Adam Ozimek (@ModeledBehavior) header
Adam Ozimek (@ModeledBehavior) profile picture

Chief economist at @InnovateEconomy. Host of the EconTwitter Water Cooler, live on twitter spaces and downloadable here: https://nitter.cf/t.co/toyjfDruRu

Lancaster, PA
Joined June 2010
Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)Photo by Adam Ozimek (@ModeledBehavior)
Adam Ozimek retweeted
I’m telling you. There’s something of a change happening after the last election. Rather than prioritizing the interests of retirees over young workers, it looks like Democrats will prioritize near-retirees instead.
Let's lower the age of Medicare eligibility to 50. I have a bill to do it. 🙋🏼‍♀️
171
209
35
5,211
374,854
Adam Ozimek retweeted
Séance may be the first NES game in history with an AI language model inside A tiny neural network (400K params) runs on the NES's own 6502 and writes every word the spirit says. No Internet. Works on a real console! Play in browser or download to NES: bxfoundry.com/seance_nes
3
5
1
25
3,796
Adam Ozimek retweeted
Starbucks is doooomed
1
3
1
12
4,603
Adam Ozimek retweeted
MIT Professor @rrwilliams reveals that OpenAI and Anthropic's AI-generated mathematical proofs contain something radically novel that researchers still can't fully explain: "I think that was just a problem in the writing, to be honest. So the writing was quite bad." "There are a group of us who understand, I would say, more or less all of the pieces. And most of those pieces we can see the origins of them. We can see what work they came from." "There's sort of one piece in I think both Anthropic and OpenAI solutions, which seems radically novel. It's still understandable. It's just a new thing and it's unclear where it came from or what it means." "At this point I would say that we have a fairly good understanding of the OpenAI proof." @MIT
24
53
10
587
117,130
Results > promises requires permitting reform In Jan 23 Biden announced $1.6B for a bridge repair. But permitting (NEPA) delayed the project until 2024, and a lawsuit delayed construction until after Biden left office A $1.6B announcement but voters never got to see results!
At an event in Kentucky, President Joe Biden (D) unites Sen. Minority Leader Mitch McConnell (R-KY), Sen. Sherrod Brown (D-OH), former Sen. Rob Portman (R-OH), Gov. Mike DeWine (R-OH) and Gov. Andy Beshear (D-KY) to tout a $1.6 billion grant to improve the Brent Spence Bridge.
4
53
3
193
11,843
Adam Ozimek retweeted
The market puts very little weight on the possibility of explosive econ growth coming soon.
Those interested in making bets on explosive economic growth due to AI can now make these on Kalshi without emailing @ben_moll. Liquidity is thin, but with the right incentives / subsidies I think these markets can be highly informative. kalshi.com/markets/kxgdpyear…
9
10
1
146
11,856
Bernie Sanders' data is misleading. The $15,000 per year is *individual income* - totally separate from a spouse who may be bringing into the household a $200,000 income and $2 millon pension. Also, the $15k excludes most govt benefits and undercounts retirement income.
Over 20% of seniors live on less than $15,000 a year. Nearly half of older workers have no retirement savings. Meanwhile, Musk pays the same into Social Security as someone earning $184,500. Absurd. We must apply the Social Security payroll tax to all income above $250,000.
3
47
6
302
12,433
a) This is wrong: the 301s (forced labor), 338s (Canada) & 232s (pharma/chips) were recent, and uncertainty is still historically high. b) His boss literally threatened Mexico ("CARAVANS") this morning. c) He's openly admitting they made harmful "wide swings" earlier.
Greer: "The feedback we get from the Democrats on the other side is that people they don't like the appearance of recklessness in the tariff program. To be clear, we haven't made wide swings in the tariff program for many months."
2
20
3
73
11,149
Hard to say what the pattern is here theargumentmag.com/p/do-mode…
35
80
11
790
74,682
Adam Ozimek retweeted
While I take the new METR report very seriously as an update toward shorter timelines, it's worth noting that significant parts of AI 2027 are not playing out — in particular, the stock market predictions. - They predicted the stock market would be up 30% in late 2026 — it's instead around 10% with 1.5 months left - They said the stock market surge would be driven by "[fictional] OpenBrain, Nvidia, and whichever companies have most successfully integrated AI assistants." - This doesn't seem right. Nvidia was up around 15% before the Jensen's scandal broke, only modestly outperforming the broader market. It has since dropped, and not just because of the scandal; there's been a wider AI draw down. - I don't think there's an equivalent of OpenBrain that is leading the stock market. Google is down; Amazon and Microsoft are doing decently well (though clearly not leading a market that's up 30% overall). - I don't think there are any clear "companies that have most successfully integrated AI assistants" that are driving stock market gains. Additionally: - The junior software engineer market is genuinely tougher, though it's unclear if it really counts as being in "turmoil" - Company revenue does look pretty strong, with e.g. Anthropic looking to earn well beyond OpenBrain's projected $35b - There are some AI-related protests planned in DC, though I'll be surprised if the 10k prediction isn't off by at least 1 OOM - Lots of people still think AI hype will blow over This is consistent with the view that while many AI forecasters have been good at predicting AI capabilities gains as measured by benchmarks, they've been less accurate at forecasting real-world impacts.
31
25
5
469
29,866
Genel olarak, Ha-Joon Chang’in Kicking Away the Ladder kitabının neredeyse tamamı, tarihsel açıdan hatalı ve yüzeysel bir okumaya dayanıyor. Aslında lise yıllarımda Chang’in çalışmalarını severdim. Ancak lisans döneminde Profesör Koyama ve Rubin ile birlikte yazdığı kitabı okurken dipnotlarda Irwin’in çalışmalarına rastladım. Ardından, Irwin’den başlayarak alakalı literatürü okumaya başlayınca; son 20 yılda iktisat tarihi literatüründe bebek sanayi korumacılığı hipotezinin ampirik olarak büyük ölçüde geçersiz kılındığını görmüş, Chang’in adeta insanları kandırdığını fark etmiştim :D Kitapta resmen doğru ve olgusal bir tane bile bir şey yok. İngiltere’deki gümrük kanunlarını bile doğru aktarmaz! (sheilagh oglieve, Chang’in bu çarpıtmasını çok iyi afişe etmişti:D) Örneğin Prusya meselesinde, gümrük vergilerinin birincil rolü oynamadığını kabul etse de Chang, bunların “olumlu etkilerini” vurgulamaktan geri durmaz. Oysa bu konudaki nispeten yeni bir çalışma, Prusya örneğinde gümrük vergilerinin etkilerini ampirik olarak ayrıştırarak devletin uyguladığı gümrük vergilerinin etkisiyle “bileşim” ve “fiyat” etkisini birbirinden ayırıyor. Bu ayrıştırmanın ardından, gümrük vergilerinin sanayileşme üzerinde kayda değer bir etkisinin kalmadığını görüyoruz:
Reading for the next iteration of my series on state capacity in Prussia and it is interesting to see where popular account such as Ha-Joon Chang get their image of Prussia as a case of state-led economic development from and why it is largely wrong.
5
26
2
98
7,279
Adam Ozimek retweeted
Paul Krugman’a blind spot on inequality everywheremillionaire.substa…
3
9
1
47
8,713
Adam Ozimek retweeted
Saddened by this. Now the bill will pass, progressives will look weak, and will, to top it off, be wrong on the merits.
NEW: The Congressional Progressive Caucus is officially opposing BAAJA "unless significant improvements are made"
28
84
1
1,487
49,004
Adam Ozimek retweeted
There are two investment booms right now. Everyone knows about one of them. The other is oil and gas. All other business investment is up at a 2% rate over the last year.
Replying to @econvitals
Real data-center-related investment rose 21 percent over the past year and oil, gas and mining 13 percent, both above the 2 percent for all other business investment. Oil and gas are a much smaller share, so even rapid growth barely moves the overall economy.
7
21
3
71
10,407
Adam Ozimek retweeted
The "editorial" parts of NEJM, Science, etc. do so much harm to public perception of science. Why do this? Are we going to publish (basically the exact same essay) from a Pentecostal preacher? Science is too important for human prosperity to do this: kevinbryanecon.com/trust.htm…
15
58
11
480
86,539
Adam Ozimek retweeted
🆕 New on Econvitals: a Macroeconomic Simulator that runs the Fed’s FRB/US model on the shock you pick, from rate hikes to tariffs. A 1-point rate hike that lasts a year: real GDP down by as much as 0.58%, unemployment up by as much as 0.32 points. econvitals.org/tools/macro-s…
1
8
3
61
33,775
Adam Ozimek retweeted
Replying to @coeff_giving
@coeff_giving is looking to support incubators for new kinds of orgs across our different areas! On abundance and growth, some examples of the kind of orgs I hope could come out of these incubators 👇
We think the world is short of thousands of high-impact nonprofits, and the people to run them. So we’re launching a Request for Proposals for incubators: organizations that identify promising ideas, recruit and train founders, and support new organizations through launch. There are two streams: 1️⃣ Scaling existing incubators: For incubators, studios, and accelerators that want to run a larger or dedicated cohort, expand into a new cause area or geography, or offer seed grants. 2️⃣ Seeding new incubators: For individuals or teams who want to start one, especially in geographies where many high-impact organizations could operate but few incubators exist, like India and Nigeria. Initial applications should take 2–3 hours to complete. Apply by December 18: coefficientgiving.org/funds/…
5
29
4
150
15,548
“Leftists are better at winning” flopped within 36 hours so they flipped to “actually the winners were leftists” but that also flopped so now six days after the election they’ve pivoted to “let’s get people angry at the guys who just won tough races.”
50
65
9
1,009
32,600
Adam Ozimek retweeted
I've been won over pretty much completely to Douthat as TV interviewer. He not only extracted more information than most people do when sitting down with this president, but he managed to drop some pretty funny asides without making the conversation about himself.
Two thoughts on this: 1. A man so consumed with hatred that he’d rather go to hell than forgive his enemies is a tragedy. I’d pity him if he was my neighbor or president. 2. This is the most interesting Trump interview I’ve seen in a long time. Feeling good about my prediction that @Douthat60 would be great on 60 Minutes.
3
14
2
279
24,400