Paper from OpenAI says hallucinations are less a problem with LLMs themselves & more an issue with training on tests that only reward right answers. That encourages guessing rather than saying “I don’t know” If this is true, there is a straightforward path for more reliable AI.

A little flavor of the speed of AI compared to previous disruptive technology...

We have now made public our AI Retrainability (AIR) index values and occupation rankings by both 2-digit and 6-digit SOC. Public link here: tinyurl.com/mvuku2td

I finally understand my paper.

I agree! There is also potentially greater scope for policy ideas like AI insurance (Ioana Marinescu) or Automation Adjustment Assistance (sam manning ) to work relative to TAA. While trade shocks are spatially concentrated, AI shocks are likely disperse --> fewer adjustment frictions.

Mostly harmless constant returns to scale pecuniary multipliers.

What happens when online job applicants start using LLMs? It ain't good. 1. Pre-LLM, cover letter quality predicts your work quality, and a good cover gets you a job 2. LLMs wipe out the signal, and employer demand falls 3. Model suggests high ability workers lose the most 1/n