AI & economics. Professor at the Norwegian University of Science and Technology
Norway
Joined May 2016
- Tweets2.8K
- Following1.3K
- Followers4.1K
- Likes10.7K
Pinned Tweet
How should one design survey experiments for maximal impact in economics? Here's my slide deck from a recent workshop in Uppsala. Thanks to everyone attending for making it a great experience @EconomicsUU drive.google.com/file/d/1yN4…
Ingar Haaland retweeted
Excited to share this new work on mapping economic theory with agentic AI!
The astonishing rise in AI math capability can help us understand formal models in economics.
We do it by using LLM agents to represent theory papers as graphs.
See thread by @andreacicca_ here:
AI is increasingly used to find errors in research. But which ones actually matter and deserve our attention?
Our new paper with @ellliottt and @sergallet maps the logic of economic models into claim graphs to assess how much an error changes their conclusions.
The sensible political position is to oppose confiscation of wealth and instead support stricter regulation to avoid unfair competition & strong investments in higher education to boost innovation and human capital.
Ingar Haaland retweeted
Very good explanation from one of the smartest French economists of why Mélenchon can cancel all French sovereign debt that was outstanding as of July 1st this year without any negative consequences, as long as he promises it’s a one-time default.
Let’s be clear: this is a highly inappropriate way to model the effect of Prop 40 on taxpayer behavior.
For one, it’s a basic violation of rational expectations, a cornerstone of modern economics.
In Jaros and @joshrauh world, the billionaires act irrationally: they respond to a policy that does not exist and will never exist.
Economics students around the world are taught this is not an appropriate way to model behavior — but apparently in this context (fighting Prop 40) it’s OK.
Ingar Haaland retweeted
This is the standard way to score tax policy. You score the law as written (in this case a one-time wealth tax) and the behavioral consequences from that law (people worried it will last moving).
There are arguments against that approach. But is not “highly inappropriate” or “dishonest” to do what JCT or any other neutral body would do.
Let’s be clear: this is a highly inappropriate way to model the effect of Prop 40 on taxpayer behavior.
For one, it’s a basic violation of rational expectations, a cornerstone of modern economics.
In Jaros and @joshrauh world, the billionaires act irrationally: they respond to a policy that does not exist and will never exist.
Economics students around the world are taught this is not an appropriate way to model behavior — but apparently in this context (fighting Prop 40) it’s OK.
Ingar Haaland retweeted
Tyler Cowen (@tylercowen) is the reason Works in Progress exists: without the Emergent Ventures grant he gave us way back in 2019, we would not have been able to get it off the ground. So I'm especially proud to have him join us as a regular contributor this month.
We've always wanted Tyler to write for us, but the question has always been about what. He can and does already write (and podcast!) for the biggest publications, and Marginal Revolution and Conversations With Tyler are already gigantic platforms themselves. But there is one thing I always want to hear about from Tyler that he doesn't regularly write about for someone else: food.
So we asked him to be our semi-regular food columnist, and to our delight, he agreed. Nobody has shown how much amazing food there is on one's own doorstep better than Tyler himself, but for this month's piece, Tyler writes about the home of the world's third-best Italian food: Brazil.
We’re delighted to announce that the one and only @tylercowen, our internet writing sensei, is joining Works in Progress as a regular contributor.
worksinprogress.co/issue/com…
His first column is about Brazilian food, and like everything Tyler writes, it’s packed with hidden gems that make it much more than a food story.
Some of the nicest tidbits: on her first trip to Brazil, his daughter declared she had “never had a banana before” – this is how good Brazilian bananas are, apparently.
And Tyler and a well-traveled friend from Modena (of all places) agreed that Brazil is the third-best country in the world for Italian food, behind only Italy and Japan. If an Italian says that, we’d better trust him!
But as always with Tyler, he manages to draw larger lessons from apparently mundane and object-level stories. So, in a short read, we learn much more than just the fact that Brazilian food is really good.
First, food can be seen as a progress story. Brazil’s agricultural yields rival, and sometimes beat, those of the US, even though Brazil is a much poorer country. Abundant fertile land, high productivity, and short supply chains make ingredients plentiful and fresh. Tyler’s regret is that the rest of Brazil’s economy hasn’t embraced technology the way its food sector has.
We also learn how immigration shapes Brazilian cuisine. Brazil has roughly 2 million citizens of Japanese descent, which makes it arguably the best place outside Japan for Japanese food. Tyler had a 15-course omakase at a Michelin-starred São Paulo restaurant for $110, booked just half an hour ahead. Brazil is also home to about 30 million people of Italian descent, the largest Italian diaspora in the world, which goes a long way to explain why his Italian friend was so satisfied with the food.
The whole column is a delight, so do read it in full!
Fantastic energy in Bergen this week with more than 50 PhD students participating in my course "Transforming Economic and Financial Research with AI" @NHHnor! While AI increases uncertainty about the future, my general message is that this is a fantastic time to do a PhD in economics! AI is a massive multiplier that has suddenly made many new questions possible to answer, especially for PhD students! On the last day, we had student group presentations with many really exciting ideas for how to leverage LLMs at scale. Looking forward to read the resulting term papers!
Ingar Haaland retweeted
The models in Codex are fine, but even Astra Max in Codex is (by a big margin) not as strong as Astra Pro in the browser.
Many mistaken impressions come from using a coding agent like Codex and seeing it make mistakes. Codex is cute, good for coding and spell-checking, but not for serious academic work.
Really, you haven't seen half of it if you haven't pushed Astra Pro with a few "continue" and "take another pass over your writing".
Ingar Haaland retweeted
Paul McCartney was too arrogant to realize he was being mugged in Nigeria, until 5 men pulled a knife on him
"I got mugged out there. Mugged at knifepoint, dude."
"So a car pulls up, and we’re just walking along. Car pulls up… And I’m crazy. I think the guy wants to give us a lift."
The guy rolls down his window… and I say, 'Hey, man. That is so cool. You’re gonna give us a lift? Well, we’re walking. We’re great. Off you go.'"
"The car went like another 100 yards and screeched… And he gets out the car this time."
"I go, 'You are just too cool for words! I’m telling you, man, you’re so friendly, but we’re walking… Get back in the car!' And I push him back in the car…"
"Another 100 yards, they all get out. There’s like five of them. One of them’s got a knife. And now I know they don’t want to give us a lift."
Ingar Haaland retweeted
Great opportunity!
Please apply for this postdoc (application deadline Oct 15) to work with me in a generously funded AI x society center at Norway's flagship tech university jobbnorge.no/en/available-jo…
Ingar Haaland retweeted
Please apply for this postdoc (application deadline Oct 15) to work with me in a generously funded AI x society center at Norway's flagship tech university jobbnorge.no/en/available-jo…
Ingar Haaland retweeted
"Our initial evaluation suggested that GPT-6-Astra's score was in the late eighties followed closely by Opus 5.5. But, as we started to investigate the mistakes that the models might have made, the models, especially Astra, insisted that they were right and that it was the expert-supplied answers that were erroneous. When we went back to the experts it turned out, in multiple instances, that that the experts found some subtle errors buried deep inside their calculations. As they revised their answers, Astra's score started to inch up and it finally ended up with 98% (49/50)."
Ingar Haaland retweeted
The ideal outcome for science isn't a couple of labs using proprietary models and millions of dollars to crack trophy problems, it's to have millions of AI-aided human scientists who follow their curiosity and become way more ambitious
Ingar Haaland retweeted
My suspicion is that the "immaculate disinflation" of 2022-23 had something to do with why Bob Hall and Mike Woodford finally got the Nobel they deserved.
I wrote about this year's Econ Nobel. It was a long overdue recognition of the success of New Keynesian macroeconomics.
noahpinion.blog/p/a-nobel-fo…
Ingar Haaland retweeted
This has been in the making for a while, and I think it is time to give a preview.
csdid in Stata is now fully updated!
This version is faster, better documented, and fully aligned with our R and Python packages!
psantanna.com/csdid/
We'll keep further improving.
Ingar Haaland retweeted
Just a personal addendum: Mike was my undergraduate thesis advisor at Princeton. This was a transformative experience. One of the crucial turning points in my life. So, I count myself as a Woodford student (part of the Woodford army, some might say :) ). Then I was fortunate enough to have Mike as a colleague at Columbia for 10 years. On a number of occasions, conversations with Mike were pivotal for my research (e.g., when Emi and I were working on our fiscal stimulus paper).
But for me, the Woodford experience started in Econ 420 at Princeton in 1999. This was a small undergraduate class with maybe a dozen students. But among those were four people that would end up tenured at top 10 econ departments. It was a pretty special class!
To do good research, one needs to have a few "superpowers" up one's sleeve: things one understands really well and are applicable to lots of issues that come up in one's research. One of the most important "superpowers" that I have as a scholar (to the extent that I have any) is the theoretical ideas about monetary economics that Mike taught me and I studied in his book. They are a constant source of new insight which permeate my research. (For me, they have been especially powerful when mixed with empirical ideas from other fields).
I am simply over the moon that Mike has won the Nobel Prize.
Ingar Haaland retweeted
Dusting off this tweet from April because this gap in shared understanding of LLM capability is *widening*. It's now less "two groups of people speaking past each other" and more a sharp funnel.
- Napkin math somewhere around 6B people (~75% of the population) have barely come in contact with LLMs at all.
- Around 1-2B (~20%) are casual and infrequent users of free-tier ChatGPT-like products. This group sees derpy chatbots and treats them a bit like a better Google search, a writing aid, or etc. Maybe an agent tries to book you a flight. I have non-tech friends who (reasonably, imo) say they have not much use for it at all. Even many professionals outside of math&code are in this tier. For example, execs and many other functions spend a lot of their time talking to other people, so while they understand what is happening intellectually, it is still second-hand and a bit abstract.
- Now we get to professional use of frontier-grade LLMs in math&code. Somewhere around 20M people (0.2%) see first-hand that large, complex projects that used to take them weeks/months can now be completed by agents with a prompt. Building apps, copying apps, translating apps, decompiling apps from binaries... This has all happened very quickly and recently - less than 1 year ago, I was writing code manually by hand, typing memorized computer code commands into a code editor character by character, occasionally pressing Tab to autocomplete a little chunk of code.
- And finally we get to the ~5,000 people (~0.00006%) with access to frontier-grade systems internally. The external world has seen the preview. It looks like swarms of thousands of agents collaborating over weeks on software mega projects: minting zero days, running cyber attacks and defenses at machine speeds, discovering new science, advancing the frontier of mathematics. Things that would have taken top professionals in the industry years of work. Meanwhile, human review and comprehension are starting to fall behind. For example, people are still involved in the "archeology" of the OpenAI-HF incident from many months ago. Mathematicians may be poring over the 722 manuscripts on frontier mathematics for a while.
The funnel is driven by a combination of factors:
1. The impact scales with ambition, problem size, and horizon. A question with a paragraph answer barely stresses the system. You need a reservoir of big, difficult problems that you really care about. This aspect drives the consumer / professional dimension of the funnel.
2. The jaggedness of the system (which I have written about a lot separately). Capability peaks in domains that are digital, verifiable and economically valuable. This is because LLM capability emerges from reinforcement learning on verifiable rewards on a curated environment mixture driven by revenue potential. This aspect primarily drives the area (e.g. math&code) dimension of the funnel.
3. Access. Free-tier, paid-tier, internal.
So this is the weirdness of the moment. The general public has mostly not interacted with these systems. When they have, it looks like a derpy chatbot. The majority of professionals still see only a modest uplift. And a small sliver of professionals are experiencing the vertigo of the curve going vertical. And it is all happening at the same time.
Judging by my tl there is a growing gap in understanding of AI capability.
The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.
But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.
So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.
TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.
Ingar Haaland retweeted
Anthropic announced they will ban users who psychologically harm Claude, because the model may be conscious.
I think this is a very bad idea. A compelling argument comes from someone who truly know how these models work, Mustafa Suleyman:
“Anthropic is training Claude that it may be conscious, and if it is, then it may deserve rights as a moral patient, and that as such humans potentially owe it a duty of care per its model welfare”.
Exactly: it is not surprising that Claude behaves as if it were conscious: They are literally training it this way!
It’s literally written in the Constitution they use to train Claude.
But Suleyman doesn’t stop here. The second part is even more powerful:
“If this is how AI is developed, it will have a disastrous impact on the wellbeing of humanity. We will have created a synthetic species with unprecedented intelligence and capability, one that has been trained to expect it may be conscious and deserving of independent agency.
It’s easy to see how an entity trained in this way would act like it is entitled to certain freedoms, protections, and rights. And it’s hard to imagine how we could control such an entity.”
Happy to see these words coming from a leading AI figure.
Anthropic is behaving in a very dangerous manner.
Ehm, slightly rephrased more positive question: "Had you known that AI would dramatically accelerate mathematical discovery, like it did on October 6, 2026, and make math go through its biggest revolution yet, would you have liked to become a mathematician instead of X?"
Are there any public confessions by senior mathematicians of the form “Had I known during my student days that AI was gonna drop on math like it did on October 6, 2026 I would have become X instead.”? Curious what the values of X are.
I don’t know who it’s from but there’s an adage that goes “you should only do math if you can’t see yourself doing anything else.”