🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

In medical settings, researchers & institutional users often cite medical examination benchmark results as a proxy for AI model performance across a range of medical tasks. This is a woefully misguided practice - as with actual doctors, test performance != clinical competence!

4/ It’s not just that LLMs fail on difficult IMO shortlist problems — their mistakes tend to follow consistent patterns. After a detailed review, we grouped the most common failure modes into the following categories: (Concrete examples can be found in our draft.)

I'm happy to share that I will be joining Berkeley Public Health as an associate professor this summer. Grateful for the opportunity and sad to leave the Berglas School of Economics at Tel Aviv University, my home for years as both student and faculty.

Forthcoming in AEJ: Applied Economics: "The Long-Term Impacts of Mixing the Rich and Poor: Evidence from Conscript Dorms" by Elias Einiö. aeaweb.org/articles?id=10…

For o3/o4-mini x.com/littmath/statu…

🧵How are the 2025 US tariffs affecting consumer prices? Our new paper with Paola and Franco Vazquez uses daily retail data linked to country-of-origin info to provide some evidence: 1️⃣ Rapid Pricing Responses: Prices adjusted quickly following major tariff announcements

We measure “effectiveness” of primary care providers in the VA Research along two dimensions: mental and physical health adverse events. We find the two are strongly correlated across providers and that more effective PCPs use less resources.

The Lucas Critique as it applies to AI in medicine (risk prediction models that guide interventions may not appear accurate if the interventions work) ai.nejm.org/doi/full/10.10…

Medicare spends over $1T per year, so even small cuts to wasteful spending represent huge savings. Maggie Shi and I have a new NBER on a policy targeting *just medically unnecessary physical therapy* and it would have easily have made the Department of Government Efficiency leaderboard. Quick🧵

Microsoft claims their new AI framework diagnoses 4x better than doctors. I'm a medical doctor and I actually read the paper. Here's my perspective on why this is both impressive AND misleading ... 🧵