their employment rise? Depends on Income vs. Substitution effects:

100

Max Reith @maxreith.bsky.social · 19/09/2025

Want to dig deeper? This is a short summary of a paper by Phil Trammell and Anton Korinek: (www.nber.org/papers/w31815). I recently had the pleasure of taking Phil’s course on Transformative AI at Stanford DEL!

nber.org

Economic Growth under Transformative AI

Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers,…

030

Max Reith @maxreith.bsky.social · 19/09/2025

• Unlocking robots: AI-led breakthroughs might also unlock humanoid robots, bringing explosive growth via the substitution channel described in 1)

110

Max Reith @maxreith.bsky.social · 19/09/2025

Why yes: • Returns to scale: Picture one AI containing the knowledge of thousands of scientists. Unlike human teams, the AI wouldn’t face coordination costs, could parallelize research effortlessly, and tap into knowledge from multiple fields instantly, thus accelerating discovery.

130

Max Reith @maxreith.bsky.social · 19/09/2025

• Limited parallelizability: Some breakthroughs depend on earlier ones: You can't invent a car wo inventing the wheel first. Research may not scale w AI. • Physical constraints: Science needs hardware and experiments, which AI might not be able to substitute. Research might not be fully automatable

110

Max Reith @maxreith.bsky.social · 19/09/2025

2) Research Automation You've probably heard of this one: AI invents new technologies, improves itself and drives explosive growth. Could it work? Maybe... Why not: • Harder ideas: It could get harder and harder to discover new ideas. Even with AGI, the rate of discovery might go down.

120

Max Reith @maxreith.bsky.social · 19/09/2025

But imagine AI that turns capital into a substitute for labor (think robots doing most jobs). Capital could expand wo human bottlenecks, creating room for accelerated growth. How much? Depends on whether capital productivity grows too. If so, growth could take off long run, though 3000% is a stretch

230

Max Reith @maxreith.bsky.social · 19/09/2025

Do tech optimists have a point? Within standard economic growth models, AI could drive explosive growth through one of two mechanisms. 1) Labor Substitution So far, it seems like capital and labor mostly complement each other, which limits the returns to additional capital given fixed labor.

181

Reposted by Max Reith

Ethan Mollick @emollick.bsky.social · 18/09/2025

A cautiously optimistic result on AI and disinformation. A week before 2024 UK elections 13% of all voters used AI to ask about political topics. A randomized trial found this may be good: using AI led to similar gains in true knowledge as doing web research, regardless of model & prompt used.

26818

Reposted by Max Reith

Alexander Doria @dorialexander.bsky.social · 15/09/2025

> be a language model > all you see is tokens > you don't care, it's all abstracted away > you live for a world of pure ideas, chain of concepts, reasoning streams > tokens don't exist.

210512

Reposted by Max Reith

Thomas Dietterich @tdietterich.bsky.social · 14/09/2025

We need new rules for publishing AI-generated research. The teams developing automated AI scientists have customarily submitted their papers to standard refereed venues (journals and conferences) and to arXiv. Often, acceptance has been treated as the dependent variable. 1/

48525

Reposted by Max Reith

Ethan Mollick @emollick.bsky.social · 06/09/2025

We are starting to see some nuanced discussions of what it means to work with advanced AI in its current state In this case, GPT-5 Pro was able to do novel math, but only when guided by a math professor (though the paper also noted the speed of advance since GPT-4) The reflection is worth reading.

39114

Reposted by Max Reith

Grace @gracekind.net · 06/09/2025

Never ask a man his age, a woman her salary, or GPT-5 whether a seahorse emoji exists

942079421

Reposted by Max Reith

Pekka Lund @pekka.bsky.social · 16/08/2025

I like the way Anthropic approaches these questions. "We remain highly uncertain about the potential moral status of Claude and other LLMs, now or in the future. However, we take the issue seriously...Allowing models to end or exit potentially distressing interactions is one such intervention"

anthropic.com

Claude Opus 4 and 4.1 can now end a rare subset of conversations

An update on our exploratory research on model welfare

2153

Max Reith @maxreith.bsky.social · 14/08/2025

LLMs are getting better at long term reasoning. This is a big deal, and opens the door for LLMs to perform more tasks in the real world.

130

Reposted by Max Reith

Ethan Mollick @emollick.bsky.social · 08/08/2025

Suddenly retiring every other model without warning was a weird move by OpenAI … and they did it without explaining how switching models worked or even details of various GPT-5 models …and they did it after many built workflows & training & assignments around older models, maybe breaking them. Odd

1411315

Reposted by Max Reith

Paige Bailey (webpaige.dev) @dynamicwebpaige.bsky.social · 06/08/2025

Reminder: DeepMind's Gemma 3n model is performing about as well as Gemini 1.5 Pro – better, in many cases! – and is only 4B parameters in size. The best model 6 months ago is now small enough to be run on a laptop. Now play that forward 6 months from now, thinking about the best models of today. 🤯

1132

Reposted by Max Reith

Jessica Leight @leightjessica.bsky.social · 31/07/2025

Fascinating new paper shows that papers reporting statistical significance get at least 60% more media attention #econtwitter #econsky: from Brodeur Cook @nikolaimcook @taylor_wright, a short 🧵

165

Max Reith @maxreith.bsky.social · 28/07/2025

Pretty impressive! It caught things that o3 and Gemini didn't. Wish I had had this before my submission...

061

Reposted by Max Reith

Helen Toner @hlntnr.bsky.social · 22/07/2025

AI companies are starting to build more and more personalization into their products, but there's a huge personalization-sized hole in conversations about AI safety/trust/impacts. Delighted to feature @mbogen.bsky.social on Rising Tide today, on what's being built and why we should care:

1156

Reposted by Max Reith

Helen Toner @hlntnr.bsky.social · 23/04/2025

New on Rising Tide, I break down 2 factors that will play a huge role in how much AI progress we see over the next couple years: verification & generalization. How well these go will determine if AI just gets super good at math & coding vs. mastering many domains. Post excerpts:

162