Econ Blog

your daily dose of economic commentary

On Evaluating Understanding and Generalization in the ARC Domain

3701. On Evaluating Understanding and Generalization in the ARC Domain

Melanie Mitchell discusses the evaluation of understanding and generalization in machines using the Abstraction and Reasoning Corpus, highlighting challenges and a new benchmark called ConceptARC.

Weekly Collection of Interesting Things (5/1/2023)

3702. Weekly Collection of Interesting Things (5/1/2023)

The author shares a collection of links and resources related to econometrics, finance, and AI, highlighting various academic materials and seminars.

Running LLM AIs on Your Own: Next Steps for the Mildly Interested Passerby

3703. Running LLM AIs on Your Own: Next Steps for the Mildly Interested Passerby

Nick HK discusses how to use large language models programmatically, providing guidance for beginners interested in exploring their capabilities beyond casual use.

Weekly Collection of Interesting Things (4/24/2023)

3704. Weekly Collection of Interesting Things (4/24/2023)

Paul Goldsmith-Pinkham shares a collection of interesting links, resources, and papers related to economics and data visualization techniques.

Do half of AI researchers believe that there's a 10% chance AI will kill us all?

3705. Do half of AI researchers believe that there's a 10% chance AI will kill us all?

Melanie Mitchell examines the validity of claims regarding AI researchers' beliefs about the risk of AI causing human extinction, highlighting survey methodology issues.

Controls in Difference-in-Differences Don't Just Work

3706. Controls in Difference-in-Differences Don't Just Work

Nick HK discusses the complexities and misconceptions surrounding the use of covariates in difference-in-differences research designs, emphasizing that they require careful application.

Weekly Collection of Interesting Things (4/16/23)

3707. Weekly Collection of Interesting Things (4/16/23)

Paul Goldsmith-Pinkham shares resources and papers related to data analysis, algorithmic fairness, and experimental design in economics and machine learning.

Corporate Political Responsibility in a Captured Economy

3708. Corporate Political Responsibility in a Captured Economy

An argument that corporate political responsibility is crucial in a captured economy where businesses influence political rules to maximize profits, undermining free market principles.

Standard Causal Inference Research Designs and Animated Plots

3709. Standard Causal Inference Research Designs and Animated Plots

Nick HK discusses various econometric methods for causal inference and illustrates their effects using animated plots to enhance understanding.

Common R Errors

3710. Common R Errors

Nick HK discusses common errors encountered while using R programming, providing troubleshooting tips and solutions for beginners and educators in data visualization.

3711. Weekly Collection of Interesting Things

Paul Goldsmith-Pinkham shares a collection of interesting links and papers related to economics and finance, highlighting recent developments and research in the field.

Thoughts on a Crazy Week in AI News

3712. Thoughts on a Crazy Week in AI News

Melanie Mitchell discusses recent AI news, focusing on misconceptions from politicians and public reactions to advancements in AI technologies like ChatGPT and GPT-4.

Ordinary Least Squares, Translated into English

3713. Ordinary Least Squares, Translated into English

Nick HK explains Ordinary Least Squares (OLS) in simple terms, translating complex mathematical concepts into accessible language for better understanding.

Overdebunked! Six Statistical Critiques That Don’t Quite Work

3714. Overdebunked! Six Statistical Critiques That Don’t Quite Work

Nick HK discusses common misconceptions and critiques regarding statistical analyses, emphasizing the importance of skepticism while cautioning against unfounded criticisms.

Robustness Tests: What, Why, and How

3715. Robustness Tests: What, Why, and How

Nick HK discusses the importance and purpose of robustness tests in econometrics, emphasizing their role in validating assumptions and ensuring reliable results in data analysis.

Supercore Inflation is Worth Watching, but it is Probably Not a Good Policy Target

3716. Supercore Inflation is Worth Watching, but it is Probably Not a Good Policy Target

Ed Dolan discusses the implications of supercore inflation on Federal Reserve policy and questions its suitability as a target for monetary policy decisions.

Why the Abstraction and Reasoning Corpus is interesting and important for AI

3717. Why the Abstraction and Reasoning Corpus is interesting and important for AI

Melanie Mitchell discusses the challenges AI faces in forming concepts and abstractions, highlighting the significance of the Abstraction and Reasoning Corpus for advancing AI research.

Did ChatGPT Really Pass Graduate-Level Exams?

3718. Did ChatGPT Really Pass Graduate-Level Exams?

Melanie Mitchell discusses ChatGPT's performance on various graduate-level exams, highlighting its passing grades and the implications for AI's understanding and capabilities.

Did ChatGPT Really Pass Graduate-Level Exams?

3719. Did ChatGPT Really Pass Graduate-Level Exams?

Melanie Mitchell discusses the performance of ChatGPT on graduate-level exams, questioning the implications of AI's test results compared to human capabilities.

Why the Inflation of 2021-22 Did Not Spiral Into Stagflation and Implications for Policy Going Forward

3720. Why the Inflation of 2021-22 Did Not Spiral Into Stagflation and Implications for Policy Going Forward

The discussion highlights key differences between the 2021-22 inflation and past stagflation episodes, emphasizing implications for future economic policy.